nternat,onal p on Laser nstrumentat,on

0 downloads 0 Views 32MB Size Report
May 18, 1992 - 8-1. Geometric. Analysis of Satellite. Laser Ranging. Data,. B. Conklin et al., BFEC ..... 8- .... Space Flight Center (GSFC) 1.2 Meter Telescope. Facility, ..... David R. Edge. BFEC ...... at 40 rpm. If the laser fire timing is controllable ...... manual interface. The median is good basis for a filtering program, because.
Conference

in

Publication

3214

nternat,onal p on Laser nstrumentat,on ternationai Workshop Annapolis, Maryland May 18-22,1992 (NASA-CP-3214)

EIGHTH

INTERNATIONAL WORKSHOP RANGING INSTRUMENTATION 741 p

ON

LA3ER (NASA)

--THRU-N94-15552N94-15625 Unc|as

HI/19

0171410

!

=

ir

_rT q_[1,;.....................

;,-,,

.........

!!'!

i ii

= _,= ;_-:

::

iiiiii'_

, ......

, ...............

=

=l_r

i

_i

---

;_

NASA

Conference

Publication

3214

Eighth International Workshop on Laser Ranging Instrumentation

Compiled

and Edited

by

John J. Degnan Goddard

Space

Flight

Greenbelt,

Center

Maryland

National Aeronautics and Space Administration Goddard Space Flight Center Greenbelt, Maryland 20771 1993

Proceedings

of on

the

Laser

Eighth

International

Ranging

Annapolis,

Instrumentation Maryland,

May

18-22,

TABLE

Workshop

USA

1992

OF CONTENTS Page

Foreward

......................



No. . vii ,.,

List

of Participants

Workshop

.........................................

Agenda

vm

..........................................

xv

Scientific Applications and Measurements Requirements Laser Tracking for Vertical Control, P. Dunn et al., Hughes Laser Ranging NetwOrk Performance Determination at D-PAF, F.-H. Laser

Ranging

Satellite

Application

to Time

and to Other

Applications

ofSLR,

University

and Routine Orbit Massman et al., DGFI Transfer

Japanese

Space

B. E. Schutz,

Using

of Today's

Satellite

Greenwich

Laser

Observatory

Adaptive Median M. Paunonen,

Filtering Finnish

SATCOP

Planning

Laser

for Space

Mission

Ranging

et al.,

Research,

CRL

.....

1-34

..............

1-43

Royal Greenwich Graz ...............

Systems,

P.J.

Dunn

Obs .....

2-1 2-15

et al., 2-23

of Data

Distribution,

A.T.

Sinclair,

................................

for Preprocessing Geodetic Institute Software

Package,

2-34

of Time Series .......................... S. Bucey,

Measurements, 2-44

BFEC

..............

2-51

Technology

Nd: YLF Laser for Airborne/Spaceborne NASA/GSFC ............................................ Alternative

Wavelengths

for

New Methods of Generation et al., Czech Technical Simultaneously Train, An

1-19

of Texas

Hughes STX ........................................... SLR Data Screening," Location of Peak Royal

1-1

.................

H. Kunimori

Timely Issues Satellite Signatures in SLR Observations, G.M. Appleby, Work at Graz on Satellite Signatures, G. Kirchner, SLR The Precision

.............

Geodetic

Programs,

Center

STX

University

Xiangchun

Light

Source

Ranging,

Ranging,

J.L.

of the Passively et al., for

Shanghai

Laser

Ranging,

Dallas

et al., 3-1

K. Hamal,

Czech

of Ultrashort Laser Pulses for University ...............................

Compression

Yang

Improved

Laser

Laser

Mode-Locked Institute

Ranging,

Univ ......

3-7

H. Jelinkova 3-9

Pulsewidth

of Optics

K. Hamal

Technical

et al,

and

and Fine Czech

Pulse

Mechanics

. .

3-15

Technical

.............................................

3-19 iii PI_CEI:]_flG

PAGE

BLANK

NOT

FILMED

Epoch

and Event

Preliminary

Timing

Results

from

lntercomparison Detector

the Portable

Standard

with MOBLAS-7,

Satellite

Laser

et al.,

BFEC

M. Seldon

Ranging ................

Technology

Performance S. Cova

Optimization of Detector Electronics for Millimeter et al., Politecnico di Milano .............................

Tracking Capabilities of SPADs for Laser Ranging, Politecnico di Milano ...................................... How to Squeeze High Quantum A. Lacaita et al., Politecnico The Solid

State

Detector

Technology

for

Technical

Calibration

University

Ranging, 5-1

F. Zappa et al., 5-19

Picosecond

Laser

out of a SPAD, 5-25

Ranging,

I. Prochazka, 5-31

Atmospheric

T. Varghese et al., BFEC ................................... The First Satellite Laser Echoes Recorded on the Streak Czech

Laser

Efficiency and High Time Resolution di Milano ...........................

Czech Technical University .................................. Streak Camera Based SLR Receiver for Two Color

Measurements, 5-36

Camera,

K. Hamal

5-47

Techniques/Targets MTLRS#1

USSR

Campaign,

P. Sperber

Test Results

from

LAGEOS-2

Optical

Characterization

Using

Pulsed

T. Varghese et al., BFEC ................................... Analysis of TOPEX Laser Retroreflector Array Characteristics, MOBLAS

Multiwavelength

Syst6m

Characterization,

Ranging/Streak

Wavelengths

for

Two

V. Husson,

Atmospheric

U. Schreiber Millimeter

Color

TWo Wavelength Technical New

Dispersion

et al.,

Accuracy

Ranging,

J. Degnan,

Satellite

University

with WLRS

Fundamentalstation

Satellites

for

Laser

Wettzell

Two Color

Ranging

in Multiple

Data

State-of-the-Art

Analysis/Model

Using SPAD,

Lasers,

T. Varghese,

BFEC

6-33 6-47

BFEC

. . . ..........

6-59

NASA/GSFC 1.2m

.........

Telescope

7-1

Facility, 7-15

Wavelength

Mode, 7-28

J. Degnan,

NASA/GSFC

I. Prochazka

et al.,

. . .

7-36

Czech

......................................

7-52 Generation

Geodesic

Satellites, 7-56

Errors Laser

Range

Modeling

Applications, S.M. Klosko Geometric Analysis of Satellite

et al., Laser

Hughes Ranging

Improvement

Satellite

6-9

.....................

Ranging,

Perspectives for High Accuracy SLR with Second G. Lund, AEROSPATIALE ..................................

SLR

Mironov

Cameras

TWo Color Satellite Laser Ranging Upgrades at Goddard's T. Zagwodzki et al., NASA/GSFC ............................. Measuring

et al., 6-1

ETALON-1, -2 Center of Mass Correction and Array Reflectivity, N.T. et al., Main Astron. Obs. of the Academy of Sciences ..................

Optimum

et al.,

..................................

Experience and Results of the 1991 IfAG .................................................

Historical

4-1

of SLR Accuracy,

A Possible

for

Geodetic

and Oceanographic

STX ....................... Data, B. Conklin

New Step,

iv

M. Kasser,

et al., ESGT

BFEC .........

.....

8-1 8-15 8-23

Operational Software Developments On the Accuracy of ERS-1 Orbit Predictions,

R. Koenig

etal.,

DGFI

...........

9-1

Compensation for the Distortion in Satellite Laser Range Predictions Due to Varying Pulse Travel Times, M. Paunonen, Finnish Geodetic Institute .............. Timebias Corrections to Predictions, R. Wood Herstmonceux Castle ...................................... Formation Poisson

of On-Site Filtering

Normal

of Laser

Points,

Ranging

G.M. Data,

Appleby R.L.

Networking

NASADOSE Laser Ranging System McDonald Observatory ........... Operational

Lunar

Laser

A. Novotny,

Streamlining,

Application of the Robust Shanghai Observatory

et al.,

Ricldefs

Computer

Upgrade

Satellite

Laser

Ranger

Royal

etal.,

Czech

Greenwich

McDonald

Technical

Univ .......

Control Computers, .....................

D. Edge

etal.,

Estimate in SLR Data .....................................

BFEC

Obs...

Obs .....

R.L.

McDonald

x-y

Offset

Observatory

9-49

T. Detong

etal., 9-57

Guiding

Stage for

the MLRS,

P.J.

Shelus

et al.,

.....................................

LLR-Activities

in Wettzell,

Fixed

Upgrades/Developments

Station

U. Schreiber

Matera Laser Ranging Observatory BFEC ............................................... Performance

of the

Orroral

Geodetic Ranging

10-1

Upgraded Other

of Shanghai

Shanghai

Observatory

Status-Report

on WLRS,

Based

Research New

Laser

Xia

for

Laser

Technology

Upgrades/Developments Upgrades,

Results

of the MTLRS-1

Observatory

Luck,

G. Kirchner

Station,

etal.,

Space

SLR Graz

R. Eichinger

IfAG

Research 11-37

F.M.

Yang

et al.,

.......................

Location,

G.C.

11-51

Gilbreath

etal.,

Naval

Upgrade,

Receiving Transmitting Control

for Satellite

11-54

at Wuhan

P. Sperber

System, System,

System

et al.,

for

Geodesy

. . 11-31

11-44

Satellite

of Seismology

System

Based

J. Mck.

S. Schillak et al., ...........................

et al.,

Syslem

The new MTLRS

System,

11-6

in Graz,

Ranging

Satellite

TLRS-3

The new MTLRS#1

Ranging

...................................... Institute

10-14

et al.,

.................................

Mobile

Transputer

Laser

Improvements

R. Dassing

Ranging

et al.,

T. Varghese

.......

.....................................

of Ranging

Zhizhong

An Overview,

10-6

Wettzell

11-1

Satellite

Laboratory

Progress

Fundamentalstation

(MLRO);

Orroral

Observatory and

et al.,

of the Borowiec Laser Station, of Polish Academy of Sciences

Development

Ground

9-33 9-43

Lunar Laser Ranging Data Processing in a Unix/X Windows Environment, R.L. Ricldefs etal., McDonald Observatory ........................

Upgrading Center

9-26

Ricldefs

...............

Preprocessing,

9-19

Ranging

A Computer-Controlled

SUB-CM

9-9

Group, 9-13

Upgrading etal., HP

at SLR Stations,

etal.,

Laser

Ranging

........................

BFEC et al.,

11-60

.................... IfAG

Station,

12-1

.................

12-17

P. Sperber

et al.,

IfAG

...............

12-26

P. Sperber

et al.,

IfAG

...............

12-33

MTLRS,

E. Vermaat

..............................

V

et al.,

Kootwijk 12-40

Airborne

and Spaceborne

Airborne

2 Color

Systems

Ranging

Experiment,

P.S.

GLRS 2-Colour Retroreflector Target Design AEROSPATIALE ........................................ Effects

of Turbulence

NOAA

Wave

on the Geodynamic

Propagation

Millar

Ranging

Checkout

et al., Mars

Equipment

NASA/GSFC

Laser

Altimeter

J.H.

Chumside,

Spaceborne

13-33

(MOLA),

B.L.

Johnson

Systems,

J.C.

Jr. et al., 13-49

Laser

Altimeter

Smith

on a Single

13-52

Photon

Ranging

Technique,

University .............................. J.L. Burton, NASA/GSFC

I. Prochazka 13-74 13-78

...................

Presentations

Satellite

Laser

Station

University

Helwan

The Third

Generation

Performance

McGarry

Summaries

Scientific Timely

Cech

et al.,

No.

and

7836,

Speed

14-14

Microchannel

Space

et al.,

Flight

Plate

Station

Mobile

(GSFC)

Conference Business

1.2 Meter

........................

Ranging

Requirements

Sununary/Resolutions Workshop

14-29

...................

15-2 15-6 15-7

...........................

................................ ...............................

........................................

Upgrades/Developments

Telescope

15-5

Cameras

Upgrades/Developments

Meeting/Next

Tubes,

15-1

......................................

Ranging/Streak

System

Photomultiplier

14-20

Center

NASA/GSFC

and Measurements

Timing

et al.,

...................................

SLR Data Analysis/Model Errors Operational Software Developments Fixed

Systems,

H. Fischer

Detector Technology ........................................ Calibration Techniques/Targets .................................. Multiwavelength

TLRS

...............................

..........................................

and Event

Laser

Technical

14-2

.............................................

Technology

Lunar

Czech

.........................................

Applications Issues

Epoch

BFEC

on the Goddard

J.F.

Potsdam

of High

et al.,

Report

Session

M.

14-1

Potsdam

Comparison

T. Varghese Facility,

1992,

Upgrades, MOBLAS ...................................

SLR Station

GeoForschungsZentrum

Station

Status

.............................................

Optical Attenuation Mechanism R. Eichinger et al., BFEC

Laser

System,

...................................... Based

et al., Czech Technical Multi-Beam Laser Altimeter, Poster

for

13-1

G. Lurid,

............................

Development of the Mars Observer Laser Altimeter NASA/GSFC ..... ....................................... Bench

.........

Performance,

13-17

Laser

Laboratory

et al., NASA/GSFC

and Predicted

15-8 15-9 15-10 15-11 15-12 15-13

.............................. ............................

15-14 15-15

................................

16-1

...............................

17-1

vi

FOREWORD

At long last, As Chairman,

the Proceedings of the Eighth International Workshop are I tried very hard to obtain 100 percent of the presentations

"ready" for publication. in printed form so that

they could be distributed in these proceedings. In spite of the fact that the original submission deadline of 1 August 1992 was extended twice into early 1993 and numerous personal contacts were made, there are still several fine papers missing. Nevertheless, this volume contains the vast majority

of the presentations,

I desperately

wanted

workshop

to avoid

in Australia.

and I felt I could the embarrassment

Thank

publication these

any longer. proceedings

Besides, at the

1994

you to all who contributed.

One does not take on the job of chairing

a major

I wish

a number

to take

not delay of distributing

this opportunity

to thank

international

meeting

of people

who

without

made

a lot of help,

the Annapolis

and

meeting

a success. A special

thank

you

goes

to Miriam

Baltuck

and

Joe

provided funding support for the meeting. Not only overall success of the workshop, but it permitted greater

Engeln

at NASA

did this contribute participation from

Headquarters

who

substantially to the many of our foreign

colleagues. I alse

wish

to express

my thanks

to Karel

kindly offered his laboratory as a meeting to Ivan Prochazka for serving as unofficial also

to Program

their

busy

I am grateful papers include

Committee

schedules

members

and coming

Hamal

site for the Program recording secretary

Christian

to Prague

also for the support

Veillet

to help

of the session

and

plan

Thanks

also to Sarah

Wager

and Deborah

in selecting

the site for the meeting,

and general

coordinating

Last,

but certainly

during the entire itself (in countless

through

their

Committee in January during our deliberations. Ben

chairmen

who

for taking

1992, and Thanks time

were

responsible

from

for soliciting

presented in their sessions. Ben Greene, Tom Varghese,

of Westover

with hotel

and travel

I want to thank

my secretary,

Mrs.

effort - from mailing the initial ways), through the preparation

Consultants

for their

arrangements

final

distribution.

International

Workshop

I couldn't

have

done

it without

I. Degnan on Laser

Ranging

vii

Diana

Elben,

circulars, through of the proceedings

Chairman Eighth

Greene

who

These Jean

assistance

for the meeting,

activities.

not least,

support meeting

of Prague,

Georg Kirchner, Christian Veillet, Erik Vermaat, and Richard Eanes (for standing in on occasion).

Williams

helping

University

the workshop.

and for organizing and summarizing the material Bob Schutz, Andrew Sinclair, Helena Jelinkova,

Gaignebet, Karel Hamal, Ron Kolenkiewicz, Jim Abshire, Mike Pearlman, Carroll Alley,

John

of the Technical

Instrumentation

her.

for her wonderful supporting the for publication

8TH INTERNATIONAL WORKSHOP ON LASER RANGING INSTRUMENTATION MAY

James

Abshire

NASA/Goddard Code 924

Space Flight

Cntr.

Greenbelt, MD 20771 USA Phone: 301-286-2611 Robert

Greenbelt, USA

Space Flight

MD

20771

AI-Hussain

428-2 Ridge Road Greenbelt, MD 20770 USA Phone: 301-474-4787 Carroll Alley Department of Physics University of Maryland College Park, MD 20742 USA Phone: 301-405-6098 Fax:

30!-699-9195

OF

PARTICIPANTS

Wiard

Beek

Baltuck

NASA Headquarters Code SEP-05 Washington,

D.C.

Cntr.

Tammy Bertram NASA/Goddard Code 726.1

20541

Cntr.

Greenbelt, MD 20771 USA Phone: 301-286-8119 Fax: 301-286-2429

Spaziale

Jim Churnside

Space Flight

Cntr.

Brion Conklin BFEC 10210 Greenbelt Rd/Suite Seabrook, MD 20706 USA

Rd/Suite

700

Greenbelt, USA

MD

Space Flight 20771

Phone: 301-286-8591 Fax: 301-286-9200

.°,

Vlll

700

Phone: 301-794-3510 Fax: 301-794-3524

Seabrook, MD 20706 USA Phone: 301-794-3466 Fax: 301-794-3524

NASA/Goddard Code 920

NOAA - Wave Propagation R/E/WP1,325 Broadway Boulder, CO 80303 USA Phone: 303-497-6744 Fax: 303-497-6978

Greenbelt, MD 20771 USA Phone: 301-286-7052 Fax: 301-286-4943

Via Ospedale 72 Cagliari, 09124 ITALY

Boulevard

Lanham, MD 20782 USA Phone: 301-286-4076 Fax: 301-286-1620

John Bosworth

Steven Bucey BFEC 10210 Greenbelt

Jean Eugene Chabaudie Ave. Copernic Grasse, F06130 FRANCE Phone: 33-93-365869 Fax: 33-93-368963 Jack Cheek 4400 Forbes

Matera, 75100 ITALY Phone: 39-835-377209 Fax: 39-835-339005

Jack L. Bufton

39-70-72-5246

Space Flight

Giuseppe Bianco Centro di Geodesia P.O. Box 11

Cenci

via Tiburtina, Rome, 965 ITALY Phone: 39-640693861 Fax: 39-640693638

Apeldoorn, 7300 AN NETHERLANDS Phone: 31-5769-8212 Fax: 31-5769-1344

Aldo Banni

Phone:

Alberto

Kootwijk Observatory P.O. Box 581

NASA/Goddard Code 901

G.M. Appleby Royal Greenwich Observatory Madingley Road Cambridge, CB30EZ ENGLAND Phone: 44-223-37437 Fax: 44-223-374700 Miriam

LIST

Afzal

NASA/Goddard Code 924

Fahad

18 - 22, 1992

Sergio Cova p.z.a. Leonardo da Vinci Politecnico di Milano Cntr.

Milano, 20133 ITALY Phone: 39-2-23996103 Fax: 39-2-2367604

32

WilliamCrawford

Kenneth S. Emenheiser BFEC 10210 Greenbelt Rd/Suite

Seabrook, MD 20706 USA Phone: 301-794-3495 Fax: 301-794-3524

John J. Degnan NASA/Goddard Space Flight Cntr. Code 901 Greenbelt, MD 20771 USA Phone: 301-286-8470 Fax: 301-286-4943

Don Cresswell

Domenico

BFEC 10210 Greenbelt

Centro Spaziale Di Matera ITALY Phone: 39-835-334951 Fax: 39-835-3771

Joe Engeln NASA Headquarters Code SEP-05

BFEC 10210 Greenbelt

Rd/Suite

Rd/Suite

700

700

Seabrook, MD 20706 USA Phone: 301-794-3493 Fax: 301-794-3524 Henry A. Crooks BFEC 10210 Greenbelt Rd/Suite Seabrook, MD 20706 USA Phone: 301-794-3500 Fax: 301-794-3524

Howard Donovan BFEC 10210 Greenbelt Rd/Suite 700

Peter

Cntr.

Wettzell

David R. Edge BFEC 10210 Greenbelt

Rd/Suite

Winfield

Richard Eichinger BFEC 10210 Greenbelt Rd/Suite

Seabrook, MD 20706 USA Phone: 301-794-3474 Fax: 301-794-3524

Ave. Copemie Grasse, F06130 FRANCE Phone: 33-93-365849 Fax: 33-93-092613 Thomas

700

Fischetti

2609 Village Lane Silver Spring, MD 20906 USA Phone: 301-871-2425 Fax: 301-871-0269

Cntr. for Space Research University of Texas - Austin Austin, TX 78712-1085 USA Phone: 512-471-5573 Fax: 512-471-3570

Koetzting, 8493 GERMANY Phone: 49-9941-603112 Fax: 49-9941-603222

Rd/Suite

Feraudy

Eanes

Seabrook, MD 20706 USA Phone: 301-794-3474 Fax: 301-794-3524

M. Decker

Dominique CERGA

Dunn

Richard

Greenbelt, MD 20771 USA Phone: 301-286-6041

BFEC 10210 Greenbelt

700

Hughes STX 4400 Forbes Blvd. Lanham, MD 20905 Phone: 301-796-5036 Fax: 301-306-1010

Space Flight

Reiner Dassing Fundamental Station

Washington, D.C. 20541 Phone: 202-453-1725 Fax: 202-755-2552

Seabrook, MD 20706 USA Phone: 301-794-3491 Fax: 301-794-3524

E. Cuot Avenue Nicolas Copemic Grasse 06130, FRANCE Phone: 33-93-126270 Fax: 33-93-092615 Joseph Dallas NASA/Goddard Code 726.1

Del Rosso

700

Seabrook, MD 20706 USA Phone: 301-794-3495 Fax: 301-794-3524

Seabrook, MD 20706 USA Phone: 301-794-3508 Fax: 301-794-3524

ix

700

J.C. Gaignebet GRGS/CERGAIOCA Av Copernic Grasse, F06130 FRANCE Phone: 33-93-365899 Fax: 33-93-368963 Virgil F. Gardner NASA/Goddard Space Flight Code 901 Greenbelt, MD 20771 USA Phone: 301-286-8437 Fax: 301-286-4943 Luciano

700

Garramone

Centro Spaziale Di Matera ITALY Phone: 39-835-334951 Fax: 39-835-3771

Cntr.

G.C.Gilbreath Code 8133

Jean Louis Hatat GRGS/CERGA/OCA

Michel Kasser 18 Allee Jean Rostand

4555 Overlook Avenue, S.W. Washington, DC 20375 USA Phone: 202-767-2828 Fax: 202-767-1317

Avenue Nicolas Copernic Grasse 06130, FRANCE Phone: 33-93-126270 Fax: 33-93-092615

Evry Cedex, 91025 FRANCE Phone: 33-160780042 Fax: 33-160779699

Carl Gliniak

J. Michael Heinick BFEC 10210 Greenbelt Rd/Suite

Waldemar EER Systems Corporation 10289 Aerospace Road Seabrook, MD 20706 USA Phone: 301-306-7840 Fax: 301-577-7493 Ben Greene Electro Optic Systems 55A Monaro Street

Pty. Ltd.

Van S. Husson BFEC 10210 Greenbelt

Ludwig Grunwaldt Telegrafenberg A 17 Potsdam, 0-1561 GERMANY Phone: 37-331-0325 Fax: 37-332-2824

Seabrook, USA

Graz, A-8042 AUSTRIA Phone: 43-316-472231 Fax: 43-316-462678

of Sciences

MD

Rd/Suite

Steve

Helena

700

20706 Rol f Koenig DGFI

Jelinkova

Technical University Brehova 7

Frascati, 00044 ITALY Phone: 49-6-94180386 Fax: 49-6-94180361

of Prague

115 19 Prague 1 CZECHOSLOVAKIA Phone: 42-2-84-8840 Fax: 42-2-84-8840 of Prague

Seabrook, USA

MD

20706

301-794-3495

S. Jessie

NASA/Goddard Code 901

115 19 Prague I CZECHOSLOVAKIA Phone: 42-2-84-8840 Fax: 42-2-84-8840 William Hanrahan III BFEC 10210 Greenbelt Rd/Suite

Lawrence

Space Flight Cntr.

Ronald

Bert Johnson 700

Lebedev Physical Institute Russian Academy of Sciences Leninski Prospect Moscow, 53117924 RUSSIA Phone:

Greenbelt, MD 20771 USA Phone: 301-286-2052

NASA/Goddard Code 924

Marstallplatz 8 800 Munich 22 GERMANY Phone: 49-8153-281353 Fax: 49-8153-281207 Yuri Kokurin

Hamal

Technical University Brehova 7

Klosko

Hughes STX 4400 Forbes Blvd. Lanham, MD 20706 Phone: 301-794-5284 Fax: 301-306-1010

Phone: 301-794-3470 Fax: 301-794-3524

Paci Guido ESA/ESRIN via G. Galilei

Phone:

Feng Hesheng Chinese Academy

Georg Kirchner Lustbuhelstr 46

P.O. Box 110, Kunming Yunnan Province, PEOPLE'S REP. OF CHINA Phone: 86-0871-72946 Fax: 86-0871-71845

Queanbeyan, NSW 2820 AUSTRALIA Phone: 61-6-2992470 Fax: 61-6-2992477

Karel

700

Seabrook, MD 20706 USA Phone: 301-794-3469 Fax: 301-794-3524

Kielek

Warsaw University of Technology Faculty of Electronics Warsaw, 00667 POLAND Fax: 48-22255248

Space Flight

Greenbelt, MD 20771 USA Phone: 301-286-6179

Cntr.

7-95-132-7147 Kolenkiewicz

NASA/Goddard Code 926

Space Flight

Greenbelt, USA

20771

MD

Phone: 301-286-5372 Fax: 301-286-9200

Cntr.

DannyKrebs NASAIGoddard Space FlightCntr. Code726.1 Greenbelt, MD 20771 USA Phone:301-286-7714 HirooKunimori 4-2-1Nukui-Kita-Machi Koganei Tokyo,184 JAPAN Phone: 81-423-27-7560 Fax:81-423-21-9899 MauriceLaplanche

Grant Moule

Michael Maberry Institute for Astronomy P.O. Box 209

Electro Optic Systems 55A Monaro Street

Kula, HI 96790 USA Phone: 808-878-1215 Fax: 808-878-2862

Queanbeyan, NSW AUSTRALIA Phone: 61-6-2992470 Fax: 61-6-2992477

Paul Malitson BFEC 10210 Greenbelt

Alan Murdoch BFEC 10210 Greenbelt

Rd/Suite

700

Pty. Ltd.

Rd/Suite

Seabrook, MD 20706 USA Phone: 301-794-3505 Fax: 301-794-3524

Seabrook, MD 20706 USA Phone: 301-794-3497 Fax: 301-794-3524

Jean Francois CERGA/OCA

Reinhart

700

CERGA/OCA Ave. Copernic Grasse, F06130 FRANCE Phone: 33-93-426270 Fax: 33-93-092613

Franz-Heinrich

AUSLIG, P.O. Box 2 Belconnen, ACT 2616 AUSTRALIA Phone: 61-6-2357285 Fax: 61-6-2575883

Fax: 301-286-4943

Pty. Ltd.

Queanbeyan, NSW AUSTRALIA Phone: 61-6-2992470 Fax: 61-6-2992477

Aerospatiale CA/TO/I 100 Blvd, du Midi 06322 Cannes La Bocca FRANCE Phone: 33-9292-7856 Fax: 33-9292-7190

Pamela

Cntr.

2820

Millar

NASA/Goddard Code 924

Space Flight

Greenbelt, MD 20771 USA Phone: 301-286-3793 Fax: 301-286-2717 Joseph Miller 1130 Freeland

Road

Freeland, MD 21053 Phone: 410-357-5818

xi

Space Flight

Greenbelt, MD 20771 USA Phone: 301-286-9283 Fax: 301-286-4952

Timothy May Electro Optic Systems 55A Monaro Street

Lund

Jan McGarry NASA/Goddard Space Flight Code 901 Greenbelt, MD 20771 USA Phone: 301-286-5020

Carey Noli NASA/Goddard Code 935

Massmann

Pfarrangerweg 4 Petershalisen, D-8037 GERMANY Phone: 49-8137-5965 Fax: 49-8153-28-1207

John Luck

Neubert

Telegrafenberg A 17 Potsdam, 0-1561 GERMANY Phone: 37-331-0325 Fax: 37-332-2824

Ave. Copernic Grasse, F06130 FRANCE Phone: 33-93-365849 Fax: 33-93-092613

Dr. Kasimir Lapushka Riga SLR Station University of Riga Riga LATVIA

Glenn

Mangin

Cntr.

Antonin Novotny Technical University Brehova 7 115 19 Prague 1 CZECHOSLOVAKIA Phone: 42-2-84-8840 Fax: 42-2-84-8840

of Prague

Jacek Offierski P.O. Box 581 7300 An Apeldoorn NETHERLANDS Phone: 31-5769-8211 Fax: 31-5769-1344 Thomas Oldham BFEC 10210 Greenbelt

Rd/Suite

Seabrook, MD 20706 USA Phone: 301-794-3499 Fax: 301-794-3524

700

Cntr.

Klaus Otten P.O. Box 581

Francis Pierron OCA/CERGA

7300 An Apeldoom NETHERLANDS Phone: 31-5769-8211 Fax: 31-5769-1344

Ave N. Copemic Grasse, F06130 FRANCE Phone: 33-93-365849 Fax: 33-93-092613

Linda

Gary D. Robinson BFEC%CDSLR, Suite 750 10210 Greenbelt Road Seabrook, MD 20706 USA Phone: 301-794-3467 Fax: 301-794-3524

Pacini

NASA/Goddard Code 726

Space Flight

Norris Laboratory Ave.

Washington, D.C. 20375-5000 USA Phone: 202-767-2828 Fax: 202-767-1317

Jocelyn Paris CERGA Avenue Nicolas Copemic Grasse, 06130 FRANCE Phone: 33-93-126270 Fax: 33-93-092615 Kamoun Paul Le Rocazur, Rue Cntr.so Nice, 06100 FRANCE Phone: 33-92-92-7517 Fax: 33-92-92-7620 Matti Paunonen Ilmalankatu IA

E. Pop Sidlerstr.

5

Bern, 3012 PEOPLE'S REP. OF CHINA Phone: 4131658591 Fax: 4131653869

Technical University Brehova 7

of Prague

115 19 Prague 1 CZECHOSLOVAKIA Phone: 42-2-84-8840 Fax: 42-2-84-8840

Rd/Suite

700

Peter Pendlebury MOBLAS-5 Tracking P.O. Box 137 Dongara, 6525 AUSTRALIA Phone: 61-99-291011 Fax: 61-99-291060

Station

Randall

Stanislaw Schillak Astronomical Latitude Borowiec 91

Observ.

Komik, 62-035 POLAND Phone: 48-61-170187 Fax: 48-61-170219

Wettzell

Kotzting, Munich, 8493 GERMANY Phone: 49-9941603113 Fax: 49-9941-60322

Cntr. for Space Research University of Texas Austin, TX 78712 USA Phone: 512-471-4267 Fax: 512-471-3570 Bernard Seery NASA/Goddard Code 726

Street

Cambridge, MA 02138 USA Phone: 617-495-7481 Fax: 617-495-7105

5844 Five Oaks Pkwy St. Louis, MO 63128 Phone: 314-233-0421 Fax: 314-232-3393

Bob E. Schutz

Seabrook, MD 20706 USA Phone: 301-794-3478 Fax: 301-794-3524

R. Pearlman

J. Roessler

Ulrich Schreiber Fundamental Station

Ivan Prochazka

U.K. Rao BFEC 10210 Greenbelt

Helsinki, 00240 FINLAND Phone: 353-0-264994 Fax: 353-0-264995

SAO 60 Garden

James Pirozzoli Naval Research 4555 Overlook

Greenbelt, MD 20771 USA Phone: 301-286-4685

Michael

Cntr.

Ricklefs

Space Flight

McDonald Observatory University of Texas Austin, TX 78712-1083 USA Phone: 512-471-1342 Fax: 512-471-6016

Greenbelt, MD 20771 USA Phone: 301-286-8943

Giancarlo

Seabrook, MD 20706 USA Phone: 301-794-3494 Fax: 301-794-3524

Paul J. Seery BFEC 10210 Greenbelt

Ripamonti

p.z.a. Leonardo da Vinci Politecnico di Milano Milano, 20133 ITALY Phone: 39-2-23996103

xii

32

Rd/Suite

700

Cntr.

Michael BFEC 10210

Peter Sperber Fundamental Station

Selden Greenbelt

Rd/Suite

Selker

NASA/Goddard Code 726.1

Space Flight

Wettzeil

Koetzting, 8493 GERMANY Phone: 49-9941-603205 Fax: 49-9961-603222

700

Seabrook, MD 20706 USA Phone: 301-794-3499 Fax: 301-794-3524 Mark

Erik

Cntr.

Charles A. Steggerda BFEC 10210 Greenbelt Rd/Suite

Vermaat

Kootwijk Observatory P.O. Box 581 7300 An Apeldoom NETHERLANDS Phone: 31-5769-8211 Fax: 31-5769-1344

700

Huib Visser P.O. Box 155

Greenbelt, MD 20771 USA Phone: 301-286-1013

Seabrook, MD 20706 USA Phone: 301-794-3489 Fax: 301-794-3524

2600 AD Delft, NETHERLANDS Phone: 31-15-692160 Fax: 31-15-692111

Victor Shargorodsky Science Research Institute for Precision Device Engineering

Mark Torrence STX 4400 Forbes Boulevard

Thomas Varghese BFEC 10210 Greenbelt Rd/Suite

Lanham, MD 20706 USA Phone: 301-794-5213 Fax: 301-794-1010

Seabrook, MD 20706 USA Phone: 301-794-3498 Fax: 301-794-3524

J. Utzinger Sidlerstr. 5

Scott Wetzel BFEC 10210 Greenbelt Rd/Suite Seabrook, MD 20706 USA Phone: 301-794-3492 Fax: 301-794-3524

53, Aviamotornaya Street Moscow, 111024 RUSSIA Phone: 7-95-273-47-19 Fax: 7-95-273-19-37 Peter Shelus University of Texas at Austin Austin, TX 78712 USA Phone: 512-471-3339 Fax: 512-471-6016 Andrew

Berm, 3012 PEOPLE'S REP. OF CHINA Phone: 4131658591 Fax: 4131653869 M.R. van der Kraan P.O. Box 155 2600 AD Delft, NETHERLANDS Phone: 31-15-692269 Fax: 31-15-692111

T. Sinclair

Royal Greenwich Observatory Madingley Road Cambridge, CB30EZ ENGLAND Phone: 44-223-374741 Fax: 44-223-374700 David

Carolus Vanes P.O. Box 581

E. Smith

NASAIGoddard Space Flight Code 920 Greenbelt, MD 20771 USA Phone: 301-286-8671 Fax: 301-286-9200 Jay Smith NASA/Goddard Code 924

Space Flight

Greenbelt, MD 20771 USA Phone: 301-286-8525

Cntr.

Cntr.

Roger Wood Satellite Laser Ranging Herstmonceux Castle Hailsham, East Sussex ENGLAND Phone: 44-323-833888 Fax: 44-223-374700

700

700

Group BN271RP

7300 An Apeldoorn NETHERLANDS Phone: 31-5769-8211 Fax: 31-5769-1344

Yao Xingdia Changchun, PEOPLE'S REP. OF CHINA Phone: 0431-42859

Christian CERGA

Fu-Min Yang Shanghai Observatory 80NanDan Road

Veillet

Ave. Copernic Grasse, F06130 FRANCE Phone: 33-93-365869 Fax: 33-93-368963

,°o

Xlll

Shanghai, 200030 PEOPLE'S REP. OF CHINA Phone: 86-21-4386191 Fax: 86-21-4384618

WenweiYe WuhanSLRStation XiaoHongshan 430071 PEOPLE'S REP.OFCHINA Lu Yu-Lin Changchun, PEOPLE'S REP. OF CHINA Phone: 0431-42859 Xia Zhizhong Wuhan SLR Station Xiao Hongshan 430071 PEOPLE'S REP. OF CHINA Phone: 86-027-81342 Fax: 86-027-712989 Thomas Zagwodzki NASA/Goddard Space Flight Code 715 Greenbelt, USA

MD

Cntr.

20771

Phone:

301-286-5199

Ronald

Zane

University of Hawaii P.O. Box 209 Kula, HI 96790 USA Phone: 808-878-1215 Fax: 808-878-2862 Barbara STX

Zukowski

4400 Forbes Boulevard Lanham, MD 20706 USA Phone: 301-286-2779 Fax: 301-286-2929

xiv

WORKSHOP EIGHTH

AGENDA

INTERNATIONAL

WORKSHOP

ON LASER

Sunday

Evening,

6:00-10:00pm 8:00-10:00pm

May

RANGING

INSTRUMENTATION

17

Registration/Orientation Session

(Governor

Chairman

Meeting

Calvert

(Calvert

Inn)

Chamber,

Governor

Calvert

House)

Monday

Morning,

May

18

8:30-10:00am

Registration

(Joint

10:00-11:30am

Welcome/Orientation

Senate

- John

Welcome/IntroductionsWelcoming

Hearing

Room

Lobby)

Degnan

J. Degnan,

Address

- M. Baltuck,

Address

- J. Bosworth,

Program Head,

Chairman

Geodynamics

Branch,

NASA

Headquarters Welcoming

Manager,

NASA

Crustal

Project Orientation

- John Degnan

Last Minute Poster

Agenda

Papers

Submission

Schedule

Conference

Rooms/Splinter

Facilities Local GGAO

(A-V

Phone

for Proceedings Meetings

equipment,

Number

xerox,

for Workshop

Tour/Cruise

Restaurants/Local

- August

Attractions XV

etc.) Participants

1, 1992

Dynamics

Monday

Afternoon,

l:00-3:30pm

May

18

Scientific

Applications

Applications Texas

ofSLR,

SLR Tracking Richard

B. E. Schutz,

of Lageos

J. Fanes,

Center

of SLR to Gravity

Topography

Determination,

Tracking

for

Requirements

Center

and Etalon:

et al.,

Applications

Laser

Past Results

for Space

Field Modeling

Control,

Univ.

and Future

Trends,

Univ.

et al.,

P. Dunn

et al.,

Hughes

and Routine

LASSO Laser and

at the D-PAF,

Experiments, Ranging

Other

Laser

Application

Timely

Issues

- Andrew

Signatures

Greenwich

Earth

Orbit

et al., DGFI

OCA/CERGA

Transfer

Using

Programs,

Hiroo

TV Time

Transfer,

Geodetic

Kunimori

Satellite

et al.,

John McK.

CRL

Luck,

Sinclair

in SLR Data,

G. M. Appleby

et al.,

Royal

Observatory

Work at Graz Lustbuhel

on Satellite

SLR Data

Qualify

SLR Data

Screening

Greenwich

Signatures,

Control, for

P. Dunn

G. Kirschner,

et al.,

Normal

Points,

for

Preprocessing

Hughes

Observatory

STX

A. T. Sinclair,

Royal

Observatory Median

Measurements, SATCOP

for

STX

Observatory

Satellite

Adaptive

Veillet,

to Time

Space

Support

Performance

Ch. Reigber

Christian

Japanese

Ranging

Geodetic

Network

of Texas

NASA/GSFC

ERS-I

Ranging

Mission

Filtering

M. Paunonen, Planning

xvi

Finnish

Software

of

and Sea Surface

European

: Laser

Schutz

Research,

Research,

D. E. Smith

Vertical

for Space

- Bob

ESA's Intentions for Laser Tracking of Future Observation Satellites, Dr. Paci, ESA

Determination

4:00-6:00pm

& Measurements

of Time Series

Geodetic Package,

Institute S. Bucey,

BFEC

On'oral

Tuesday Morning, 8:30-10:30am

May

19

Laser

Technology

Nd: YLF Laser al.,

- Helena

Jelinkova

for Airborne/Spaceborne

Laser

Ranging,

J. L. Dallas

et

NASA/GSFC

Picosecond

Laser

Alternative

Wavelengths

Sci.

Transmitter,

and Physical

Laser

for

Jelinkova,

C. Veillet, Recent

Ranging,

Ranging,

Systems

K. Hamal,

J. Gaignebet,

of Ultrashort

of Nuc.

Ranging

Faculty

of Nuc.

Sci.

Laser

and Physical

to the Moon

OCA/CERGA Pulses for

Ranging,

H.

Engineering

and Meteosat3

at OCA LLR Station,

OCA/CERGA

Analyses

Geoscience

Laser

of Generation

Faculty

Multi-Pulse

Laser

QUANTA

Engineering

Two Color

New Methods

for

J. Ferrario,

and Laser

Laser

Oscillator

Ranging

System

Breadboard (GLRS),

Test Results

J. Gaignebet

for

the

et al.,

OCA/CERGA Simultaneous

11:00am-

12:00pm

Pulse

Train,

Epoch

and

Compression Yang

Event

Timing

Results of Accurate Lustbuhel Streak

Camera

Preliminary Intercomparison

Fu Min,

Results

Mode-locked

Shanghai

Observatory

Tests

Resolution,

from

Pulsewidth

and

- Ben Greene

Timing

Timing

of Passive

at Graz,

J. Gaignebet,

the Portable

with MOBLAS-7,

xvii

G. Kirchner,

Standard

M. Seldon

Observatory

OCA/CERGA Satellite et al.,

Laser

BFEC

Ranging

Tuesday

Afternoon,

1:30-3:30pm

May

19

Detector

Technology

Performance Ranging,

- Thomas

Optimization S. Cova

of Detector

et al.,

Tracking Capabilities Politecnico di Milano

Varghese

Politecnico

of SPADs

Electronics di Milano

for

Laser

for

Millimeter

(Invited

Ranging,

Talk)

F. Zappa

et at.,

How to Squeeze High Quantum Efficiency and High Temporal Resolution out ofa SPAD, A. Lacaita et al., Politecnico di Milano Solid

State Detector

Prochazka, Streak

Faculty

Camera

of Nuc.

Based

Atmospheric

Varghese

et al.,

BFEC

Temporal

Analysis

Sci.

of Picosecond

Faculty

of Nuc.

Sci.

Techniques/Targets

Experiences

and Results

P. Sperber

et al.,

(presented

and Physical System

and

System

Target

6:30pm

Buses

leave

7:00-10:00pm

Barbecue/tour

10:30pm

Observatory Buses return

High

Pulses

- Jean

of the MLTRS#1

I.

Accuracy

Measurements,

Reflected

T. K.

from

Satellites,

Engineering

Gaignebet USSR

Collocation

Campaign

Correction and Array Reflectivity, Astron. Obs. of the Ukrainian Acad.

of

by B. Schutz)

Characterization

NASA SLR Network V. Husson, BFEC New

for

Physical

Ranging,

Engineering

Delay

Laser

Test Results from LAGEOS-2 Optical Lasers, T. Varghese et al., BFEC Historical

Laser

IfAG

ETALON 1, 2 Center of Mass Nikolai Mironov et al., Main Science

Picosecond

Differential

Calibration

1991,

for

SLR Receive

Multiwavelength

K. Hamal, 4:00-6:00pm

Technology

Using

Concept

of the

of the NASA

Collocation

Based

for GGAO

Characterization

on Fizeau

Pulsed

SLR Network

and Special

Effect,

Using

Analysis

of the Techniques,

V. Shargorodsky

tour Goddard

(GGAO) to hotel

,°,

XVII1

Geophysical

and

Astronomical

Wednesday

Morning,

8:30-10:30am

May

20

Multiwavelength

Ranging/Streak

Two Color

Laser

Ranging:

Gaignebet,

OCA/CERGA

Optimum Wavelengths NASA/GSFC Two Color Telescope Two

Color

Laser

Ranging

Faculty

Satellite

of Nuc.

Millimeter Accuracy NASA/GSFC Low

Pulse

Faculty New

Spread

Ranging

Satellites

for

and

Possibilities

for

High

at Goddard's

Physical

1.2m

WLRS

Using

SPAD,

I. Prochazka

et

Engineering

Two

Retroreflector

J.

NASA/GSFC

U. Schreiber,

Physical

Sci.

Hamal

J. Degnan,

Upgrades

Sci. and

Laser

Ranging,

et al.,

Laser

of Nuc.

Satellites,

ll:00am-12:00pm

Ranging

at Wettzell,

- Karel

and New Developments,

Two Color

T. Zagwodzki

Two Wavelengths al.,

Potential

for

Satellite Facility,

Cameras

Color

Ranging,

Array,

J. Degnan,

I. Prochazka

et al.,

Engineering

Precision

2 Color

Ranging

to Geodesic

G. Lund

SLR

Data

Analysis/Model

State

of the Art SLR Data

SLR Modelling

Errors,

Errors Analysis

R. Eanes,

- Ronald

Kolenkiewicz

at GSFC,

S. Klosko,

Center

for Space

Hughes

Research,

STX

Univ.

of

Texas Geometric Analysis NASA/GSFC Improvement ESGT

of Satellite

of SLR Accuracy:

xix

Laser

Ranging

A Possible

Data,

New

Step,

J. Degnan

M. Kasser,

et al.,

Wednesday 1:30-4:00pm

Afternoon,

May

20

Operational

Software

On the Accuracy Compensation

of ERS-1 for

Due to Varying Institute Timebias

Developments

Pulse

Travel

Computer Technical

et al.,

Upgrade

Application Detong,

Laser

Roger

DGFI

Predictions

Finnish

Wood,

Operational

Geodetic

Satellite

Shanghai

Shelus

et al.,

Lunar

Laser

Environment,

Laser

LLR Activities Station

J.F.

Orroral

LLR

System University

- Christian X-Y

Ranging

Offset

Obs., Data

Ricklefs

in Wettzell,

Ranging

Mangin,

Antonin

Obs.,

Estimate

McDonald

Multi-Wavelength

Randall

Royal

L. Ricklefs

Novotny,

Control

et al.,

Czech

Computers,

of Texas

D. Edge

in SLR Data

et al.,

BFEC

Preprocessing,

T.

Observatory

Ranging

R.L.

G. Appleby,

Data,

Streamlining,

of the Robust

Laser

Points,

Ranging

McDonald

A Computer-Controlled

LLR,

Range

M. Paunonen,

at SLR Stations,

NASADOSE

Ricklefs

Lunar

Normal

of Laser Ranging Univ. of Texas

Networking Univ.

Upgrading

4:30-6:00pm

Laser

et al.,

Herstmonceux

Poisson Filtering McDonald Obs.,

HP

Times,

Kirchner R. Koenig

in Satellite

to Predictions,

The Formation of On-Site Greenwich Obs.

R.L.

Predictions,

the Distortion

Corrections

Ranger,

Orbit

- Georg

Veillet Guiding

University

Processing et al.,

Stage for

the MLRS,

in a Unix/X

McDonald

Obs.,

Windows University

Wettzell

Laser

et al.,

to the Moon

and METEOSAT

3 at OCA

Orroral

Obs.

J. McK.

XX

of Texas

U. Schreiber

OCA/CERGA

Activities,

P.J.

of Texas

Luck,

Geodetic

Ranging

Wednesday

Evening,

7:30-10:30pm Thursday

May

20

WEGENER/CSTG Morning,

8:30-10:30am

May

Meetings

21

Fixed

Station

Design al.,

Splinter

Upgrades/Developments

Principles

of Fully Automated

EOS Systems

Status

Luck,

Laser

(presented

Performance

Ranging

Upgraded

Geodetic

Upgrading of the Borowiec Research Center of Polish Progress

Hesheng,

Observatory,

Orroral

Laser

Improvements

B. Greene

G. Bianco

Ranging

in Graz,

et

et al.,

System,

J. McK.

G. Kirchner

Laser Station, S. Schillak Academy of Sciences

in the Work of the Yunnan

Yunnan

Development

Systems,

Obs.

Sub-CM Ranging and Other Laser Station Graz

New

Ranging

by T. Varghese)

of the

Orroral

Degnan

Inc.

of the Matera

ASI/CGS

- John

Laser

et al.,

Ranging

et al.,

Space

Station,

Feng

Observatory

of Shanghai

SLR Station,

Yang Fu Min,

Shanghai

Observatory WLRS

Status

Report,

NRL SLR Activities,

ll:00am-12:00pm

R. Dassing

and U.

C. Gilbreath,

NRL

Status

of Tokyo

Station,

Hiroo

Mobile

System

Upgrades/Developments

TLRS-3

System

Upgrades,

Results

of the MTLRS-1

R. Eichinger, Upgrade,

A Transputer Based Control Delft Uni_,. of Technology Presentation al.,

Kunimori,

of the Highly

ESGT

xxi

French

WLRS

CRL - Erik

Vermaat

BFEC

P. Sperber

@stem for

Mobile

Schreiber,

et al., IfAG

MTLRS,

E. Vermaat

SLR Station,

et al.,

F. P[erron

et

Thursday

Afternoon,

May

21

Airborne

l:30-4:00pm

and

Airborne

Laser/GPS

Krabill

et al.,

Airborne GLRS

Spaceborne

GLRS-R

- James

Abshire

of the Greenland

Ice Sheet,

W. B.

NASA/GSFC Ranging

Experiment,

B Extension

2 Color

Performance,

Studies,

Retroreflector

P.S.

Millar

K. Anderson, Target

et al.,

NASA/GSFC

GE/ASD

Design

and Predicted

G. Lund

Turbulence Churnside

Effects

Development al.,

Mapping

2 Color Phase

Systems

on the Geodynamic

of the Mars

Observer

Laser

Laser

Ranging

Altimeter,

System,

B.L.

J.H.

Johnson

et

NASA/GSFC

Bench Checkout C. Smith Single

Photon

Studies,

Equipment

Ranging

for

Spaceborne

Systems

I. Prochazka

et al.,

for

Faculty

Laser

Mars

Altimeter

Altimetry

of Nuc.

Systems,

and Atmospheric

Sci.

and Physical

Engineering Small

Spacecraft

Measurement Satellite

from

Operational

Altimeter

Low

to Satellite

Measurements, 4:00-6:00pm

Laser

Earth

Laser

Orbit,

Boarding

6:30-9:30pm

Conference

9:30pm

Return

Friday

Morning,

May

time

System

NASA/GSFC

Splinter

for Conference

Dinner

to Annapolis

Bufton,

et al.,

for

Meeting/Joint

LASSO Splinter Meeting/Arundel Poster Session/Governor Calvert 6:15pm

J.L.

Concepts

Ranging

J. Abshire Software

Instrument

for

NASA/GSFC

Lunar

Gravity

Hearing

Conference

10:30am-12:00pm

Business

Dinner

Cruise

(Annapolis

Cruise Harbor

12:00pm

Adjourn

Summary/Resolutions Meeting/Next

Workshop

xxii

Room

Room/Maryland Inn Inn/Calvert Chamber

22

8:30-10:00am

Topography

- Michael - Carroll

Pearlman Alley

Harbor)

J.

POSTER PRESENTATIONS

Satellite

Laser

Station

Helwan,

NRIAG,

Helwan,

Egypt,

and

Czech

Tech.

Univ.,

Prague,

Czechoslovakia 1.2 Meter

Telescope

Zagwodzki

et al.,

Ranging Plate

Data

Facility,

Space

Flight

Center,

Greenbelt,

Md.,

T. W.

NASA/GSFC

Quality

Photomultiplier

The Optical

Goddard

Attenuation

Improvement Tube,

from

T. Varghese

Mechanism,

High

Speed

et al.,

Detection

BFEC

R. Eichinger,

,,°

XXIII

BFEC

Using

6m

Core

Microchannel

Scientific

Applications and

Measurements

Requirements

N94-15553 LASER Peter

TRACKING Dunn

and

FOR Mark

VERTICAL

Torrence,

CONTROL Hughes

STX Corp.,Lanham,MD

Erricos Pavlis, Univ. of Maryland,College Park,MD Ron Kolenklewiez and David Smith, GSFC LTP,Greenbelt,MD ABSTRACT The since

Global

the

first

dimensional

Laser

Tracking

MERIT

Network

campaign

positlons

in late

stations very

when

to that small

predicted

effect.

measuring

the

we can

now

ranglng resolve

at monthly

data

accuracy

centimeter-level

intervals.

separately from the with a formal standard

of high

three

In this

analysis,

by recent

models

attention

component,

of post-glacial

must and

be applied

the

survey

the

horizontal components, error of 2 ram. using

The rate of change in the vertical can be resolved level of several geophysical effects. In comparing

Particular

vertical

and

LAGEOS

observatories considered stations

years of continuous observations. mm./year, which is at the expected in this

provided

1983

of participating

station height estimates have been can be determined by the strongest

of the

has

rebound, to data

observatlons

to a few the behavior

we find

and

survey

are

and eight

no

correlation

quality

critical

control

components

of

the geodynamle results. Seasonal patterns are observed in the heights of most stations, and possibility of secular motion at the level of several millimeters per year cannot be excluded. such motion must be considered in the interpretation of horizontal inter-slte measurements, can

help

to identify

seasonally

mechanisms

which

can

cause

variations

which

occur

linearly

with

the Any and

time,

or abruptly.

INTRODUCTION LAGEOS laser horizontal motion processes

and

ranging measurements at the observing stations

regional

deformation

movements are as large Island which lie astride ability

to measure

usually

been

independent

from annual as the network Accurate

processes allows loading

vertical

and the

to calibrate

the

stations.

of SLR

stations

altimeter

per work

year The

(Frey

and

Bosworth,1988).

fast moving stations boundary. The SLR to a few mm/year,

they time

significantly to our knowledge of helped to improve models of tectonic

boundaries

motions

because

variations.

directly grain

provide

of the

but

geodesic

horizontal

horizontal

control

can

of pre-

or

of post-glaclal The

scale

to establish instruments

assist

the

horizontal

post-selsmlc rebound

of an Earth-centered a global by determining

mission.

I-I

the

vertical the

radial

height

have

has

systems

component

et

tectonic

determination

can

Easter their

are

of atmospheric

system The

tectonic

averages (Smith have improved. in monitoring

Accurate

datum.

and

measurements

investigation

reference

lengths

rates

positioning

events. and

The

such as Huahine and data have demonstrated

values (Chrlstodoulldls et al., 1985) to quarterly has grown and observation and force models

detection

measurement

at the

network altimeter

in this

of vertical

progressed al., 1990)

at plate

as 17 cm/year between the Paciflc/Nacza plate

centimeter

used

have added and have

be defined

can

also

of the

orbit

also pressure in a

be employed of the

Degnan(1985) hasdescribedthe varioustechnicalmethodsof accuraterange which

include

as accurate bench-mark. points

we employ

in the

distance errors

that

range

the

and

laser

detection

scheme.

The

and

would

bias

the

magnitude

of the

the

final

center and be preserved

position and

returns

corrections

Errors station

in station

position,

is normally

distributed.

and

usually

normal

points,

yielding normal points with this system characteristic

signature

of the

would

satellite

We have between

time-keeping

although

round-trip

time

can

modern

degrade

systems

measurement

out in the horizontal directly affect their calibrated strict

quality

Figure

control

1, and

component.

their The

network,

standards.

The

positions

listed

observations

from

determination of 3-dimensional improved resolution of the rate DATA ANALYSIS

In our satellite squares

analysis,

model

subsequently positions, motion

each

over

of mass al., 1992)

about and

the

The

time

more

have

locations

array

range with on

the

photon, leading edge detection methods may

of the transfer

difficult

subjected

of these 1, with

strong

measurement

of equations (Putney, 1990) spans

horizontal

stations

of

are

They

will

observations are

shown

allow

of a year

tend

sky coverage, our analysis

emphasis

now

time

at the level of positioning model. Systematic errors

their

stations

component

for epoch

to control.

particular

in

to cancel

but will to the best

to particularly on

on

the

the

world

map

of

vertical

us to reduce to a month,

with

expanded

an

the and

interval thus

for

provide

effects

of thermal

effect

(Rublncam,

(Ries

value

of a numerically

integrated

range Information in a least computed with an accurate

The

on the with

frame

from

satellite an

I-2

initial

which

resulting

Includes

of 398600.4415

et ai.,1992).

to Include

perturbations

1990)

solution

a month.

in a reference

tides

drag

the

linear

system

is

three-dimenslonal coordinates of the tracking station parameters estimated at various tlme Intervals. The

adopted

constant ocean

constrains

which satisfies all of the for orbits independently

of approximately

is computed Earth

supplemented by third body Mercury through Neptune.

Yarkovsky

by

a longer will vary

retro-reflector

of stations with adequate treatment we have restricted

and

these

resolution

GPS

positions from a quarter of any station movement.

SLR

time

gravitational

with

be causecl

adopted a value of 251 ram. the satellite's center-of-mass

insignificant error satellite perturbation

are

in Table

solved to yield monthly together with other geodetic

of the satellite

relativity

be

Any

METHOD

trajectory. A system sense is developed

perturbation

is an in the

for range

in the

the

using

position measurements height estimates. In this

observatories

must

by a few millimeters.

synchronized to the microsecond, which accuracy currently dominated by errors the

for the

system

return, although

also depend on the instrument. for the correction for the offset

differing

a ground in the normal

estimates

detection

and its reflecting surface, which would be expected from the multiple detection MOBLAS systems. Lower power transmitters with alternative require

m_surement refraction, as well

atmospheric

transmitter

of satellite

which would delay the detection of the than that from a Gausslan distribution,

range measurement will (Fitzmaurice et al., 1977)

delays

will affect

instrument's

distribution

pattern

path

between a system's electro-optical in the original observations will

analysis,

of each

to ensure

skewness

for electronic

in our

Characteristics

monitored errors value

calibration

surveys of the Any systematic

which

stations.

the

careful

The

significant the

were

sun

GEM-T3 the

perturbations moon,

represented

satellite

spin

effect

together

by a model axis

of general

for GM,

geopotentlal

LAGEOS and

the

km3/sec2

orientation

the

model

product (Lerch

et

was with

of the

the

planets

Earth

of 22 degrees,

decreasing by 50% every 6 years. To satisfy remaining track acceleration was adjusted every 15 days, as well track

component

acting

once

parameter been found

is related to the to accommodate

adequately

described

determined

by the

determined

parameter

repeating (Anselmo

(cf. Eanes with

the

same

measures

position

dynamics

The

radiation

the

value.

The

values

unusual

than

this

could

vertical

components

loading

at appropriate

this semi-diurnal of stations with

at favored

of the

times

locations

function variation

in it's

in 1991,

and

of the

day

(or night).

International

Earth

Rotation

Service

was

workers on both

crossing

associated

with

indicates

a have

The

once

per

revolution

level

Standards:

each

month.

McCarthy,

1991),

averaged over the monthly have an effect on stations and

for one

shown

have suggested that would lead to chaotic spin

by as much as ten 3, the root mean square fit of precision it is possible to

(IERS

rotation

of

equator

observations

centimeter

were adjusted included, and fixed

amplitude

results.

at the

applied

Earth

in which they polar motion

is a well-

regularly

signature

from those

recent

Iess(1991) gradient

these

stations

angle

particularly

effect would be very small when adequate sky coverage, but would

were taken from a global solution system with the effects of dynamic

the

along-track

of orbital

plane,

explain

was

2. This

and

models.

Bertotti and and gravity

selected

acceleration

in Figure

of the

has

of the once-per revolution acceleration the direct effect, and have formal errors

again

help

alongalong-

adjustment

along-track

is shown

behavior

unusual and

revolution

1 plcometer/scc2,

by these

continues in 1992. due to eddy currents and

per

have been modelled by several et al., 1991) using theories based

equatorial

The

in 1989

effects, a secular amplitude of an

by Yoder et al.(1983) and which have not yet been

of secular period

cosine

in the

pressure.

1992,

once

affect monthly orbital fits to the ranging observations but when modelled according to the values of Figure data remains below five centimeters, and with this

Ocean although estimates

or

The

components determined

behavior

This

of about.

predicted

perturbations

satellite's

in 1991

perturbations centimeters, each month's

by the

orbit.

experimental

uncertainty

well

as a typical

that the irregular behavior torques on the spacecraR

track

the

heating.

very

the orthogonal more weakly

size and

in the

solar

is not

unmodelled

resolve

over

a formal

direct

in 1990

Figure 3 shows estimates, which are

change

of the

orbit and

excitation vector described in the behavior of LAGEOS

et al., 1991).

full network

and

commencing

solar

revolution

patterns in the early part of the signature et al., 1983; Afonso et al., 1989; Scharoo

Earth-reflected

about

per

eccentricity variations

unmodelled as the phase

orientation

parameters

daily in the J2000 in which the UT1 day

of each

position which

month

(EOP)

reference time published to establish

a

longitude frame. In the global solution the station position for each site was estimated, motion was modelled according to Smith et al.(1990), resulting in a consistent reference

but its frame

throughout the eight year experimental period. In both the monthly analysis which yielded the height values presented

the system

was set by fixing the horizontal position Maul (latitude). The results for monthly both of the flducial stations at Greenbelt and

if there

were

Wahr(1981) according LASER

DATA

and

from

the

components of Greenbelt (latitude and longitudes) and values of station height are reported only if coverage for and Maul reached a minimum of nine LAGEOS passes, each

effect

solution for EOP and the stations' reference

individual

of solid

station.

Earth

tides

A nutatlon at the

series

stations

was

according

to

also

computed

analysis

contains

1981)

QUALITY

of the

calibrated

data

adopted

to Wahr(

Each well

was

adequate

global here,

CONTROL

observatories system

that

whose has

been

vertical

motion

in operation

1-3

since

was late

monitored 1983.

In this During

the

lifetime

of each

a

station, continuous improvements are made to the system through up-grades in hardware and software. Any disturbance at an instrument is monitored with accurate resurveys of the system's eccentricity

(optical

in the surveyed eccentricity Center

center

offsets

are

respect

of the

in Table

System

tower

MOBLAS

2. They

(CDDIS)

to an associated

calibration

for the various

listed

Information

with

distance

have

ground

used

instruments

been

in December

from

and

their

as well

delay

fielded

retrieved 1991

marker)

for system

as of any change

correction.

The

by the Goddard

the Crustal

Space

Dynamics

correctness

will

Flight

Data

directly

affect

the

estimated heights given in Table 1, as well as any measure of vertical motion. The remaining observatories in the network were assumed stationary during the eight year period and their positions refer to the optical axis of each telescope, which is the estimated parameter in our reduction. the

Any

marker

improved

information

positions,

and

on eccentricity

it is not necessary

to repeat

other hand, techniques for direct estimation values at the outset of the analysis to connect Information

concerning

calibration

by subsequent

improvements

might

analysis,

exert

bias

parameters,

and

updates using linear shiRs the initial time-consuming

to efficiently

reduction

process.

On the

require accurate eccentricity occupation at a site.

system

is accessible

through

the

to compensate

position.

Subtle

of our

for any effects

engineering

problems

using the original by pass-by-pass

analysis

facilitates

based on the partial derivatives computation of normal equations.

retro-actlve

time-of-flight observations, or longer term range or

the

of range Several

that

in the detection

incorporation

of historical

or clock bias computed in corrections to the released

data were required. In particular, range corrections cm to each measurement up to March 1986 to allow

to Arequlpa observations for the improved survey

were applied: of the calibration

tower

noted

as well

until

which

time

Range

errors

in the CDDIS improved of this

description

system

delay

magnitude

of this

would

of a few ram/year,

is closely

monitored.

The

an abrupt

change

when

earlier,

frame

of Figure

year

data

earlier

in station

span,

observations

ANALYSIS

The month

independent

variation

the vertical

estimates

In our

and

force, laser

of several

measurement

analysis from

of height are given

an average

the measurements

ranges

centimeters and

is modelled

in quarterly

solutions

bias

held

fixed

shown

at the

was

motion locations

position

clearly

shown

is

seen

in the

estimated

4 indicate

at

1988).

at other

in station

this

at a value

in Figure

Earth

In amplitude

are models.

to Marini

1-4

and

stations 5a,b

with

and

semi-major

are qualified

orientation

according

three

in Figures Earth

standard deviation based on the final fit of the range the ranges themselves are formally accurate to better signatures

level:

1988

lower

over

error

the

13

in the

magnitude.

values

of the distance scale

of range

anomalies effects

July

(Husson,

of vertical

of similar

maintained was

offset

any estimates

of engineering

used

3 cm

this

4

MEASUREMENTS

monthly

seen

was

of the station

of the correct

OF VERTICAL

in millimeters

data

the height

the monthly

affect

indication

as another

Indicated

the possibility

to a subsequently

Arequipa

4. When

and

compelling

height

uncorrected

procedures

significantly

at the rate

most

station,

calibration

occurring

data

up-date

in the processing of the raw range measurements As corrections to the calibration procedures are

it is necessary

the design

be used

full data

of each

system must be remedied in a pre-processlng stage but many of the data corrections can be represented timing

the

characteristics

on station

can

of station velocity will the positions of each

CDDIS, although it is has already been used and is thus embedded in the normal points. uncovered

surveys

c. The

axis

by error

the

lowest

least

significant

of 6378136.3

estimates

The Murray

due

effect

of twice

to uncompensated

of atmospheric

(1973)

who

assumed

figures

m. appear

observations to each orbital than a centimeter, systematic

observed

month-to

their

formal

arc. Although residual errors

refraction

In on

a spherically

the

on

stratified

atmosphere

based

on surface

range corrections due the refractivity dependence of the range correction elevation variations

angle, which is the in station barometer

pressure

measurements.

formula, the formula should

lower limit accuracy

zenith only

Herring

(1988)

not

be separable

The

possibility

for stations Huahine collocation

which

from

the wet

for most of the systems. However, or in the effects of lateral gradients

of errors

in the

have

undergone

component

in nearby

adopted

of system

must

The

collocation

transportable

calibration

working TLRS-2

location. eccentricity

likely

Considerable and

to their

quoted

individual

most

values

listed

in Table

height

statistic

component

1 are has

only

reduced

been

of motion

used

(see,

a mean we see

as upon

the

of the

precision and

Arequipa

Post-glacial

observations,

of the

and reliability

of the

Earth

from

factor

in earlier

lower

for some

significant

compared

on the

fit of the

about

the

fit is substituted. measuring

of height

the

horizontal scatter

'quality' careful

scatter

mean The

the

Considering

as a measure of the system stability and values

melting

variation

values

work

et al., 1990).

of these

the

a linear

and at their

sites but, because technique, this

based

height

test

MOBLAS-1

not

error

when

the

and

for up-grades

height

3 are

or two

3 of Smith

to the

accurate

standard

(or uniformly moving) value that it depends as much on

testify

rebound

control

Table

Quincy

a collocation

in the

in Table

of scatter

by a millimeter

less

be noted

formal

particularly

Greenbelt,

to Greenbelt undergo

early,

shown the

measures

as a quality

height about performance,

can

rates twice

The

for example

station's station's

Yarragadee

are

line.

of the

motion

height

which

to a straight

influence

uniform

estimated

uncertainties,

usually

could

sites coincided with machine as well as its

returned

do not

as

shows more variation than the other by employing a precise reposltionlng

to the

from

of the

periodically but

position minimized

to be due

deviation

are

MOBLAS-7,

The Huahine errors are

behavior is more measurements.

stations,

systems

against

be considered,

such

American ranging

any long term in the atmosphere

instruments.

also

occupation,

(see Table 2). The system changes at the North tests which cross-calibrated each instrument's

eccentricity.

that

The SLR systems refraction which

microwave

eccentricities

changes

shown

range correction and the elevation be a few millimeters at 20 degree

(see Abshire and Gardner, 1985) will directly affect the vertical estimates. thus be used to monitor aberrations in the dry component of atmospheric would

has

of a

of the calibration

at Greenbelt,

instruments. of continental

ice sheets

starting

roughly

18,000 years ago produces changes in the gravity field as it affects the long-term evolution of the LAGEOS orbit and have been reported by Yoder et al. (1983) and Rubincam (1984). Wagner and McAdoo due

(1986)

present

to post-glacial

complete

enough

predicted

by the

compared

the

for convenience and observed

model

model

nor

America due could

the

observations

lack

explain

modelling observations,

SLR

sites.

taken

from

height

detection

the

by tide model

rates

model

maps The

for the

rate

of Wu and

values

Figure

of vertical

from

the

gauge

results.

data

in their James

of the

at Greenbelt (see

Trupin

treatment and

1-5

uplift and

laser

1991) been

(1990)

lithosphere horizontal

: the

absence

noted have

data

sinking

at each from

other

position model

is

observatory and

the

SLR

are

stations,

between the rate expected hand,

of eastern of higher

by Wagner shown

this

McAdoo(1986)

On the

confirm

of radial and

can be seen the 4 mm/year

systems.

can has

Morgan

of the properties of the have also indicated that

SLR

of change

Peltier(1983),

5 of Wagner

estimated

capability taken

of a lithosphere

assumptions and they

Ice-2

by tectonic plate. Very little correlation values of up-lift, even in Europe, where

the

as required

to the

viscosity

the

been

the

is within SLR

on

all the

have

3 with

uniform

based

to include model

in Table

arranged modelled

a simple

rebound

and

in more

neither

from the

North degree

terms

McAdoo detail

and

how

can cause disagreement with sea level motions due to post-glacial rebound in

North America

and

Fennoscandia

can

amount

to 4 mm/year

from

plausible

models.

This

movement is predicted 10 ram/year, and both

in the Hudson Bay region where vertical movement can amount to over components are clearly within the resolution capability of a modern SLR

system

this

region

in an extended

that

further

investigation

occupying

It is possible instrument

error

apparent assume

which

would

rate, of 4 mm/year a stationary

alias

of the

into

observed

vertical

campaign. SLR

the vertical

at Arequipa

component

and

observations

will

component Is large

expect

uncover

of station

enough

accurate

that

baseline

stations. Only explicit separation from the vertical component will allow the definition of accurate horizontal motion.

a source

tx_sition.

no SLR

analysis

measurements

by considering

of

However,

the

should to distant

geodesic

lengths

CONCLUSIONS

The eight

stability

year

analysis

time does

signatures nature,

of the span

height can positioning campaigns exercised

that

in the height

of position

vertical

variations

do not produce

motion

suggested results

significant

represent

term

in close to retain

proximity

to the SLR

the full scaling

as they benefit from tracked. On the other

reception

center

from

strong hand,

orbital accurate

supplemented resolution

These

a ground

marker

to be in high

Earth

orbit

by concentrated of any subtle

vertical

in the next

tracking motion

1-6

variations

also

this

of a geophysical

estimates

of station

as to establish scale for on a global scale, such as GPS quality

control

which

is not so stringent

in both

space

As observations orbiting

be improved.

ETALON

from

techniques

be

are of each and

Network can satellites

LAGEOS

spacecraft,

must

in the analysis

when multiple satellites position measurements

is critical

few years.

in an

Periodic

which the Global Laser Tracking number of retro-reflector-carrylng

of the currently should

and

rebound.

accurate

ranges

geometry relative

observatories

by 2 or 3 ram/year

The data

of the laser

SLR

of post-glacial seasonal

trends.

Observatories.

accuracy

must be carefully monitored. The capability with control vertical scale will grow with the increased expected

strongest

Is bounded

by models

may long

at the

help in the calibration of satellite altimeters as well tech'niques which degrade as a function of distance

of GPS networks simultaneously instrument's

component

suggests

not confirm

apparent but

radial

2 are time

REFERENCES

Abshire,J.B. IEEE Trans.

Afonso,

and C.S.Gardner,"Atmospherlc on Geose. and Rem. Sense.

Refractivity GE-23, 1985

G,F.Barlier,M.Carplno,P.Farlnella,F.Mignard,M.Milanl

LAGEOS

Seasons

and

Eclipses",

Pressure

for a Spacecraft

Bertotti,B.,and

L.less,"The

Chrlstodoulidis,D.C.,

D.E.Smith,

P.J.Dunn,"Observing J.Geophys.Res.,

90(B

Degnan,J.J.,"Satellite Geose. and Rem.

Eanes,R.J.

and

Ranging

M.M.

of LAGEOS",

Laser Ranging: GE-23, 1985

Fltzmaurice,M.W.,P.O.Minott,J.B.Abshire Geodynamic Satellite (LAGEOS)",

NASA

Frey H.V. and J.M.Bosworth,"Measurlng Dynamics Project", Earthqu. and Volc.

Six Month

James,T.S.,and 1990

NASA

Goddard

Space

Coefficient

and

Syrup.

Flight

and of the

IERS

Adv.

in press,

i-7

Testing

Motions:

Report,

1992

Res.,

IEEE

Trans.

Field

from

of the

Laser

NASA's

October Rebound",

Tracking, Altimeter 104555, 1992

Note

3, Central

I0(3), in the

Ranging",

on

Satellite

1991

M.H.Torrence,"Preclslon Space

Laser

Gravitational

to Post-Glacial

Tech.

to Radiation

and

Prospects",

Crustal

due

M.M.Watkins,"Progress

Earth",

of

1991

Satellite

No.3,Vlenna,

Report",BFEC

and Center",

from

H.E.Rowe,"Prelaunch Paper 1062,1977

Contemporary 20(3), 1988

due

M.H.Torrence

Future

Models of the Earth from Satellite GEM-T3 and GEM-T3S", NASA TM

Ries,J.C.,R.J.Eanes,C.K.Shum Gravitational

and Tech.

D.D.(ed.), IERS Standards(1989), de Paris, 1989

Effects

96,2431-2440,

of Earth's

IUGG

Motions

Putney,B.,R.Kolenkiewlcz,D.Smith,P.Dunn the

Ass.

Orbital

Perturbations

Deformations

Status

Processing

W.J.Morgan,'Horizontal

Lerch,F.J," Geopotential Gravity Observations: McCarthy Observatoire

Data

and

Variability

XX Gen.

Ranging",

29,27,1983

S.M.Klosko,

Current

Watkins,"Temporal

Mech.

J.Geophys.Res.,

R.Kolenkiewicz,

Observations",

Husson,V.S.,"First

Celest.

Tectonic Plate Motions 1 I),9249-9263,1985

Sense.

A.M.Noblli,'Orbital

Shape",

Rotation

A.M..Nobill,"

Laser

7(5),501-514,1989

and

of Complex

in Satellite

and

Ann.Geophys.A

Anselmo,L.,B.Bertotti,P.Farinella,A.Milani

Laser

Corrections

197-203,

Crustal

1988 Geo.

and

R.Lett.

Surface

Bureau

of IERS,

Orbit

Determination

1990

Determination

17(7)

of the

at

Rubincam,D.P.,"Postglacial Lower

Mantle",

Rublncam

Rebound

J.Geophys.Res.,

D.P.,"Drag

on

Observed

89(B2),

the

LAGEOS

Satellite",

and

J.Geophys.Res.,

95(B

R.Noomen,"On

96,729-740,

Trupin A.S.," Gravitational

Wagner,C.A. Intersatellite

13),22013-22041,

The Effect Potential",

Forced

Geophys.J.R.Astron.Soc.64, Wu,

P.,and

Constraint

Deep

303,

Long 1991

Variation s. 91(B8),

95,4881-4886,1990 the

Along-track

Satellite

Period

Acceleration

of the

and Ranging to LAGEOS",

Laser

Tides

In the Earth's July, 1986

of an Ellipsoidal

Glacial

Mantle

Isostatic

Viscosity",

on

the

Gravity

Rotating,

Adjustment Geophys.

Yoder,C.F.,J.G.Williams,J.O.Dickey,B.E.Schutz,R.J.Eanes Earth's Gravitational J2 Coefficient from Nature

of the

Earth's

Elastic

Field

and

Rotation

Detectable

Oceanless

and

With

GRM

Earth",

1981

Peltier,W.P.," on

Viscosity

1991

of Global Change and U. Col. Ph.D. Thesis,

Nutatlons

Effective

1990

and McAdoo,D.C.,"Time Traeking_,J.Geophys.Re

Wahr,J.M.,"The

the

Geophys.Res.,

Smith,D.E.,R.Kolenkiewicz,P.J.Dunn,J.W.Robbins,M.H.Torrence,S.M.Klosko R.G.Williamson,"Tectonic Motion and Deformation from J.Geophys.Res.,

and

1984

Satellite",J.

Scharoo,R.,K.F.Wakker,B.A.C.Ambroslus LAGEOS

by LAGEOS

1077-1087,

LAGEOS

1983

I-8

and

the

J. R. Astron.

and

Free

Air Gravity

Soc.74,

Anomaly

as a

1983

and B.D.Tapley,"Secular Non-tidal Acceleration

Variation of of Earth Rotation",

TABLE

1 :

STATION

POSITIONS

LATITUDE DEG

MNSEC

LONGITUDE

HEIGHT

ST.ERR.

DEG

METERS

MILLIMETERS

MNSEC

ST.DEV

NO. MONTHS

283

i0 20

19.931

2

16

69

30

239

3 19

1107.119

2

18

67

32

53 30

243

34

39

1839.746

2

20

73

7090

-29

2 47

115

20 48

242.080

2

16

69

HUAHINE

7123

-16

44

l

208

57 32

46.110

5

23

22

AREQUIPA

7907

-16

27

57

288

30 25

2492.945

2

17

52

MATERA

7939

40 38

56

17

536.551

2

19

60

WETZEL

7834

49

12 52 41

661.842

4

24

45

GRAZ

7839

4742

15 29 36

540.125

3

20

55

RGO

7840

50 52

76.114

3

21

69

SIMOSAT0

7838

33 34 40

100.175

4

25

51

GREENBELT

7105

39

114

QUINCY

7109

39 58

MON.PEAK

7110

YARAGADEE

8 42

16 42

20

3 135

I0 56

13

I-9

TABLE

2 : ECCENTRICITY

OFFSETS START

GREENBELT

7105

QUINCY

MON.PEAK

MOBLAS-7

GROUND

1

1

84

3 22

16

-26

3169

84

3 22

85

7 29

17

-32

3169

85

7 29

89

10 12

17

-31

3168

89

10 12

90

7 25

35

-40

3162

90

7 25

91

12 31

-14

-33

3153

4

6

90

7 23

-7

-5

2613

TLRS-4

90

7109

MOBLAS-8

84

I

l

86

9 18

-29

11

3124

86

9 26

91

3 17

-27

12

3138

TLRS-4

91

3 19

91

8 19

-5

0

2651

MOBLAS-8

91

l l 18

91

12

II

-19

5

3184

91

12

12

91

12 31

-35

-3

3184

7090

MOBLAS-4

841

TO 7918

7121

TO 7123

MOBLAS-5

7121

MOBLAS-

7123

TLRS-2

MARKER

7105

UP(mm)

84

1

88

4 30

-33

-15

3210

4 30

91

12 31

-33

-16

3213

84

1

87

8 13

87

813

91

12 31

3 3

11 10

3185 3177

88

HUAHINE

N(mm)E(mm)

7918

71 I0

YARAGADEE

STOP

I

1

84

l

1

86

3 13

8

I

3662

87

7 14

87

10

8

0

0

1453

88

3 16

88

9

I

0

0

1437

89

424

89

9

3

0

0

1482

90

3 15

90

8 20

-I

3

1459

91

4

91

9

-2

4

1482

5

DISTANCES X(mm) -14419 1458

Y(mm) 5137

Z(mm) 9457

807

501

i-I0

4

TABLE TECTONIC

3 : COMPARISON

PLATE

N. AMERICAN

STATION

WITH

POST-GLACIAL

REBOUND OBSERVED

MODEL

GREENBELT

3

1.7

QUINCY

3

1.5

MON.PEAK

I

2.6

2

HUAHINE

I

3.2

4

AFRICAN

MATERA

I

2.3

EURASIAN

WETZEL

4

- 1.5

3

GRAZ RGO

4 4

.9 -.2

2

AUSTRO-INDIAN

YARAGADEE

I

1.4

2

S. AMERICAN

AREQUIPA

-2

4. l

2

UNKNOWN

SIMOSATO

-3

2.2

4

PACIFIC

MODEL

I-II

+b2 mm/year 2

2

2

0 0

>.

_o

0

"I" Q. .Q r-

O"


-

i |

8 8 $|1

>-

'OO O"

e e

I el I

cO OR

e,,,,-,

ee

I, I 0

i

e e e e I

e

("4 CO O"

e

|° 01

0

i e I

i' ,i'

C) 'CO O-

t

|

I

i

Ii

:

e

i,', : O

!

0 U")

i

I

I

I

I

I

e

_

I

'i

e

lllllllll

I (:D

CO r-OO"O

(LUO) S_B AInU OIAI

1-15

0 ..............................................

+ ............

_.._

.............................................

I C I_

0 0

I=.

m MH

0

i

0 o.

L

J l In /'_zl

0 0

0 0

0 0

0 0

0 0

(s,e),Om!ll!m) ),q6!eH 1-16

0 0

C_ O_

_IHH

HHi '

i k.

O_ _lm'

CO 0

.C

......................................................... _ ""............................................. CO

4)

0

>, h4_

............................................................... ,.._._ ............................................. r,,, !

O

a_

2 .j I

:

O U)

CO

0

0

0

0

0

(_1

T-

O

O_

CO

sJe),eu_!ll!U_ 1-17

(M O_

Ii

01

m_ !

! I

i

I

J 0 O_

i

i I

t

i I

E

O_ ............................................. t.......... _ ......... +...................... ._ ...................... CO

_m

0

{J .c: !

Im ....... "I" ....................................

i"_

O0 ¢0

..................................................

,

_

t

m G) >,

0

H._

J= _a

tIN

GO

I

0

L. I

(D GO

!

5

'

I

In GO

[

0 0

0 0

0 0

0 0

0 0

sJa),eUJ!ll!m 1-18

0 0

O0

N94-15554 L

Laser Ranging Network Performance and Routine Orbit Determination at D-PAF

F.-H.

Massmann,

Ch. Reigber,

J.C. Raimondo, Deutsches

Geod_itisches

German

C.

H. Li, R. Ktinig,

Rajasenan,

M. Vei

Forschungsinstitut

Processing

(DGFI),

and Archiving Facility MiJnchner Str. 20

D-8031

Abt.

I, and

(D-PAF)

Oberpfaffenhofen Germany

Summary ERS-1 the

is now about

very

8 months

beginning

of the

(D-PAF) is coordinating determination tasks. This

paper

presents

mission.

and

details

in orbit

and has been

The

German

supporting

about

the

tracked

processing

the network

global

by the global and

archiving

and performing

network

status,

laser

network

facility

the different

the tracking data and orbit processing system at D-PAF. The precise orbits are shown and some problem areas are identified.

quality

for ERS-1 routine

the communication

to D-PAF

of the

from

preliminary

orbit

and and

1. Background On July

17, 1991,

ly launched microwave to make

the first

from

Kourou,

instruments full

use

measurements

the satellite

ERS-1

laser

in a sun-synchronous

the satellite

The

Sensing

satellite

ERS-1

orbit

Satellite

is equipped

of the Earth's

the

ranging

and an inclination of 98.5 degrees Kepler period is about 100.5 min. To keep

Remote

environment has

(ERS-1) with (see

to be known

was

successful-

a number figure very

of active 1). In order

precisely.

is carrying a laser retro-reflector and, as an experiment, the Prare Range Rate Equipment). After the failure of the Prare system due

diation damage bit determination.

is flying

(ESA)

Guayana.

for the monitoring

of the

this purpose ERS-1 (Precise Range and

European French

ground

maintenance manoeuvres weeks depending mainly

tracks

(SLR)

circular with

within

measurements

orbit

different

(quasi-polar) ground

+-1 km deadband

have to be executed, on the solar activity.

1-19

are the basis

which

take

repeat

at a mean cycles

(first 6 months place

usually

For system to a ra-

for the precise

altitude (see

of 785 km

table

+- 2 km) every

or-

two

1). The

so called to four

Figure

1: ERS- 1 Satellite

Wind

1DHT Antenna

Table

1: ERS- 1 Orbit

Characteristics Repeat

Revolutions per cycle Revs per day Mean Semi-Major Axis [km] Mean Inclination [Degrees] Mean Eccentricity Mean Argument of Perigee Mean local solar time of descending node Longitudinal phase (ascending node) [degrees] (1) Venice (2) ICE Orbit Duration

3 days 43 revs 14 + 1/3 7153.138 98.516 1.165 xl0 -3

Cycle

35 days 1.501 revs 14 + 11/35 7159.4'965 98.5429 1.165 xl0 -3

I 176 days 2521 revs 14 + 57/176 7156.30 98.5114 1.165 xl0 -3

90.0 deg

90.0 deg

90.0 deg

10h 30min

10h 30rain

10h 30min

24.36 East O) 128.2 West _)

20.9605

not decided

26/07/91-12/12/91 23/12/91-30/03/92 16/12/93-01/04/94

1-20

¢1_' (2) (2)

East

2/4/92-15/12/93

8/4/94-...

2. ERS-1

Laser

The ERS-1 by many periods.

Tracking

satellite different

Network

is tracked institutes.

by a network In table

of globally

2 the stations

distributed

are listed

stations

along

which

were

with their ERS-1

funded tracking

In figures 2 to 4 the geographical distribution of the stations is plotted for the first three orbit phases (see table 1). As can be seen, most of the tracking stations are located in Europe and North

America.

The

southern

hemisphere

of them is located orbital passes.

in the south

Figure

Tracking

2: SLR

is covered

African

region.

Coverage

within

The

only by up to four plots

show

the Commissioning

also

the

Phase

SLR

stations

actual

(Venice

and

tracked

none ERS-1

Orbit)

9O

6O

5O

- 120

-60

50

1-21

120

180

Table 2:ERS-1

ID q181 1863 1864 1873

Tracking

Network

Location

Name

Stations

System

Potsdam Maidanak Maidanak Simeiz

Germany, Europe Russia, Russia, Russia, Europe

fix fix fix fix

1884

Riga

Latvia, Europe

fix

1893 1953

Katzively Santiago de Cuba

Russia, Europe Cuba, Carribean

fix fix

7046 7080

Bear Lake Fort Davis

USA, North USA, North

7090 7097 7105 7109 7110

Yarragadee Easter Island Greenbelt Quincy Monument Peak

Australia Chile, Pacific USA, North America USA, North America USA, North America

7210 7236

Haleakala Wuhan

Hawaii, USA China, Asia

7403 7512 7542 7810 7811

Arequipa Kattavia Monte Venda Zimmerwald Borowiec

Chile, South America Greece, Europe Italy, Europe Switzerland, Europe Poland, Europe

7824

San Femando

Spain, Europe

fix

7831

Helwan

Egypt,

fix

7835 7837 7838 7839 7840 7843

Orasse Shanghai Simosato Herstmonceux Orroral

France, Europe China, Asia Japan, Asia Austria, Europe Great Britain, Europe Australia

7882 7883

Cabo San Lucas Ensenada

Mexico Mexico

TLRS-4 TLRS-4

7907 7918

Arequipa Greenbelt

Chile, South America USA, North America

fix TLRS-3

7939 8834

Matera Wettzen

Italy, Europe Germany, Europe

MOBLAS MTLRS TLRS

mobile laser system modular transportable laser system transportable laser system

Graz

America America

North Africa

1-22

TLRS-4 fix

MOBLAS-5 TLRS-2 MOBLAS-7 MOBLAS-8 MOBLAS-4 fix fix TLRS-3 MTLRS-1 MTLRS-2 fix fix

fix fix fix fix fix fix

fix fix

Tracking

Periods

910720 910724 920421 910723 920501 910731 920225 920401 920422 910808 911127 910906 910822 911121 920318 910731 911018 910802 911211 910725 920504 910731 910929 920404 910731 920328 910729 910719 910807 911002 920107 911029 920505

.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

910803 920226 910717 910816 910720 910722 910719 910725 911117 920119 920414 920319 911113 920109 911129 920107 920404 910724 910918 920305

.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

910829 911028 920226 920515 910922 911015 910822 920124

920311

920306

920222 911010 920426 910917 910818 911006 920121 920327 911131

910729 911117 920127

911126 920208 911129 920115

910918

Figure

3: SLR

Tracking

Coverage

within

the First

Ice Phase

(Ice

Orbit)

90

6O

3O

..

,_ k./---_

.--.

--_0

V_

_.,

"

.W'/"(L

_........ -7_--,_-..

_._l_.

'V

)

,

,

. _i

_ _



I

_

_-._.

_"

I

/

.-_.,._:

_-/

....

_ '_ _"_.;'_k-'-_

-_,

I

-60

,

-90 -180

I

-120

Figure

4: SLR

-60

Tracking

0

Coverage

within

the first Month

60

120

of the Multidisciplinary

180

Phase

i

-180

-120

-60

0

1-23

60

120

180

!

3. Data Soon

Flow

after

to D-PAF

the tracking

a first data

Systems of a satellite

screening

then from each the measurements

pass the SLR

is performed

data is preprocessed

by most of the stations;

pass about 50 data were compressed

points were selected (quick-look (Q/L) into onsite normal points (ONP).

This Q/L or ONP data is then transmitted to the Data two centers collecting and forwarding the data: the data ted via

existing

second center the European these them The telex,

internal

links

to the CDSLR

transmission

Span,

Bitnet,

is performed Intemet,

from

CDDIS

and

by

It has to be noted,

that

able for all partners. performed the EDC According ties

(PAF)

and

merges

all three

parties

In the activities

to the work D-PAF

using

all

distribution

them

currently

the full tracking data or Span. Again D-PAF

exchange

the

their

mission

the four

the telecommunication

system

management

the trackfng subsystems:

and

system orbit

data

to have EDC

European

generation

orbit earth

determination methods, quality prediction different gravity solution

to CDDIS;

and

merges

communication

links:

set (full-rate (FR)) is retrieves the FR data files.

the full was

processing

of operational tasks

information

avail-

not established

D-PAF

D-PAF

and archiving

precision

orbital

has

faciliproducts

set up a number

system

(TOS),

of incoming

which

data,

consists

generation

of the

subsystem: differential orbit correction by numerical control, generation of numerical and graphical products orbit

extrapolation,

generation

of orbit

equation

systems,

1-24

quality

control

following

of normal

points

integration

prediction

formats, generation of time bias functions, quality control modelling subsystem: processing of surface gravity data, of normal

all data from China. From

(TCS)

determination

subsystem:

the

(DMS)

preprocessing subsystem: preprocessing from laser and altimeter data orbit

received

when

and for a processing of SAR data. In order to fulfill these of processing systems. Used for the orbit generation are:

the data

forwarded

available

with the directly

of the

between for

or

III and ftp.

beginning too.

is responsible

then

the collected data files on a daily basis have sent their data directly to D-PAF.

GE/MARK

EDC

and

(EDC) at DGFI. EDC is collecting from stations in Russia, Cuba and

After having applied all necessary corrections transmitted to the two centers, usually by CCT sets

data)

Analysis Center at D-PAF. There are from the CDSLR stations is transmit-

headquarters

is at the European Data Center systems (EUROLAS) and also

two data centers D-PAF is retrieving with data from those stations which data

at the station:

reduction

sets

in and

4. Station

Performance

At D-PAF

all incoming

data

is checked

for quality

and statistics

are generated.

On

a weekly basis Q/L reports are generated presenting incoming data statistics and data quality check information. This paragraphs will present some of the statistical information. As already served by

seen in figures 2 to 4 only a small part of the ERS-1 orbital path is obthe ground stations and this is mainly located north of the equator.

Especially over South America second phase and in the first tracking From the

in that part the figures

stations,

while

7 presents

only

week.

mainly

Figure

in figure

about

were

only

in the number

shows clearly was not in

is visible

For an arc of 7 days

the ascending

arcs

can be tracked

the variation

trend

that

a few descending

passes

It ranges from 0 to 13 and March 92, when ERS-1 corresponding

from the first phase there

to the is no

of the world.

it is also visible

is that the descending Figure

the situation became worse month of the multidisciplinary

tracked.

during

of stations

The

observed

reason

probably

daylight. observing

ERS-1

per day.

9 which

SLR

depicts

the

number

of passes

are available.

Stations

per Day

14

,tLil 0 0

100

150

o 3OO

200

elapse time [days]

1-25

by

a decrease for the months October 91 to sunlight for many of the stations. A

30 to 60 passes

7:ERS-1

arcs were

o=,,

per

There is also a wide range in the number of passes being tracked by different stations. As can be seen from figure 8 this varies from 1 to 217. But one has to keep in mind that not all stations were continously tracking, some were occupied only for a short while (e.g. Monte Venda, see table 2). The tracking contribution of some stations to the total ERS-1 tracking ranges from a few percent to almost 15 percent (Grasse). Figure 8: Acquired

ERS-1 SLR Passes per Station

st.R 0/t. o.d ONP Data (92/0S/13) Gralllle

Herstrnon. Potsdam

I I II IIIII I I II Iiiii I []II mmmmm

I I mi IIII [][] II III

m

m

217

,, I



i:-::i!!/.(.Fi_ ::_i_:_ _i-.-_: !ii-.::! _i

Yarragadee

mmmIl

I I II

Gmz

,,

Maul J4elwan

Mon. Peak

ii:,.'-:" _ :..;.:_'!_ _!_::'_ g9

:);i-.:_i ?1_4

Greenbelt Maldanak

71

Riga S. de

Cuba

[]

Zlmmerwald

5

Shanghai Slmasato

[] I [] 33

Mh Venda

Matera Qulnoy

3

Wuhan

28

Bear Lake

$

Easter Iw.

;

L

Adrmqulpa

N

Slmelz Kattavla

_a.S.Luoas WeHzell S.Femando

Orrorol rnsenada Mc.

Donald

Borowte© Greenbelt

Maldanak Arwquipa

o

50

100

150

200

no. of passes

:I_uROt_S

_Na,o-net

_Ru,,;*.

_O,h.,-.

DGFI Not only the data quantity is varying different. There are some stations with cm and less. Figure mal point accuracy

from single

10 depicts the station for some stations.

station to station, but also the data quality is shot precision of about 20 cm and others with 1

performance

1-26

in terms

of single

shot and

15 sec. nor-

Figure

9: Acquired

ERS-1

SLR Passes

per We_k

8O

70

DGFI

O 10.

Figure [¢m]

10:ERS-1

Tracking

Precision

per Station

(cm)

26

dam type

/

stad(mnarn8

1-27

15 soe. _ormnl truant

5. Orbit

Determination

In order generated

to provide ERS-1 and distributed:

orbit

predictions

at D-PAF

and

precision

time

orbits

the following

ERS-1

orbital

products

are routinely

bias functions

preliminary orbits precise orbits The

latter

details

two

about

proctucts the

description (1990).

of

The

data

overall

are official

ERS-1

all

radar

orbit

ESA

products,

predictions

altimeter

can

and

tracking

flow for the orbit determination Figure Int.

Organis.

11: Orbit ESA

while be

data

in Kt_nig

products

processes

Determination

Stations

the first one

found

is an internal et.al.

(1992).

is presented

is explained Data

in Bosch

in figure

Flow

SLR Stations

Data

Reception

Dlstrtb.

of Orbit

Reformatting

Predictions

RAFD

Screening

Orbit

Prediction

Plausibility Merging, Normal

Generation

Check Screening Point Generation

Orbit Adjustment Quality Control Comp.

with

Archiving

Submission

1-28

MMCC

of Orbits

of Orbits

one.

Orbits

More

A detailled

11.

et.al.

The

processing

schemes

are outlined

in figure

Figure

12 for manoeuvre

12: Processing

(for manoeuvre F

S

S

D

M

I

D

I

I

I

orbit

I

I

arcs

PRC

I

I

I

I

I

5-7

day orbits

_--

5-7

day overlap

arcs

Orbit

I

I

I

I

I

I

I

I

I

I

I

I

data

and

on a weekly are

points

be seen

periods

the

usually

basis

by using

available

from

figure

weekly

arcs

Q/L

within

(15 sec. bins for SLR,

six orbital

coefficients, stations

I

I

I

I

I

I

I

I

two

plus

radar

altimeter

Both

for RA-FD)

data

another

and

weeks.

10 sec. bins

14, the use of RA-FD

are computed

laser

elements

altimeter

range

at epoch,

improves

7-day

data after

fast

sets

are

first qual-

the global

arc which

bias

one

solar

radiation

cover-

is overlap-

orbital

(if RA-FD

data

is used)

and

of bad

data

cm. The

Figure 13 presents the quality control results from overlapping large values in November 91 are resulting from a period with few

coefficient, station

daily/halfcoordinates

only).

resulting

a very

I

®

The

only

I

days and is used for quality assessment (see figure 12). The models for the orbit are described in Zhu and Reigber (1991) and Massmann et.al. (1992). Solve-for are

drag

I

@

free

(for new

I

@

I

(3

In manoeuvre

daily

I

@

for arc with only a few SLR passes.

parameters

I

@

age especially

ping by 3-4 determination

I

@

I

are computed

As can

->

Determination

into normal

ity checks.

I

@

_i

(RA-FD)

I

@

@

compressed

I

I

7 day overlap

delivery

D

I

_ _'///ii/II/ilIA.---

@

orbits

M

I

10 days

_-_I

7 day orbits

Preliminary

S

I

2

PRL

5.1 Preliminary

S

I

I0 days

,

3--4 days

Scheme

F

I

, orbit

periods.

free periods)

M

I

free

laser

fits are usually

passes.

The

around

same

50-80

cm, in case

hold for the end of December

1-29

coverage

up to 130

arcs for 1991 and 1992. no useful RA-FD data and 91.

Figure

13: PRL

Precision

Assessment

by Overlapping

PRL

Radial

RMS

Arc Comparison

Differences

Ioo

I PRL

Cross-track

RMS

Differences

PRL

40(

_x

2ix

loc

1-30

Cress-treck

RILLS

Differences

Figure

14: Input

Example

for the Preliminary

(January

20-27,

• .go

Orbit

Determination

1992)

!

6O

3O

\

i

i

I -go -1BO

-120

5.2 Precise The

Orbit

precise

15 sec.

availability in between

0

12o

60

18(

Determination

orbits

normal

-60

are based

points.

The

up to now orbits

only

on laser

are computed

with

data

which

a delay

of all FR laser data at D-PAF. The arc length two manoeuvres and is usually in the range

have

of 3-6

been

months

compressed depending

into on the

is choosen to fit as good as possible of 5-7 days. The quality control is

also performed by overlapping arc comparison. Models and solve-for parameters are the same as for the preliminary orbits, except the earth rotation parameters, the geomagnetic indices and the solar flux data, which are now the official final values instead of predicted or preliminary ones. Figure 15 presents now while figure comparing erally one due

figure 13 and 16 one has to be careful, can say that the PRC results are more

to better

ponent

the resulting rms orbital fit values for the precise orbit arcs generated 16 shows the quality estimates from the overlapping arc comparison.

input

of the orbit

data seems

and

a more

intensive

to be accurate

data

because the arcs are not the same. But genhomogeneous and do not show large spikes handling.

to 50 to 60 cm.

1-31

up to When

The

most

interesting

radial

com-

Figure

15: PRC Orbital

Fits

(_)

F

i

i +

I

200

210

!

i

.....

220

230

240 DlyI

Figure

16: PRC

Precision

250

260

270

In 1991

280

q,_w_,_m._

Assessment

by Overlapping

Arc

,_e_,

Comparison

PRC Radial RMS Differences m)

r

18(

16(

-----

14(

/

12: 10(

---

,

:

_

I

2OO

210

220

230

PRC Cross-Track RMS Differences

_

200 (°m)

1_.

i

|

I

IR--

240

II

_! 250

260

270

PRO Along-Track RMS Differences SOo_orn)

.

O_m In 111_

_

1-32

270

6. Conclusion Soon

after

the ERS-1

launch

DGFI/D-PAF

ed its capability to generate good SLR stations have demonstrated tracking

there

preliminary and the SLR data.

and demonstrat-

quality orbits on an operational basis. On the other hand the their capability to track ERS-1 and provide high quality

to now

are

a few

precise

orbits

there

are only

things

that

is very

a few

could

much

stations

identified

Russia.

The temporal passes) were capability

coverage suffers from tracked by the stations.

of the station,

to the

ing periods information

data

partly

coverage

the fact A reason

tracking

spatial

hemisphere

accuracy and

of the

temporal

of which

generated

coverage

only

one

of

has a

only one tracking station can be found and the northern hemisphere some gaps can be

that mostly the for that is partly

restrictions

it is difficult

with high solar activity. derived from Spot2 Doris

The

by the

in the southern

record. On the African continent in North Africa (Helwan). On

over

be improved:

limited

good tracking that is located

Due

the orbit computations

data.

Nevertheless

Up

has started

to model

and partly large

ERS-1

ascending ERS-1 arcs (night the missing daylight tracking

man drag

power

problems.

perturbation,

Tests are initiated to improve this data or by using crossover altimeter

especially

by using data.

dur-

either

drag

References Bosch,

W.,

Flechtner,

F.,K_Snig,

Wilmes, H., 1990. ment. D-PAF/DGFI, K_Snig, R., Li, H., racy

Massmann,

of ERS-1

Orbit

Massmann, F.-H., Reigber, 1 Orbit Determination Symposium Zhu,

R.,

Massmann,

The German Document F.-H.,

F.-H.,

PAF for ERS-I: ERS-D-PSD-30000, Raimondo,

Predictions,

Positioning,

S.Y. and Reigber, Ch., D-PAF. D-PAF/DGFI,

1991. The Document

and

Reigber,

Ch.,

Schwintzer,

P.

Product Specification 1.1, Miinchen. Ch.,

1992.

On

and

Docu-

the

Accu-

this proceedings

Ch., Raimondo, at D-PAF,

on Satellite

J.C.

Reigber, RAT Issue

J.C. and Rajasenan, C., 1992. Operational ERSProceedings of the Sixth International Geodetic Columbus,

Ohio

r.-

German PAF for ERS-I: ERS-I ERS-D-STD-31101, Miinchen.

1-33

Standards

used

at

N94-15555 Laser

ranging

application

to time transfer

and to other Japanese Hiroo

KUNIMORI*,

" Commumcations

Research

Laboratory,

Safety

Academy,

satellite

space programs

Fujinobu TAKAHASHI*, and Atsushi YAMAMOTO**

Tokyo ""Maritime

using geodetic

Toshikazu

4-2-1 Nukui-kita,

ITABE*

Koganei-shi,

184 Japan

Wakaba-Cho

5-1 Kure-shi,

Hiroshima

737 Japan

ABSTRACT Communications Research Laboratory (CRL) has been developing a laser time transfer system using a satellite laser ranging (SLR) system. We propose Japanese geodetic satellite 'AJISAI', launched in 1986 as a target satellite. The surface is covered not only with comer cube reflectors but also with mirrors. The mirrors are originally designed for observation of flushing solar light _'eflected by the separate mirrors while the satellite is 'spinning. In the experiment, synchronized laser pulses are transferred via specified mirror from one station to another during the satellite is up on the horizon to both stations. The system is based on the epoch timing ranging system with 40 ps ranging precision, connected together with UTC(CRL). Simulation study indicates that two stations at thousands of km distance from each other can be linked

with

signal

from AJISAI AJISAI pass.

strength

mirrors

give

of more many

than

chances

10 photons

and

to two stations

the distributed to link each

images other

of laser during

In other topics on the application to Japanese space programs, Retro-reflector for Advanced Earth Observation Satellite (ADEOS) and RendDezVous docking Experimental Technology Satellite-VII (ETS-VII) are briefly presented.

1. Laser

Time

Transfer

via Geodetic

Satellite

beam

a single

In Space mission of

AJISAI

1.1 Introduction Users of time comparison techniques

and frequency standards have been able to take a variety of time even if they pursue the highest accuracy. The precision of GPS common

view observation, for example, have precision have been attained by radio (Ref.1). The requirement space geodesy, relativity

got to several few nano seconds techniques such as two way time

and 1 ns or higher transfer via satellite

of such an extremely high precision may exist in deep space navigation, physics and astronomy which always demand extreme precision and

accuracy in their measurements. In addition to the radio techniques, time transfer using optical pulses has recently been suggested as one of potential for giving precision of sub-nano or picosecond time transfer to overcome the maximum bandwidth of radio waves (Ref.2). Satellite laser ranging (SLR) have progressed as one of space geodetic techniques in the field of global geodesy and geophysics. Some SLR stations measure the range to as precise as 30 ps or higher (Refs. 3,4). Since the SLR system is essentially a highly precise epoch recorder transmitting and receiving of optical pulses, a combination of two or more SLR systems has potential as a highly precise time comparison and transfer system.

1-34

Illlqdlll_

m

¢_cit

(a) Fig.1

There

are

(b)

Configuration of Laser time transfer (a) One way uplink (LASSO) (b) Two

two

configurations

considered

way

for ground-based-laser

time

transfer

system

using a satellite (See Fig.l). The first is called the one-way up-link configuration. Both SLR stations transmit a laser pulse to the time interval unit (TIU) on the satellite and the TIU measures the interval between their arrival times. Such an optical time comparison has been performed in Europe in the LASSO experiment (Refs. 5,6). In the European phase, laser echoes were successfully received at Grasse, European laser ranging sites. The second is the two-way configuration. laser pulse to the other station via a satellite with mirror reflectors. The cube reflectors for conventional SLR, and they are used to determine station to the satellite'to give support to the time transfer solution. In the

two-way

configuration

concept, calibration system performance for two-way 1.2 Target

using

methods, is based

the Japanese

SLR

satellite

LASSO experiment of France and at other Each station transmits a satellite also has comer the distance from each this paper, we examines

AJISAI,

including

time

signal strength and spatial distribution of the reflected on the CRL SLR system (Refs. 7,8), including additions

time transfer. satellite

AJISAI Table

Launch

1 • Major

Date:

specifications

August,

1986

Configuration:

Polyhedron

Weight: Comer cubes:

685kg 1436 pieces,

Number

of mirrors

inscribed

:318 pieces,

40 rpm Altitude Inclination

in sphere

Effective

Reflective efficiency: 0.85, Duration of a flash :. 5 msec, Spin rate: Launch orbit:

of AJISAI

of diameter

2.15

area :91.2 cm 2

Curvature

of mirrors

Brightness

:

Rate of flashing

:8.4-9

1.5-3.5

1-35

Eccentricity

m

star mag.

: 2 Hz

1500 km, 50 deg,

m

0.001

transfer

beam. The to be made

z





Vlau_ sp_'.

,

Fig.2

w_h mirror1

Arrangement of mirrors on AJISAI Parts of an 8.5m radius virtual sphere (mirror) is placed on the surface of the i-m radius AJISAI maintaining their latitude angles.

AJISAI is the Japanese geodetic satellite launched in 1986 by the National Space Development Agency (NASDA). The major specifications of AJISAI are listed in Table 1. The satellite is a hollow sphere 2:15 m in diameter and weighs 685 kg. The surface is covered with mirrors

as well as corner

cube reflectors.

The orbit is a circular

with an inclination

of 50 degrees

and an altitude of 1500 km (Ref.9). The mirrors were originally designed for observing flashing solar light reflected by separate mirrors while the satellite is spinning at about 40 rpm. Each mirror is a piece of a surface with a radius of curvature of 8.5 m. Three elements from the same latitude of the 8.5-m sphere at longitudes of 0, 120 and 240 degrees areput on thesurface Of a 1-m radius satellite retaining the phase angle of the original sphere (See Fig.2). The latitudes from which every set of three is taken are uniformly selected from the original surface so that a distant observer rotation. 1.3 Two way

can observe

developed

1 is ahead

from

the front

surface

flashing

three

times

per

time transfer

The equation being

the light reflected

for two-way

in radio

of clock

frequency

2) between

dt2= (tertle+tl-t2)/2

time transfer band

the clocks

+ (RIe-R21)/2c

where

• (t;j =1, 2),

t,.

Epoch of pulse departure

via laser

pulse

is in principle

(Ref. 1). The time difference of stations

1 and

2 is given

the same

d_2 (: positive

as those

when

clock

by:

(1),

at station i measured by station clock i,

(i .NE.j_ Epoch of station i's pulse reflected station j measured by station clock j,

1-36

by a satellite

mirror

arriving

at

=

=

Re :

(i .NEd') One-way distance via the satellite, and

c:

speed

pulse to travel from station

to stationj

i

of light.

The last term (Rlz-Rzt)/2c travelling is related

of a laser

is expected

paths, but not negligible to the difference between

to be nearly zero because

of the good

symmetry

of

due to the effect on motions of both satellite and stations. It the satellite position at the times the two laser pulses arrive.

It approaches zero if the two pulses arrive at nearly the same time on the satellite. This is done by controlling the firing timing at both stations. Timing control within the precision of 1-ms is necessary to obtain a station to station link by a mirror of AJISAI with 20cm x 20cm size spinning at 40 rpm. If the laser fire timing is controllable to l_s, (RIz-Rz)/2c is down to a few ps on the assumption that both the predicted position of the satellite and the clock synchronization are known with the accuracy of less than 1 _ (300 m in distance) before the experiment. The Its firing control can be performed by a fully active mode-locked laser operation. In this observation mode, we can apply the geometric method rather than dynamical one to the orbit solution using SLR range data (Ref. 10). 1.4 Laser ranging system We use an active-passive

mode-locked

optical

pulses

100

ps

wide.

The

energy

is 100

mJ

per

pulse

and

repetition

rate

mode-locking firing

can

passive the

be controlled

by a saturable

Nd:YAG mode-locking

nominal

from

laser to generate is performed

repetition

7 to 14 pps

dye is replaced

by

by

with an active

real-time

nm wavelength

saturable

rate is 10 pulses

per

software.

component,

detector

can be also gated

temporarily

by the prediction

arrival

receiver,

The timing system consists of a high performance disciplined and a time interval unit (TIU). Either a GPS timing receiver

(CRL)

can keep

a resolution

the reference

to UTC.

Timing

epoch

is measured

is transmitted

to the SLR

we must

system

consider

If the

plate

in 20-ns

via a 500-m

optical

5-MHz

used

and

of UTC.

to the clock

long-term

of photo

efficiency The MCP time steps.

quartz oscillator, or cesium clock

via cables

and electronics.

signal

a GPS of UTC with

Furthermore, clock

of the atomic the laser

and the TIU

fiber.

signal

as the reference The

5-MHz

room. Then

monitored stability

Output

(50 MHz) of the quartz

frequency

and 1-pps delay

at the clock

in TIU.

signals

and phase room.

of the system.

1-37

1 pps signal

are combined

difference

It has

The

been

oscillator again

with respect installed

clock

transmitting

is monitored

an independent measurement system. Figure 3 is a block diagram of the monitoring in the experiment. The combined 5-MHz+l-pps signal from the cesium clock room

continuously

passive

for up to 4 stop events

that the reference

and receiving point is not at the telescope reference point. The cables and electronics delay between the atomic

back

The

of 40-ps.

Calibration system In the actual experiment,

reference

The

the synchronization

rise time, 8% quantum fraction discriminator.

of photons

dye.

second.

timing can be controlled on nanosecond level (Ref.11). The receiving telescope aperture has diameter of 1.5 m. A micro-channel

multiplier (MCP-PMT) is used for detector. It has a 300-ps and the transit time jitter of less than 30 ps with a constant

1.5.

532

by

system used is transmitted

is phase-locked

by the

is used

epoch

at SLR

for the

site

to the original

and are sent signals

to get the characteristics

are

of the

s

Optics[ Fiber (50O I)

5_z

$r.ar t T I Stop

IP_ > i

5 _z

C

_ I PPS

[-'-I TIC: T[ae cs:

Interval

Cesim

Co,rater

[

'

Standard

great tiniag Telescope _._z t.ro [ [J_ER flrs ti_ial

Fig.3

The epoch occurred target

latch point

at the telescope

its physical

distance

=(te;-tle+t;-t + {

signal

at each laser pulse point

moving

from reference

The time difference

d;2

Reference

reference

is put on the telescope

( intersection

clocks

including

in addition

Optical

path delay

tL_

Electronic

delay from firing

ti.y "

Electronic

delay from receiver

and

electronic

delay,

measurement temperature.

must

it will the

wotald

from

firing

axes).

The

events

calibration

of transmitting

beam

and

precisely. term

(d';2) is given

by:

in time

common

portable errors

defined input)

reference

point

in Eq. 1, point

point

unless

design

can

from

to receiver

be the

reference point

latch gate

is changed.

value

of parameter

two

dependence

on

signal

or

out

the

stations. strength

This and

1.6 Link budget ground

In the laser time transfer experiment, the number of photons station are calculated according to the following equation: Np=(E)Jhc)(Grer where

f ,2)'(16Ae4 JnZR,2ReZq,2q,2)

E" Energy h • Planck's !"1 •

in a pulse, constant,

Transmission

(Np) detected

at the remote

(3)

_ • Laser wavelength, c • Velocity of light, beam efficiency, 1-38

re • Receiving

beam

efficiency,

E

meter

To cancel

at

i,

i.

by a distance

collocated

at

i, and

at station

1 and 2, each

point

at station

at station

can be measured

optical

receiver coming

clock

delay

to telescope

latch gate

to stop-epoch

between

The optical

be stable

(2),

}/2

point to start-epoch

the time difference

have

r)

(laser

from telescope

be evaluated.

still

reflection

calibration

to the parameter

ti. b "

to calibrate

of two rotation

as if these

2 )/2 + (R,e-R21)/2C

path delay i,

by scales,

start and stop must be treated

point must be measured

between

system

axis, as it takes partial

Optical station

In order

monitoring

(tla-tlj,)-(t2_-tza,)+(tz._+t2.r)-(tL_+tx.

: (i=1,2),

in { } in the Eq.2

MONITOR

Target reflector reflectivity, Atmospheric transmission efficiency (one-way), Total reflector surface area, A r : Received aperture surface area, Station 1 to satellite distance, R2 : Station 2 to satellite distance, Transmission beam divergence, and Reflectors beam divergence for incident fight return

1"3:

r,: A, : R1 : qt : q, : Figure

4 shows

the expected

number

of photons

for AJISAI

mirror

reflection

and

the

satellite elevation from stations, assuming the parameters have the values listed in Table 2. We also assume the position of AJISAI is roughly in the middle of the two station's common sky. From stations 2500 km apart, AJISAI is observed at above 40 degrees and more than 10 photons are expected.

Table 2 • Parameters for estimating Satellite Energy

Altitude

: 1500km

of pulse

Wavelength Transmission

the number of ohotons

(E)

(AJISAI)

: 100 mJ

(1) efficiency

: 532 nm (r_) : 0.6

Target reflectivity (r 3) Atmospheric transmission

: 0.9 efficiency

Receiving (r 4)

efficiency

,: 0.6

Reflector

surface

area (A,)

: 0.04 m 2

Received

surface

area (A)

: 0.27 m 2 (60cm

Transmission

beam divergence

aperture)

(q,) : 5 arcsec

5O

10o N

_

=

Z

ELEVATION(DEG)

B0

,o

60

30

fON

=

G v

(r2): 0.3

o im

O ZO

10

ZO



0

w

DISTANCE

Fig .4

-

2000

!

i

/

4000 BI!I'WEI'.'H

gO00 S'I'ATIOHS

ttl

o 8OO0 (kin)

Elevation and Expected number of photon v.s. distance between stations in laser time transfer via AJISAI

1.7 Spatial distribution of laser reflection

1-39

We simulate the spatial distribution of laser reflection from possibility of an optical link for time synchronization between stations. rotation project

AJISAI to study the We calculate first the

phase and the incident angle of AJ'ISAI for a given transmitting station and epoch, then the image onto the ground from all mirrors visible to both stations. The calculation

continues

according

to the given

time step.

In Fig.5, each rectangle is the image of the instantaneous laser shot and the number by the rectangle In Fig.6,

three

steps of 50 msec (one AJISAI

START TIME TIME INTERVAL INITIAL PHASE

spin)

are illustrated

successively.

8TART TIME TIME INTERVAL INITIAL PHASE

1987.8.27 12:22:0 0.0 (INSTANTANEOUS) 0.000 (DIEG) -

tgR7.11.27 12:22:0 0.IS0 (S) 0.000 (O8(31

IIo

m

ad

laser reflection on the ground for an shows the reflector number on AJISAI.

...'_1_ 114 ""-

--'""" Z

"''"

'''"

°'"°

In''" .

-_

"

w"¢_

. AJISAI

",_l.I.o

_

EO-

,_,. '_"_.-,,.,_ _._._.

___-_.L_,_..__

"_

,t,,.

,¢_j,.__-4xIl.,i...._l,,

,s,.

J

.. ¢,,...dt,

:t 4o

,dl' f'"

,



---......._,,.

_'ll,:_..,.4j_.

" .... :IIL_/_', _ -"_';_" _'_"

.;

"'41,,. ,.

"1"-

,_._

o_lt...,,0t:"

40

_ ,°i._'. Reference

#"

10

Slsllon

°

"N

90 neference

St|il()r_

_pv.

z0 I10

Fig.5

110 Longitude

1.10 Essl

140

IiO

II;0

(dog)

Instantaneous spatial distribution of reflected images on the ground via AJISAI mirror reflector. (The number next to each image indicates the mirror number on AJISAI)

A several stations'

to get

percent the link

of East every

Jzo

i_

Longitude

Asia

AJISAI

Fig.6

can locate pass.

an image,

If the timing

to Japanese

_4o

Space

but there

of every once

shot

is a chance

for many

is as accurate

every

few

seconds

as the by

Program

Table 3 lists the satellite name and its launching schedule in Japanese program. The schedule is released before the H-II rocket engine explosion which least one year delay shift from original schedule. 2.1

t_o

(deg}

Spatial distribution of reflected images in three 50 msec steps. (The number next to each image indicates the mirror number on AJISAI)

prediction, time comparison can be performed on average simultaneously determining the rotation phase of AJISAI. 2. Application

El'It

R&D space will cause at

RIS

Retro-reflector Observing Satellite)

In Space (RIS) is one of missions on ADEOS (Advanced Earth satellite which is scheduled for launch in 1996. The orbit is a sun

synchronous sub-recurrent polar-orbit with an inclination of 98.6 deg. It has a period of 101 minutes and an altitude of approximately 800 km (Ref.12). RIS is a single element cube-corner

1-40

Table

3 R&D

Satellite

Schedule released Fiscal

Launching

(1992-

by NASDA

year/Satellite

...........

2001), Curved surface

in April 1992 name /_.,.

n .......................

1992

JERS-1,FMPT

1993

ETS-VI,

1994

IML-2,SFU

1995 1996

ADEOS cOMETS

1997

TRMM,

1998

JEM-1,2,

1999

JEM-n

2001

HOPE

__..

_lat_urface

SFU

ADEOS-II X: Direction of satellite movement Z: Nadir

Fig.7 Structure of RIS ( Effective diameter" 50cm

retro-reflector

with

a diameter

of

0.5 m designed

for

earth-satellite-earth

)

laser

long-path

absorption experiments. RIS is proposed by the National Institute for Environmental Studies collaborated with CRL. In the experiment, laser beam transmitted from a ground station is reflected by RIS and received at the ground station. The absorption of the intervening atmosphere is measured in the round-trip optical path. surface for one of three

Figure mirrors

7 shows the structure of RIS. We use a slightly curved mirror forming the retro-reflector, which diverges the reflected beam to

overcome

the velocity aberration caused by the satellite movement. We have proposed a simple spectroscopic method which utilizes the Doppler shift of the reflected beam resulting from the satellite movement for measuring the high resolution transmission spectrum of the atmosphere. The wider laser beam is used for illuminating satellite and for autonomous tracking the satellite by guiding camera. 2.2 ETS-VII NASDA

will

schedule

to launch

the Experimental

Technology

Satellite

(ETS-VII)

in

1998 whose objectives are development of space robotics and rendezvous control. It consists of two satellites, the chaser and the target, each has GPS receiver for orbit control of chaser. Tentative altitude is 550kin and communication to the ground will utilize the data relay satellite. We have proposed that each satellite should be loaded with corner of which is two-color sensitive and be monitored the rendezvous process will track the satellites with switching or simultaneously mode depending within

one beam

cube reflector set, one from the ground. We on two satellites being

or not. 3. CONCLUSION

We synchronization whose

timing

have

studied system

system

with

the

feasibility

a target

is connected

satellite

of

the

of

AJISAI

with UTC(CRL).

1-41

sub-nanosecond based

on CRL

precision laser

laser

ranging

time

system,

Simulation can be linked beam from chances

study

by laser

AJISAI

indicates

beam

mirrors

that two stations

with signal is uniformly

to link each other during a single

The system to the major determined

has been

geodetic

operating

satellites

with the precision

system space

from

AJISAI

in geodetic 1990.

of km distance

than 10 photons.

on the ground pass.

information

of several

has also for time program in future.

at thousands of more

distributed

timing, however, it bring on the precise satellite range data.

current Japanese

strength

from

The

each other

images

and two stations

It requires

of laser

have

the lasec control

many

of laser fire

of the orbit as well by using the self-returned mode

since

While global centimeters,

synchronization

CRL started the SLR observations

position

of the SLR station

we are re-designing mode

operation

has

our system

as well

as

been

so that

adapting

for

REFERENCES ,

2. .

4. 5. 6. 7. 8. 9. 10. 11. 12.

Kirchner D.,1991, Two-way Vol.79, No.7, pp.983-990 Leschiutta S.,1991, No.7, pp. 1001-1008.

Time

Time

Transfer

synchronization

via communication using

laser

Degnan JJ., 1985, Satellite Laser Ranging : Current Trans. Geosci. Remote Sensing. GE-23, 4, pp.398-4

satellites,

techniques, Status 13.

_,

and Future

Pearlman M., & al. 1989 , Ground-based laser ranging measurement technology, _g.09.13...QLC,,J_-COOLFONT, section2.1, pp. 1-26. Serene B. and Albertinoli P. 1979, The LASSO Experiment, Proc.PTI'I pp.145-167.

Vol.79, Prospects,

and

1 lth conference, D l[ i

Proc.7th

Kunimori H. & al. 1991, "Satellite ,pp.303-317.

of CRL.

ranging system",

IEEE

techniques

Veillet C. 1989, LASSO - the European Phase Aug.88-Sept.89, Workshop on Laser Ranging Instrumentation, Matera 2-6. Laser

_..Qg,_.L_,

International Z

Journal

Greene B.A. 1984, Epoch timing for laser ranging, Proc. 5th International Laser Ranging Inswumentation (Herstmoncex), pp.247-250.

Vol.38,

No.2

Workshop

on

Sasaki M.and Hashimoto H. 1987, Launch and observation program of the experimental geodetic satellite of Japan, IEEE Trans.Geo.Rem., GE-25, 5, pp.526-533. Yang F.M. 1986, The proposal of strictly simultaneous satellite laser ranging, The 6-th international workshop on laser ranging instrumentation, France, pp.549-557. Electro Optic Systems 1991, EOS-V1 Active/Active Laser Characterization. Sugirnoto,A., Minato A. and Sasano satellite, CREO'91 Baltimore,.

Y.,19_1,

1-42

Retro-reflector

in space

for the

ADEOS

N94-15556 APPLICATIONS

OF SLR

B.E. Schutz Center

For Space

Research

The University of Texas at Austin Austin, TX 78712, USA

INTRODUCTION Satellite

Laser

Ranging

10 meter-level

(SLR)

first generation

to the

began,

in part, as an interesting

science

systems

that

systems.

ments

component

has a rich

of modem,

now

history

These

produce space

systems

several

application

precision

of development evolved

which

began

with order

millimeter

single

of the new laser

geodesy,

which

in the

of magnitude

shot

range

technology

contributions

with

improve-

precisions.

has become

in turn enables

1960s

What

an essential to a variety

of

areas.

Modem

space

cision

geodesy

geodetic

Ranging butions

is the beneficiary

measurements.

Aside

of technological fro SLR

and

(LLR), Very Long Baseline Interferometry also. In recent years, the Global Positioning

developments

its closely

which

related

have

technique,

enabled Lunar

preLaser

(VLBI) has made prominent science contriSystem (GPS) has demonstrated a rapidly

growing popularity as the result of demonstrated low cost with high precision instrumentation. Other modem techniques such as DORIS have demonstrated the ability to make significant science contributions; furthermore, PRARE can be expected to contribute in its own right. An appropriate question is "why should are several answers, I offer the opinion

several techniques be financially supported"? that, in consideration of the broad science

the benefactors of space geodesy, no single tations of the science areas in which space ing. The more well-known rotation/orientation, gravity name

While there areas that are

technique can meet all the requirements geodesy contributes or has the potential

science areas include plate tectonics, (static and temporal), ocean circulation,

and/or expecfor contribut-

earthquake processes, Earth land and ice topography, to

a few applications.

It is unfortunate

that the modem

this view is usually diminishing reduce

the data

physical

encouraged

budgets.

The

space

by funding

techniques

to geodetic

techniques

competition,

are, for the

parameters

from

especially

most

several

are often part,

techniques

viewed

as competitive,

in an era of growing

complementary promotes

and

but

needs

the

but

ability

to

confidence

in the geo-

in the context

of the other

interpretations.

In the following

sections,

the current

techniques.

strengths

and

prospects

geodesy

The

SLR

limitations

applications of SLR

are reviewed

are reviewed

and

speculation

about

the

future

are offered.

SLR Summary Satellite transmitter

Laser

Ranging

to a target

measures and back.

the round-trip Current

time-of-flight

and near-term

1-43

satellite

for a laser targets

pulse

include

to travel

Starlette

from

(1000

a

km

altitude),

Ajisai

km).

Future

1992

launch).

lifetime offering

(1500

km),

satellites

include

The

measured a distinct

ERS-1

geodetic

(800

kin),

Lageos

TOPEX/POSEIDON satellites

(59000

(August

(Starlette,

Ajisai,

km)

1992

by the

extragalactic Precision

and

The

and

-2 (25000

and Lageos-2

(October

Etalon)

have

a long

orbital

in thousands of years to millions of years for the high altitude satellites, thereby advantage that the satellite will be available at no cost for a very long period of

on-board radio

Orbit

Etalon-1

launch)

Lageos

time. Only the ground segment requires operation and maintenance satellite techniques (GPS, DORIS, PRARE) rely on an active space mined

and

power

sources

system,

which

usually

several

will, presumably,

years.

support. segment

VLBI,

be available

By contrast, all other with a lifetime deter-

on the

other

hand,

uses

for a very long time.

Determination/Gravity

traditional

strength

of SLR

has been

in the ability

to determine

the orbits

of target

satellite

with high accuracy. In the case of TOPEX/POSEIDON, SLR is the primary means of precisely determining the orbit to the required 13 cm in the radial component in support of radar altimeter analyses. SLR tracking of NASA altimeter satellites has been an important element in the accomplishment dependency potentially SLR, tems.

of the respective on atmospheric able to provide

mission goals, SLR, however, has a distinct disadvantage transparency. Radiometric systems, such as GPS and essentially continuous tracking. Since TOPEX/POSEIDON

Created by DORIS, are includes

GPS and DORIS, it will provide a unique opportunity tO evaluate the performance of all sysThe focus of tracking systems on future satellites may be GPS with SLR as a backup. Nev-

ertheless,

it is important

to note

that

the passive

nature

of SLR

provides

an extremely

reliable

space segment--it simply will not fail except under catastrophic circumstances. A cautionary note for the future is in order. The risk of regarding SLR as a backup system encourages the reduction in operating systems for budgetary reasons, but as systems close, it becomes more difficult to reactivate them in the event that the backup mode must be initiated. Improvements

in the

gravity

field,

including

the gravitational

parameter

where SLR has made very significant contributions. The development tion for TOPEX/POSEIDON have benefitted from the SLR data base

GM,

are

other

areas

of new models in preparathat has been accumulated

over the years, some of which has been used in previous fields and some has not. Gravity fields used for GPS applications, such as the Department of Defense WGS-84, have used SLR data from Lageos gravity have

and Starlette coefficients. relied

which have particularly contributed to determination of low degree and order Current GPS applications use GEM-T3 or other, more recent fields, which

on SLR

data.

Furthermore,

the Etalon

satellites

effects that will be somewhat similar ular GPS effect of deep gravitational

to those influencing resonance.

SLR

to the study

has made

unique

contributions

tidal and non-tidal

variations.

These

studies

area-to-mass

satellites

(Lageos

and

the

opportunity

This

ratio

diminishment

effects. Furthermore, investigate the orbit years.

enhances

can be used

the GPS satellites,

of temporal

variations

have, in part, been Starlette),

which

to identify

the

gravitational

except

for the partic-

in the gravity

made possible

diminish

to study

by the nature

the nongravitationai

spectral

field,

content

both

of low forces.

of gravitational

the study of temporal changes in gravity has been enhanced by the ability evolution over long periods of time, spanning in some cases more than

to 15 i

| 1-44 E

Earth

Orientation/Reference

Frame

All satellite techniques are, conceptually, ter of mass. VLBI, by using extragalactic ever,

if VLBI

is used

to track

satellite techniques. The has been a consideration Rotation

Service

All space

artificial

satellites,

VLBI

would

have

with the cenof mass; how-

the sensitivity

of the

Reference

techniques

Frame.

have demonstrated

sensitivity

to polar

motion.

Current

comparisons

between polar motion series obtained from SLR and those obtained from VLBI show at the 0.5 milliarcsecond level. GPS developments are underway, including the proof International

other

SLR precision and long term continuity (Lageos was launched in 1976) in adopting the SLR origin to be the origin of the International Earth

Terrestrial

geodetic

able to define a reference frame coincident radio sources, is insensitive to the center

GPS Geodynamics

Service

(IGS)

slated

to begin

in June

agreement of concept

1992.

VLBI provides long term UT1 that cannot be matched by the satellite techniques at the present time. Lageos has been demonstrated to provide independent determinations of UT1 over a 50 day interval.

The

higher

altitude

of the Etalon

satellites

suggests

that a much

determination is possible, but the sparse SLR tracking of the Etalon limited demonstrations of the UT1 capability. It is worth emphasizing ratio of Lageos

and Etalon

The nongravitational satellites

to very

Comparisons ments, either

result

forces

in much

smaller

are a significant

short term UT1 (sub-daily

have been permanent

nongravitational

factor

longer

period

of UTI

satellites has allowed only that the low area to mass

forces

in the GPS satellites

than

the GPS satellites.

that limit the use of these

to a few days).

made between SLR and VLBI reference frames using collocated instruor mobile. These comparisons have shown agreement at the 2 cm level.

FUTURE With

the launch

years,

a second

of Lageos-2, Starlette

satellite

our knowledge

about

gravity

for science

models

lites,

for example,

only

high precision of the diverse

one satellite

process

variations

New

will be launched. in the gravity

improved

developments

quick-look constellation

can be ranged

of establishing

will be two Lageos

applications

will enable

tion of site positions. provide tracking

there

priorities

and two

Starlette

Etalon

and Lageos

satellites. have

field as well as contributing

and for precision Earth

rotation

in on-site

orbit

software

have

and enhanced

provided

much

Two Lageos more

rapid

the ability

data that can be used for science applications of satellites now avail,able for SLR tracking.

at a time and the involvement

a few

to the development

determination.

determination

Within

of the science

of of

satel-

determinaof SLR to

and to assure Nevertheless,

community

in the

is essential.

No single space geodetic technique can meet all of the application requirements of the science community. Each technique provides some unique contribution, though most of the techniques overlap in some areas. In an era of diminishing ance of techniques and the prospects of losing considered

as emphasis

budgets, the determination of the appropriate balor eliminating some science applications must be

is redistributed.

1-45

The

international

further

collaboration

collaboration

work, thereby understanding

in the

will enhance

assuring the continued of our planet.

SLR community

the prospects collection

1-46

has been

for long term

outstanding.

availability

of data for the purpose

The

promotion

of a global

of gaining

better

SLR

of net-

scientific

Timely

Issues

E

R_ B

zl m m_ E

N94Satellite

Signatures

in SLR

G.M. Royal

15557

Observations

Appleby

Greenwich Madingley

Cambridge

Observatory Road

CB3

0EZ,

UK

Abstract. We examine observations vation

the evidence for the detection of satellite-dependent obtained by the UK single-photon SLR system.

distributions

from

Ajisai

and Lageos

functions and from the characteristics tions. The effects of vamfing return mental

observations

photons

of Ajisai,

is achieved.

corrections

The

are developed

of the strengths

during

which

implications

s,ignatures in the laser range Models of the expected obser-

from

SLR system, are discused a range

of these

the published

for

spread

and compared with the observausing the models and by experi-

of return,

results

satellite

levels

from

single

system-dependent

to multiple-

centre

of mass

are discussed.

1. Introduction. The

UK SLR

system

sited

at Herstmonceux,

and

run

by the

Royal

routinely observes the primary tax'gets ERS-1, Lageos, Etalon-1 sai. The single-shot precision achieved by calibration ranging Tile detection and timing hardware has recently been upgraded Avalanche Epoch

Photodio(le

is derived

rangc return

(SPAD,

at present

Procha.zk_t

from

ctal,

a Maryland

1990), 4-stop

and event

Greenwich

Observatory,

and -2, Starlette and Ajiis close to 1 cm (1-sigma). to include a Single Photon

an HP 5370 timer,

time interred

which

is also

counter.

used

to make

measurements simultaneously and independently of the HP counter. Pass-averaged rates are in general fairly low, varying from a few percent from the Etalon satellites,

through

about

20%

fl'om

targets are deliberately ters in the laser path. return, carried becaxne

Lageos

kept Under

to up

to 50%

from

Ajisai.

Returns

from

to similarly low levels (about 10-15%) using such conditions we can describe the system

the

calibration

neutral density filas a single photorl

single photon detection system. A detailed study of the system error budget was out following the upgrade of the detector from a PMT. During this investigation it clear that the observational precision of in particular Lageos and Ajisai was consis-

tently worse than distribution of the

that of the retroreflector

calibration targets. It was considered arrays on the satellites would modify

likely that the the distribution

spacial of the

range residuals, when compared with those from the fiat calibration targets. In this paper we examine the evidence for detection of satellite signatures in our range observations, compare the observations with models of the expected distributions from a selection of those satellites

regularly

required 2. This

observed,

to reduce

the

and

discuss

observations

the

implications

to the

centres

in terms

of the

of mass

of the

range

residuals

appropriate

corrections

satellites

Observations. investigation

is based

upon

the

pass-by-pass

2-1

that

are

formed

during

the preprocessingstageto errors

in the

solves

for corrections

falling stage

predicted

outside

a 3-sigma

is fitted

shown

in Figure

cMibration values are

target plotted

distribution

Also

the

values

TaJICI[T

*

II)@Z

4

pass,

in the

Li

and

I, 11

1992,

these

residuals

iterative



distribution

the SLR Gaussian

system. The distributions,

600 m from of the fitted



kurtosis

would

be 0.0 and

:,%,..,,=. II

I_RS-

13@

?'.A) INI IIIg_

4 =

lai[lr=41.

,.

qla,

8

ZZ

1.1_6 IS

k'@rl_m,l.e7

I,_

-

- :--

' "

i

_|@1

g

rm

Ig

li|_

l@

ldt._l I_IGNAm

_lCVndl,f.4l

IKNi

]Ill 114

Zl

em

I_UIITI_43

Lageos

observed range which are also

l..II|

,_n- ,.,

1

to a

3.0 respectively.

-I@

IS

4 =

and

of ranges

li_-.tU.

i@e .58r.NA

bins.

_l?

j

lul

are used

in range

JIUIIIF=3.00

I@

•rao:igo

In a final

fitted distributions are shown, along and kurtosis. For a perfect Gaussian

•t

-I0

residuals

least-squares,

is a typical

11¢111|

.

to

iteratively

the residuals

the by

due

stage

proceedings).

performance,

by grouping

which

at each

@.34

x,,,._,

oil

this process,

rejecting

distribution

Figure

•-

=l

in the residuals

are used to make a final selection of the origin',.1 distributions and their fitted Gaussian distributions

i. I "*_1

•53.13'a

ra

Sinclair,

observed

about mean

of skewness

_lGllAi

ia

during

The standard deviations of the of the data, expressed as skewness

IINI

C lib

shown

board, distant relative to the

shown on each plot. with higher moments

and

for each to the

All trends

parameters,

che¢& on system

fitted Gaussian of the observed 1.

points.

are removed

(Appleby

as a useful

distribution

of the Examples

normal

of orbit-related

band and

distribution

the parameters observations.

on-site

of the satellite

a frequency

A normal

are

orbit

to a set

of pre-processing,

to form

compute

[email protected]

p*

Aj isa i

3..@ liT.

I;l;1¢ If ..e j..r.

r,l_

tu=r.=.jw Jill r

KInlT*3.S4

_Lolon-

l

_.

lIA m

d_lll

_,., p,

m

E_

k

-i@

g _10



I|

tm -I@

Figure

1. Observed

distributions

of range

2-2

residuals

from

calibration



and

satellite

IO

em

targets.

2.1

Discussion.

From

the

butions

distributions

of the

shown

calibration

in Figure

ranges

and

1, we make

those

the

following

fl-om Starlette

and

observations.

ERS-1

The

are clearly

distri-

symmetric

and well-fitted by the Gaussian distributions, but all have a significant 'tail' of observations outside the fitted curves. Skewness values for these 3 targets are between 0.05 and 0.1. The Lageos distribution is much less symmetric, and is less well fit by the Gaussian distribution. A chi-square goodness of fit test indicates significant departure, at a 5% level of significance, from the best-fit distribution shown in the plot. The results from Ajisai and Etalon 1 show large

asymmetry,

and

are

not

at all well

fit by the

Gaussian

distributions.

Of particular

significance to this investigation, are the 'widths' of the distributions, characterized by the standard deviations of the fitted distributions. Mean values of these standard deviations for a number of observations made during November and December 1991 arc given in the Table below. that

These the

largest,

mean

values

calibration the range

of standard

ranges

have

of standard

the

deviations smallest

deviations

confirm

scatter,

being

and

from

Target

the

impression

those

given

of Ajisai

and

in Figure Etalon-1

1, the

1.1 cm to 4.8 cm.

a llll-n

Calib

11

ERS-1

12

Starlette

16

LAGEOS

18

Ajisai Etalon

32 48

Before proceeding to investigate the hypothesis that satellite signatures arc present in our observations, we first consider the possible causes of the 'tail' in the distributions, particularly evident in the calibration and Starlette data. We remark here that the existence of this tail does

not

constitute

the

thrust

of our

argument

that

we are detecting

satellite

signatures

our observations, board. We must cause the SPAD

since the tail is also present in the calibration ranges from therefore rule out such a target-induced effect and consider or the laser. In an experiment primarily designed to quantify

time-walk

a large

using

under

neutral

density

filters

In this way the average shot, as deduced fl'om Figure

2, where

expected,

the

range

as the

are

in the

single-event

Prochazka

the

(1992,

displayed

standard

strengths,

average

detector

private

deviations

increasingly

of photons

form

of the

was carried

reaching

as before.

distributions

The with

the

out

detector.

photons

originating

plots

show,

increasing

expect the contribution number of photons

receives

as

signal

of the laser in the return nearer

to the

plots also demonstrate that the extent of the strength, suggesting an origin within the laser.

communication),

2-3

ranging

was varied from about 0.5 to 50 photons per rates. A selection of the results is given in

in histogram

pulse. The with signal

calibration

number

laser pulse-width we would to decrease with increasing

leading edge of the transmitted tail in the distributions decreases However

signal

number of photons the observed return

strength, since for a given to the observational jitter train,

to vary

results

a reduction

of return

in

a flat target as probable the system

points

out that

correct

optical

alignment

120

'

'

I

'

'

'

'

I

'

'

I

I

Illll

200

IOO

(a)

n = 0.5

(b)n = 6

= 12

a=

13mm

150 0

o

60 100

E z

40

50 20

I

I

[

I

|

I

L___-_.A_

20

_

__ __l____d--k__L__.L

30

d0

20

30

40

250

200

0

200 i

--

t[

it J

150

o

._

Z]

100 I00

_

-

]

--

5o

0 20 Range

Figure

30 Residual

2. Calibration

40

20

(cm)

distributions

Range

as a function

2-4

of average

numbers

30 Residual

40 (cm)

u of returning

photons.

of the

SPAD

Resolution 3.

detector

of this problem

Satellite

We now

take

strength

system

with

the

convolved

pulse,

from

system-signature

it models of Lageos

upon

within

the

chip.

that

effects

the

of Lageos.

of changing

and

not

In this

which

(Fitzmanrice

of a computer

diffraction

cross

satellite,

estimate

pattern

tests

in polar

of each

row of reflectors

section

of the

to caxry of the mainly

et al 1977.)

simulation

distribution

parameters give, for the particular cubes, in rows, contributing to the

distance

effective

of the

polarisation,

its shape

Ajisa_ we use the results

the

centre

calibration

satellite return signatures, by For Lageos, we take the model

far-field

is the optical

We use

from

the

of expected and Ajisai.

row-by-row

Also given

of gravity.

distances

signature

of non-uniformity

experimentation.

single-photon

based

signal.

centre

we ignore

effects

in Fitzmaurice et al (1977). The cross-section and number of corner

of returning of known

further

1, and develop from the spread functions

presented the lidar

spacecraft

rings,

standard,

para_neters

orientation, orientation,

possible

Models.

as our

of cross-section

to avoid

awaits

Signature

shown in Figure convolution with

the

is essential

out

shape affects

To model

carried

out

fl'om

cubes

in their

a convolution

of our

of the

returning

pulse

the

amplitude

the

return

signatures

and

Hashimoto

by Sasald

of the

(1987). They find that the number of retroreflector sets contributing to the return signal from a given single laser pulse can only be 1, 2 or 3.5, and give the computed pulse shape in each of these 3 cases. The laser used in their simulation is gaussian in profile, of standard deviation 33 ps. From cross-section's

the published profiles, we can infer the spread distributions, consisting of lidar and dista.nccs from spa.cecrMt ccntrc of grltvity. We now have the informa.tion

required to ca.rry out a convolution with our system signature, in the same way as for Lagcos. Wc assume that the rapid spin rate of Ajisai, of 40 rpm (Sasaki and H,'Lshimoto, 1987) will ensure

that

convolve

for cvcry

our

pass

system

all 3 possible

signature

with

orientations

each

of the slitellite

of the spread

will be sampled.

distributions,

and

sum

the

Wc thus resulting

3 distributions. The

results

of the simulations

3(a) and (b), where butions, also shown convolving

our

and

Ajisai

the quoted standard deviations on the plots. For completeness

system

o

for Lageos

scpaz'ately

60

_' --

'

with

I '

each

'

of the

'

'

are shown

in histogram

form

are those of the fitted Gaussian distriwc also present in Figure 3 the result of 3 orientations

I '

'

of Ajisai.

'

'

o" =

17

.__ I

I

I ' mrn

1

-

©

40

--

20

--

0 0

,_a _. --

0

__L_ Figure

3.

(a)

SimulaLcd

2-5

L_gcos

in Figures

rangc

residual

distributions.

of the

SPAD detector is essential to avoid Resolution of this problem awaits further 3.

Satellite

We now

Signature

take

as our

shown in Figure convolution with of cross-section

the rings,

standard,

paa'aaneters

spacecraft-centre of known

of non-uniformity

within

the

chip.

Models. single-photon

1, and develop from the spread functions

orientation, presented orientation, the lidar strength of returning

possible effects experimentation.

based

system-signature

it models of Lageos

upon

the

of expected and Ajisai.

row-by-row

fax-field

calibration

distribution

satellite return signatures, by For Lageos, we take the model diffraction

pattern

tests

in polar

in Fitzmaurice et al (1977). The parameters give, for the particula_ cross-section and number of corner cubes, in rows, contributing to the signal. Also given is the opticM distance of each row of reflectors from of gravity.

distances

from

We use the

centre

the

effective

of the

cross

satellite,

system signature with that of Lageos. In this estimate we ignore the effects of changing polarisation, which

section

to cazry

of the

cubes

in their

a convolution

of our

of the returning the amplitude

pulse of the

out

of the shape mainly affects

convolved pulse, and not its shape (Fitzmaurice et al 1977.) To model the return signatures from Ajisai we use the results of a computer simulation carried out by Sasaki and Hashimoto (1987). They a given single these 3 cases. 33 ps. From cross-sections

find that the number of retroreflector sets contributing to the return signal from laser pulse can only be 1, 2 or "1.5, and give the computed pulse shape in each of The laser used in their simulation is gaussian in profile, of standard deviation the published and distances

required to carry We assume that ensurc

that

convolve our 3 distributions. The 3(a)

profiles, we Call infer the spread distributions, from spacecraft centre of gravity, gTe now

consisting of lidar have the information

out a convolution with our system signature, in the same way as for Lageos. the rapid spin rate of Ajisai, of 40 rpm (Sasaki and Hashimoto, 1987) will

for every

pass

system

all 3 possible

signature

with

orientations

each

of the

results of the simulations for Lageos and (b), where the quoted standard

butions, also shown on the plots. convolving our system separately

of the satellite spread

mad Ajisai deviations

For completeness with each of the

will be sampled.

distributions,

axe shown are those

and

sum

in histogram of the fitted

we also present in Figure 3 orientations of Ajisai.

the

Wc thus resulting

form in Figures Gaussian distri3 the

result

of

01

o

60 a

=

17

mm

l> 0

z

4O o 0 I.-,

,.o

20

X; 0

-_

Figurc

3. (a) Simulated

2-6

I

I

]

I

Lagcos r,'mgc rcsidu'al distribution

I

5O

6o

=

(_)

23

(b)

mm

'u

=

30

5O 40

_, .o

4o 30

,> o

30

2O ,,D

E

10 lO

20

30

20

dO 140

5O

I

'

'

'

'

I

'

'

(c) II

' _r

' _

I

'

'

33

'

I

30

'

'

'

'

I

(a+b+c)

I

40

....

1 cr

=

20

120

40

I00

.2

30 0

O0

'

"5 GO ,-, 20 E

,10

1o 2O

20 Range

30 Residual

Figure

3.

20

40

Range

(cm)

(b)

Simulated

Ajisai

2-7

range

residual

distributions.

30 Residual

40 (cm)

'

3.1

Discussion.

The

standard

deviation

mean of 1.8 cm, The underestimate

neglect of atmospheric the satellite, and the Ajisai well

return with

of signature satellite

photon

data

deviations

results.

with

pass

orientation

Multi-photon

(1.7 cm)

is close

There

of between

is some

circumstances,

which

or orientations

2.3 and

evidence

discussion pulse

is based

upon

to be a random distribution

Observations

Experiments

and

were

in received levels was beam and

carried

may

for a given

with

return

event that

energies

be due

ground

to filter

observations

the experimental passes are shown in histogram Gaussian distributions. We calculate

the average

taken

out using

Ajisai

since

from a population response.

it is relatively

them to solve for corrections did not remove all trcnds biases which into a number

compare

observations

to the

dominance

of of a

of true

photon

level;

the detected

formed

by the convolution

We now consider

the effects

in order to quantify the variation of the mean

in each

easy

range

a large

to the predicted orbit. from the range residuals,

However, it was indicating the

segment

separately.

The

resulting

return

rates

at intervals

scatter

plot

each of the deviations

of 30 seconds

throughout

returns and the numbcrs of pre-rct_rn per second, the true percentage return

measurements*100)/(30*10

that such

in several of the 30-second intervals the calculated return levels we cannot reliably estimate the mean

may

be fax in excess

16 estimated

for

a near

used these 30-second mean values to estimate the mean the observations in our 6 segments, and these averages

2-8

for one

- number

of noise

return number 100%

rate.

of

six segments of the fitted

the noise rate in

events)

On the assumption that the quantum efficiency of the SPAD is 20%, we calculate corrected return rates the average numbers of photons in each return. However

of the

variation

from single photon to multiple photon by altering the divergence of the laser The observations wcre filtered in the

is shown in Figure 4. The residuals from form in Figure 5, along with the standard

percentage

to obtain

varied during the passes. We found that it was (G) of segments, and use the processing software

passes by counting the numbers of satcllite detections. Given that the laser fires 10 shots each 30-second interval is then (number

Ajisai

reduction.

of systematic range to divide each pass the

fading induced by The models of the

track.

at the single

of the satellite

encrgy from the large target. The variation achieved during the expcrimental passes hence the energy density at the satellite.

standard way, by using found that this process presence necessary

observational

3.3 cm, which

in the

of a larger number of photons reaching the single-photon detector, subsequent systematic effects caused by a signal-strength-dependent reflection distance to the satellite. 4.1

to our

and observed histograms is similar. may be attributed to various causes;

Returns.

is considered

of the laser

give standard

observational

particular

The foregoing

Lageos

turbulence (Gardner, 197G); neglect of coherency single satellite orientation chosen for the model.

signatures

the

variations

4.

of the simulated

and the appearance of the simulated of the observed scatter by our model

from these we found

rate was nearly 100%. At of returning photons, which Where

possible,

numbers of photons are shown in Figure

we have

contributing to 5. For those 2

m

1

I

r

(cuo)

l_np!so_l 2-9

a+u+_I

140 ....

I''

120

n

=

''1''''1

....

]''

6

o"

''1'''

=

20

F

mm

I00

I00

L20

n = 0

o = 20 mm

;

OO

• N 6O ,o

_

4o

40

m

2O

_

,I,,,,1 0 -1_

,

,

,

,

I

, .-

,

,

-5

-I0

0

[_ lO

0

15

__Ji--L-L_/_[_L

-15

140

• ,

-lO

-5

,

,

J ,

0

,5

I0

I ....

I ....

I'''

i

I

I 15

14Q

....

I ' ' ' ' I r'T-m_]_ ....

I ....

''''l

I ' ' °

....

I ....

J

P'O

I00

n _'[_

ll_ 1

a = 13

I00

(r = IG

O0

z

Jl /% IG

120

F--

40

2O

2O I--

60

o

0 -15

"-..- 10

-5

0

fi,

10

15

JJ-J-J_

-15

14OL_ 120 140f_

_Lj_j_I"

120L

J--LLLL -10

°

'

'

1_

I

=

-to

....

I

0

*

'

'

'

I

'

'

'

'

i

' In '

=I

' 15'

*

'

I

'

'

'

'

I'"_

'

'

'

it= I '

' 16'

'

I

'

'

'

'

'

t

'

'

'

"T

,]

=

18

--_

'

100

lO01--

00'

w o

'

15

fJ o"

'

LLLL_ lO

b

"

N

GO

_

4o

60

4O

2O

20

.... 0

,,,,1,,,,

0

-lg

iO -iO

Figurc

I,,,,i

,1,

5.

-5Range

0 Residual

15

-15

( cn5 0

Distribution of reazgc turning photons.

rcsiduMs

from

2-10

Ajisai

pass

-IO

as

-5 Raw8©

,_ functioll

0 Residual

of

5 (el

avcragc

,0 numbers

10

15

n of x','-

_=

segments

where

the

average

return

photons

as >16,

but

remark

that

4.2

Model

There

and

of the

satellite

variation

of histogram

signature

remains

resemble

results

ta_uce during corrections. change model

cannot

those

numbers

we have

could

asigned

be several

the

times

numbers

of

as large.

shape

and

single-shot

Starlette

to detect

precision

returns, 1). For and

or ERS-1

a systematic

the

change

the distribution the high return histograms

(Figure

variation

with

of signal

of residuals rates little

qualitatively

distance of the of the

50 ps, to represent a random number

and

1). of satellite

mean

reflection

the passes, because the method of reducing the observations absorbs However we can use models to predict both the increase of precision

by convolution

FWHM we use

100%,

in the distributions,

from

be used

of mean reflection the time-distribution

number,

true

near

At low return rates equivalent to single photon to the 'standard' Ajisa.i distribution (Figure

quantitatively These

the

were

Discussion.

is a clear

strength. is similar

rates

dis-

any such and this

as a function of numbers of photons in each return. We returning photons, from which we may sample a variable

Ajisai

spread

distributions

the laser. To model the effect generator to pick one 'photon'

with

a Gaussian

distribution

of n photons reaching the from our time-distribution

of

detector, of pho-

tons, then reco1"d its time-location within the distribution, and repeat the process n times. We then sort this sequence of n relative event times into chronological order of arrival at the detector. time dom event the

We model

the

20% efficiency

of the

detector

by stepping

through

the

n events

in

order, at each step generating an integer ran(lore number in the range 1-5. If the rannumber is 1, the event is accepted (detected). If the random number is not 1, the next is 'tested'. selection

In this way

of a single

wc generate

photon

from

a large

a series

number

of returns

The mean and standard deviation of these event in cm. The standard deviation values are added

of event containing

times

each

an average

resulting

from

of n photons.

times arc computed and converted to range quadratically to the estimated system jitter

(0.8 cm) to fully model the observations. The results of simulations of range precision and biases from Ajisai for values of n between 1 and 50 arc shown in Figures 6a and 6b, where the results hnye bccn joined by continuous lines. The 30-second aver_Lge observed values of precision, where they Ajisai passes, are shown predicted distance

range bias curve from the satellite

reaching the detector from the finite pulse strengths change 4.3

at the

in Figure 6b expresses the expected change centre of mass as a result of increasing the

in each laser return. Most length of the laser (FWHM

single

is predicted

Lageos

photon

with

level

increasing

Centre-of-mass

We can

use

Lageos,

in the

Figure number

can be reliably cstimated (see section 4.1) from our experimental as dots on the graph and agree well with those predicted. The

the

above context

of the change of the 1.3 cm for a variation

of about

of photons

40 photons

beyond

that

a contribution between signal

per

return.

Little

point.

correction.

techniques of worldwide

6 shows the results of photons reaching

of the bias, which contains 50ps), is seen to take effect

up to an average numbers

of mean reflection number of photons

to estim_Lte SLR

the

systems

magnitude working

of a systematic at different

range-bias

return-signal

of a computation of the rmlgc bias as a function the detector, for 2 modelled laser pulse-lengths. The

effective reflection distaalce from of return level fl'om single-photons

2-11

the

satellite centre to the 40 photon

for levels.

of avcragc magnitude

of mass is about level. This result

zl_'-'l--*l

I I

I

_I

l

I

I

I

l

I

l

i

i

i

i

i

I

i

i

i

i

i

i

i

i

I

3



++ b

a)

Ajisai

i

_(a)

Prcdictcd

lind obscrvcd

bcrs of photons |

0

l

'

J

I

variation

of Ajisai

range

prccision

as a function

of avcragc

num-

rcccivcd. ]

]

0

|

l

J

,

]

I0

, ....J

,

,

t

20

,

,

i

J

30

I .J

,

J__.a__

40

50

4

-

B v

(b)

Ajisni

2

o

o (.9

_eeith

)

munbcrs

of photons

for Ajis+d.

0

-1

, 0

',

,

,

J

"-+.

I

f

,

,

I0

I

,

j

,

,___]

,

,,

j

30

20

1

,

,

,

L+

40

50

2

I

B v

laser

50ps (c)

aageos

£

o

o ¢.) ,NmnV._. COUNTER COMPUIER

DELAY GENERATOR Figure 3 Portable Standard Hardware (SLR Configuration)

2.1 Frequency Source A Hewlett Packard Cesium beam standard with a high-performance option tube and an Austron disciplined oscillator provide the frequency source for the PS. The disciplined oscillator improves the short-term frequency stability an order of magnitude when compared with use of the Cesium standard alone. 2.2 Time Interval Measurement Several HP5370 (A & B) time interval counters and two Stanford SR620 time Interval counters are available to take range measurements. The time interval counters are arranged to take multiple measurements of the same event to Improve the data quality. These counters can also be configured to measure differential time (figure 4) between two return events or a combination of events depending on the application. For these first tests 4 HP5370 counters were used.

4-3

I,ll]ql

J

_ULEH(_II4

l

!

LMI_

OPTICAL DEIECTOR #1

WAVELENGTH SEPARATOR WAt_JE_alH

#1

I_ACNSTANT

"nON

I

I

CRIMINA'TOR ]_ I

I

"nuE IL. I I COUN R il I

VARIABLE

IN'IERVAL

OPTICAL DE'lECTOR #2

L

DIGITIZER

It°"

IN'IERVALcouN11ER I

PHOTODIODE IN'IERVAL ,=I> IFRACTION I [DISCRIMINATORJ PRIMARY START PULSE

ICONSTANT

I

COUNIER

IFRACTION I IDISCRIMINATOR j IDIFFERENTIAL IRANGE AND PULSE 'AMPUTUDE IMEASUREMENT

I

TIME TIME INTERVAL COUNTER

r EPOCH TIME-TAG -dl>

I I

INTERVAL COUNTER

]M]EASA_ ]lL,M_,ze

Figure 4 Differential

Range Configuration

(Two Color etc)

2.3 Signal Processing A Tennelec constant fraction discriminator (TC-454) is used to process the MCP-PMT detector output. The discriminator is calibrated to remove systematic amplitude-dependent errors (time-walk). For these first tests the MOBLAS-7 discriminator was used and no amplitude measurement was made using the PS. 2.4 Optical Detection The PS features a detector package which can accommodate several detectors. These may be directly coupled to the optics, or indirectly using fiber-optics. During two-color ranging experiments two MCP-PMT's may be used to separate the wavelenths for differential time measurement. Other types of detectors (PMT, APD, Photodiode) can be used depending on the application, or simultaneously for detection Intercomparison. During the first series of passes the MOBLAS-7 MCP-PMT detector was used. 2.5 Amplitude/Waveform Measurement The output of the MCP-PMT may be digitized using a Tektronix 7912HB (1 GHz bandwidth) digitizer from which amplitude information is extracted. This information can be applied to the range measurements to correct for discriminator time-walk. The digitizer can also be configured to measure the differential time between two pulses while the entire waveform record of each shot is stored on disk. During the TOPEX measurements, a 4.5 GHz bandwidth digitizer coupled to a 60 GHz bandwidth photodiode was substituted for the normal digitizer to map the optical response of the TOPEX array segments.

4-4

2.6 Environmental Instrumentation The thermal environment of the PS instrumentation is monitored and stabilized using microprocessor controlled fans. Thermal sensors are arranged to measure the temperature of the air discharged by the instruments. Thermal stability can be maintained better than 1 degree Centigrade depending on the airconditioning system available. Thermal stability is essential to maintain high data quality. 2.7 Calibration Optical calibration provides a means to model Time Interval Meter non-linearities for limited dynamic range measurements such as satellite laboratory characterization and two-color differential time interval measurement. The optical calibration mechanism (figure 5) consists of a diode laser, optics and a computer-controlled microposltioner with ~ 2 _m precision. A pulse emitted by the diode laser is split into two pulses, one pulse travels a fixed distance while the second pulse traverses a known but variable range. The two pulses are each detected by a common detector and the signal processed using a common discrlmlnator. The calibration mechanism may be used two ways. For discriminator (amplitude-dependent error) calibration the separation between the pulses is held constant, while the amplitude of one of the pulses isvaried using a neutral density filter wheel. A real-time histogram of the varied pulse height is displayed on the computer CRT assuring that a good distribution of data is collected to determine the CFD time-walk curve. For range-dependent error, the amplitude of each pulse is maintained at a constant value, while the pulse-pair separation iSvaried. A number of data values are collected at each range Interval. The variation in separation step size and number of points per position is selectable and Is dependent on the desired accuracy. For LAGEOS-2 measurements 0.25 mm instrumental accuracy was desired so an appropriate step size and data volume (500 measurements/position) were used. DIODE LASER

LASER PULSE

CUBE CORNER

DIODE LASER PULSE PAIR

,2)-f (Xl)

O-R(X1)+

Ob-T(X1,)_2)

(2)

Where -_ T(X2)-T(X1)

5T(X1,X2)

2 (3)

°o'-r 0,_,>,2) = i=1

AT (X i) nph(Xi)

• T/qE(), i )

Absolute

range

sweep

to the satellite

RO R(X)

= =

f(X)

=

Dispersion

TO

=

Absolute

time

T(X)

=

Increase

in time

Measured

" A

range

at wavelength

),

factor

at wavelength

of flight of flight

s(x)

=

r/ph

=

_QE

=

Pulse profile factor Number of photons Photocathode quantum

A

=

Microchannel

5T

=

(7

=

Differential time of flight Standard deviation

AT(X)

=

Pulse

width

5-39

through

),

coupling

at X

efficiency factor

atmosphere

DATA ANALYSIS

SMOOTHING

TECHNIQUES

(3 POINT) ALGORITHM

( a(m-l)+ b(m-1)+ ¢(m-1) ) (m-=)

(4)

( a + b + c )

I b(m-1)+

(m) =

c(m-1)l

(s)

(b+c)

I a(m-l)

(m) =

.

b(m-1)

) (6)

(a+b)

m

Mean intensity "m" iterations

a,b,c

Weighting

of the pixel at location

coefficients

DETECTION:

Peak Half Maximum Centroid

(Mean Value)

5-40

p after

ORIGINAL BLACK

AND

WHITE

PAGE PHOTOGRAPH

LLI I--" Cf) >-U3 LLI > LL] (._) LLI (3£ rY _J CO CD LL] U3 CIJ < (3£ L_J r_ W W

n-uj 13_ n," (__ WO')

_w ,_.J

oN rY w I--m_

0 Lh

O_ w

w_ (.)Z w __1 0 i_. I--

o_ 5-42

5-43

l,

o rY i__jo _o,J

D_

5-44

|

m_

C

+++ ++ I.,x+Oo_.

+g

o

i

i'

i

o

:

0

:m--+++,+,_,_+,_ ('13XId/Sd)

a33dCj

d3"-JMS

("FIXId/Sd) a:13clScl33MS

5-45

SUMMARY Single photoelectron sensitive streak camera SLR receive system has been configured with a resolution approaching 250 fs/pixel for high accuracy atmospheric dispersion measurements. It has been verified that the streak camera timing characteristics are profoundly impacted by the sweep voltage. The sweep voltage is intrinsically nonlinear in the leading and trailing edges and must be corrected using optical calibration to achieve picosecond timing accuracy, centroid and half maxima (mean) analyses were found to be superior to peak detection. Real-time calibration has been found to be very effective in monitoring the timing performance and would be very advantageous for critical timing measurements.

5-46

N94-15574 THE

FIRST

SATELLITE

RECORDED

ON

THE

K.Hamal, Faculty

of Nuclear

Phone/fax

+42

Science Brehova

2 848840,

LASER

ECHOES

STREAK

CAMERA

I.Proch_ka

and Physical Engineering, Czech 7,115 19 Prague l,Czechoslovakia,

telex

121254

fjfi c, E-mail

G.Kirchner, Austrian Academy A-8042 Graz,

Technical

University

[email protected]

F.Koidl

of Sciences,Lustbuenel Austria, Phone +43

Observatory 316 472231

ABSTRACT The application is described. The integrated ranging ed.

of the streak camera with the circular sweep for the satellite laser ranging Modular Streak Camera system employing the circular sweep option was

into the conventional has been

Satellite

performed.

The

Laser

System.

first satellite

The experimental

laser

echo

streak

satellite

tracking

records

are present-

camera

and

GENERAL The laser

ranging

is an attractive

and

is gradually requested accuracy.

at 532

unit

between

reaching

millimeter the

wavelength

The

laser

level

and the long ground

ranging

precision

at present.

The

baselines

and

accuracy

for

nineties,

goal

that

is the

millimeter goals

to predict

ranging

propagation accuracy.

main

However,

to

The on

two

obtain must

the

camera

detectors 5--47

the

time interval

centimeter

to

meters

and

time interval error

and hence

models

with centimeter and

atmospheric valuable

electronic

from

atmospheric delay

wavelength

be measured

as a return

_.

to the systematic

optical

the existing

An

improvement

existing

millimeter of 20-200

photomultiplier

to determine

and discriminators,

of a streak

of the existing

detection

accuracy

the atmospheric

delay

is used

contributors

atmosphere.

simultaneously

of the implementation

: the replacement

The

time difference,

the two wavelength

signal

systems

used detectors

itself.

plate

photodiode

of 30 picoseconds.

picoseconds

and echo

microchannel

or an avalanche

resolution

of 20-35

for the current

accuracy

the fast

discriminator

interval

transmittion

factor

data permit

improvement, first

fi time

of the target final

It is expected,

The

the Moon

centimeter

as a transmitter, electric

level are the currently

meteorological

The

satellites,

technique.

one

a resolution

pulse

limiting

the contribution

to

nanometers

having

the laser

The

limiting

developing

it reaches

with an appropriate

detector

interval

Earth

by the data users, is the satellite laser ranging to artificial satellites with The existing pulsed laser ranging systems are using short pulse lasers

along

signal

rapidly

increasing,

picoseconds tube

of the artificial

data

determining

models for

into a laser

on

the

two

will be improved atmospheric

with picosecond

and discriminators

based accuracy.

ranging

model

accuracy system

by the streak

2 . : camera

with and appropriate readout and data processing system. This will increase the time interval resolution of the radar detection chain several times, the echo pulses distortion by the target geometry may be monitored and compensated. The second : the streak camera will be used for two wavelength laser ranging echo signals detection, the two wavelength range delay may be determined with the picosecond accuracy. The third : along with the detector, the streak camera may be integrated into the time interval unit. The picosecond time resolution and a high overall timing accuracy is expected I. THE STREAI_

CAMERA

CONSTRUCTION

Designing the streak camera for satellite laser ranging purpose several key problems to be solved: energy budget of the ranging chain, closely connected to the photocathode and sweep used, camera configuration, the type of deflection, input optics configuration, readout system and the output data processing. Completing the series of indoor and ground target ranging experiments, we have chosen the $25 photocathode, the circular sweep camera setup and a two dimensional TV readout system with a full frame image processing. "l'be linear sweep tube setup reaches higher limiting temporal resolution. However, the necessity of a trigger signal appearing 1030 nanoseconds prior to the photons to be detected with a subnanosecond jitter and the angular/temporal relation may cause serious difficulties when ranging to satellites. Using the circular sweep tube setup, the trigger signal requirements are moderate : the trigger signal must appear 10-20 microseconds before the optical signal, the jitter of several microseconds is acceptable. The temporal/angular dependence may be monitored, software modelled and compensated. The PV-006 streak tube having the $25 photocathode and the microchannel plate image intensifier fiber optically coupled together have been used. The streak tube is equipped with two pairs of deflecting plates acting in the mutually perpendicular direction. Applying the RF signal phase shifted to both deflection plate pairs, the circular sweep may be obtained. The 320 MHz signal, 13 Watts of pulsed RF power is used. The maximal RF power applicable on the deflection system is limited by the internal ionization inside the tube. The diameter of the circle was about 6 mm resulting in 155 psec/mm sweep speed. Due to the imperfect impedance matching of the RF driving circuit to the tube deflection system, the deflection is not perfectly circular. Nevertheless, the complex streak image processing and calibration package is able to compensate for these effects. The temporal resolution of 30-35 psec and the range difference jitter 6 psec have been achieved. The application of the streak camera for ranging purposes is accomplished by the timing and gating circuitry. The deflection signal is ON for only few microseconds before laser transmittion and again few microseconds before expected arrival of the satellite echo. All the remaining time the camera is working in the static image mode and may be used for guiding / alignment purposes, as well. The 320MHz deflection signal is produced by multiplication of the 5 MHz sine wave output of the station Cs beam frequency standard, which is simultaneously acting as an master oscillator for all the station timing electronics. This way, the precise phase synchronism of the radar electronics and the camera circular deflection is maintained. In fact, the camera may be used as a vernier to the time interval unit, rough time interval is derived from the integer number of periods from Start to Stop event, the fraction of the period may be determined from the phase difference of the Start and Stop even_. The microchannel plate intensifier gain is controlled in three steps. During the laser have - the speed - the - the

5-48

transmittion,

the

atmospheric

is set

backscattered

photocathode promise

gain

is used.

between

minimum

light.

Most

the static

to avoid

Optionally,

of the time image

expected echo arrival time, achieve maximum detection To readout

to

the

the

mechanical

mode

plate

system of

by

a com-

contribution.

for 300

Intensified

the

the input

is set to 100,

noise

of 10,000

we use the Silicon

the

in front

gain

gain and the background

image

of

shutter

the microchannel

the gain is set to maximum sensitivity.

the tube screen

blinding

At the

microseconds Target

to

(SIT)

tube

made by Haimann. The output signal was fed to the Visionetics Frame Grabber card (512x512x8bits) inserted in the IBM PC. To solve the problem of a deflection nonsymetry, the software

package

3 allows

the full frame

image

processing

and

accomplishes

the software

modelling of the image distortion, the sweep nonlinearity, the gain nonuniformity etc. On the end of the image processing procedure the temporal curve consisting of 1000 channels, 8 bits each

is generated.

satellite

image

Once

to install

SATELLITE

Lustbuehel laser,

design

streak

RANGING

Observatory,

passively

mode

locked

been Austria.

NdYAG

The

in green.

containing

60

may

milliJoules

returned

photons

package

based

ranging

on a single

hard/software

photon

remained

+ 5 C what

system

using the

compensated.

with its low

mass

telescope.

Ranging

consists

harmonic

be transmitted.

the The

System

of a modified

generator, semitrain

the

Quantel

delivering

pulses

of 5 - 7 such pulses

telescope

is a Contraves

laser

ISIT camera attached to the telescope and dielectric mirror is directing 99% of the

the remaining

avalanche

unchanged.

photons

photodiode. Thanks

enter

All the

the standard

rest

to the photodiode

of the

in significant

dark

noise

detector

satellite

single photon

ranging capability has been preserved, although system timing and gating logic has been added. image intensifier and the readout have been

resulted

of

by a 10 cm diameter Galileo telescope, the 0.5 for both return photons collection and visual

photocathode

capability, the satellite routine return rate. The streak camera The streak tube, the temperature

ranging

replaced an original visual tracking. The

on the streak

along

Satellite

Alternatively

tracking type. The laser output is transmitted meter diameter Cassegrain optics is used The streak camera for faint objects

to the

with the second

35 picoseconds

guiding. dedicated

may be software

construction

part of the tracking

integrated

of 10 milliJoules, totally

for example

SYSTEM

has Graz,

is identified,

effecP

its rugged

on the moving

camera

circle

the temporal/angular

of the camera,

it directly

LASER

The

of the deflection

in the static mode,

The compact permitted

the center

laser

response

with

reduced

kept

at

the

reduction.

RESULTS The have

been

been km,

retroreflectors ranged.

obtained see Figure

The

January

equipped first

satellite

18,1991

1. In this stage,

satellites laser

at 21:37

at the distances

ranging UT from

the semitrain

echoes

of

the AJISAI

of pulses

1,000

recorded

satellite

separated

is near

the

center

of the pattern

displayed

in the

static

kilometers camera

have

at the distance

2000

by 8.557

been transmitted. On the image the time interval of 40 milliseconds transmittion to echo detection is integrated. The transmitted laser image

to 2,300

by a streak

nanoseconds

has

covering the laser beam backscattered

pulse light

image.

moment

the

was not illuminated by the Sun. On Figure satellite echo obtained transmitting single

energy. camera

The temporal profile of the firs part of the trace is included. Please, note the streak receiver system saturation by the strong echo signal coming from the first rows of the 5-49

2 there is a streak picosecond pulse

At this

satellite AJISAI

camera record of the containing 10 mJ of

corner

reflectors.

retroreflectors.

The

the time interval echoes

coming

nonlinearity

about one nanosecond

depending

backscattered

light

satellite

echo

record

satellite

is a sphere

detector

various

illuminated

comparison

retroreflectors

compensated

has been covered

3. The

satellite.

may be seen.

response

from

be clearly

time

profile

temporal

corner

sweep

speed

spot is an static image

of

The atmospheric The

of about

STARLE-TFE

transmitted.

The

25 centimeters

is significantly

of the echo

resolution

reflector

The separate

The

has been

spread

with

orientation.

shutter.

pulse

covered

be spread over

of the sweep.

mechanical

signal

The temporal

a single

central

of the diameter

echo

The system

The

the center the

meters should

distinguished.

picosecond

with retroreflectors

kilometers.

to the AJISAI

out by Single

two pulse

on the satellite

may

thus marking

blocked

is on Figure

exceeding

the returned

on this display.

by the Sun,

1,000

saturation

not saturated

of diameter

simulations,

from

about

is a sphere

to numerical

has not been

the satellite

distance

satellite

According

is included,

as well.

may be demonstrated

resulting

in a pulse

on the streak

camera

at a

smaller

width

in The

on the of 25-30

picoseconds. CONCLUSION The first satellite for

the

circular camera ranging

first

time.

sweep

laser The

for satellite

has been

echoes laser

confirmed

employing

have been

applicability

recorded

of a Modular

Streak

has been

demonstrated.

ranging

to be satisfactory

available

laser

ranging

Camera

to our knowledge

system

employing

The sensitivity

for the purpose

: for standard

the

of the streak

geodetic

satellite

technology.

REFERENCES

1. J.J.Degnan, "Satellite Trans.on Geoscience and 2. J.Abshire, 3.

Applied

P.Valach,"

Laser Remote

Optics,

Streak

Limitations", Antibes,

Sept. 1986,

5. K.Hamal, for Ranging", Photonies,

proc.

edited

presented Cambridge,

Frame

Ranging

"Streak

of the Sixth

I.Prochazka,

Full

on Laser

K.Hamal,

Current Status and Future Vol. GE-23,pp.398-413,July

Vol. 19, No 20,pp.3436-3440,

Camera

International Workshop C.Veilet,CERGA,France 4. I.Prochazka,

Ranging: Sensing,

Camera

Image

International

Radar

Workshop

V.Lozovoi,

at the 19th International UK,Sept.

proc.of

17-21,1991,

5-50

the

Oct.1989,

Receiver,

on Laser

by J.Gaignebet,Cerga,France, M.Schelev,

IEEE

1980 Processing",

Instrumentation,Matera, as a Laser

Prospects', 1985

Seventh edited

its Performance

Ranging

by and

Instrumentation,

1987 V.Postovalov,

Congress published

" Modular

on High

in SPIE

Speed

1358-55

Streak

Camera

Photography

and

SATELLITE

LASER

FIRST STREAK

AJISAI

satellite,

RANGING

RECORDS

diameter

2.,

meters

range 1 500 km January 19, 1991 , UT 21:37 0

v

C= O

% an

sweep

dk

period 3.125

nsec, displayed

part 1 nsec

Fig.

5-51

1

...... r

"_

q



fiJI_lI

satellite

" q, lb

• •

IN

Q

picosecondlaser pulse response on streak

i88

?5

58

25

8

t

1

!

188 24B 68 128 fine [ 5 psec/pixel : 1,5 nsec/screen ] 5-52

Fig.

2

SATELLITE

L,_ SER RANGING

FIRST STREAK

STARLETTE

satellite,

RECORDS

diameter

0.3

meter

range 1 000 km January 24, 1991 , UT 21:59

100 psec 15 mm range

t -_.J

n --

•-41.

r_ _

_

.,11...,113__

.,S'Cn r'__

,.r-,..,n_r_ .IL,qn_

echo signal temporal

profile

8 psec / pixel

Fig. 5-53

3

Calibration Techniques/Targets

N94-15575 Experience

and

Results

of

the

1991

P. INSTITUT

MTLRS_I

Campaign

SPERBER

FUR ANGEWANDTE

FUNDAMENTALSTATION DER

USSR

FORSCHUNGSGRUPPE D-8493

GEODASIE

WETTZELL SATELLITENGEOD,_SIE

K()TZTING

H. HAUCK INSTITUT

FUR ANGEANDTE

RICHARD-STRAUSS D-6000

GEODASIE

ALLEE

II

FRANKFURT/MAIN

70

Abstract. In the fall 1991 the Modular, Transportable Laser Ranging System MTLRS#I was operating in the USSR for collocation of the SLR systems in Riga, Simeiz and Kazivelli. In this paper we will summarize the results of the collocation experiments and we will show our (positive and negative) experiences,

1.

we got during this campaign

in the USSR).

Introduction

The

year

upgrades

1991

Germany)

and

MEDLAS

nor

After

MTLRS#2

MTLRS#I)

upgrade

was

and

by

operated

the

OSG

to

(about

in the first

Due

half

out

SLR

system,

two

fixed

SLR

scheduled

by the IfAG,

neither

(P. Sperber

sites in

the

a WEGENER-

in 1991.

of 1991

at two

to

(operated

Netherlands) carried

fixed

the

3 km

was

systems. MTLRS#1

Kootwijk,

maesurements

the

to collocate

ranging

Systems

campaign

to make close

laser

Ranging

of MTLRS#I

choosen

Kasivelli

mobile

Laser Project

in August

we

pad

for the

(operated Dynamics

Wettzell

Riga/Latvia

year

Transportable

a Crustal

from

MTLRS#1

2.

a special

the successful

departed In

was

of the Modular

in the

et al.)

Simeiz/Ucrainea

station

the system

USSR. the

in Simeiz

(300

place

for

m distance

to

distance).

Results An

overview

about

system

arrived

the

number

and

quality

of the

MTLRS#1

passes

is shown

in Fig.

1 and

Fig.

2. The new

parts

first

data

In

of the were

week

army

34

During

the

All

this

the

quantity

completely After sytem

faced

against

rest

with

(number

Because during

a problem,

our

we

Due

observations

of measurements are

32.

installed

gorbatchov.

to stop

problems

week

were

the the

crew

was

upgrade,

it

took

not

yet

before:

In

moscow

familiar

some

with

days,

some

before

the

successfully.

we were

we had

during

which

collected

putsched

situation

in Riga

system,

to again

in Riga

showing

in an

of passes

per

for nearly faced

poor

nor

had

unclear

we were

unusual

week)

never

the

the

and one with

dangereous

of the and

soviet

military

week. extremely

performance quality

parts

political

of

(number

bad

weather

MTLRS#

1 in

conditions.

Riga.

of normalpoints

per

Neither pass)

was

moved

the

satisfying. 43

Lageos

passes

we stopped

the

observations

in Riga

on

October

3 rd

and

to Simeiz. Typeset

6-1

by .AA,t-q-TEX

Here the system was able to show its habitual performance. After 53 Lageos passes with more than eleven normalpoints per pass in average the campaign was finished on November 23 rd. The computation of the collocation was performed by the computing Frankfurt/Main. The results are summarized in Fig. 3 - Fig. 5.

center

of the

IfAG

in

Fig. 3 shows the results of the Riga collocation. We got 16 simuitaneos Lageos passes with a R.M.S. of 20 cm at the Riga system and 1 cm or 2 cm at MTLBS#I depending whether a PMT or a single photon avalanche diode (I. Prochaska et al., P. Sperber et al.) was used as receiver. Out of this 16 passes six passes with good residual overlap (for example: Fig. 4) were selected to calculate the range and epoch bias of the Riga SLR system. There is a small negative tendency in both biases, but compared to the error values and the R.M.S. of the Riga system we can't find a significant bias. Fig. 5 summarizes the results of the collocation in Simeiz. Because of problems, the fixed stations only were able to observe few passes during the collocation, but all of the passes were simultaneous with MTLRS# 1. The data of the fixed stations was not possible. 3.

are not yet delivered

to the network,

so until now a computation

Experience

In this chapter we will summarize our experience of a dedicated geodetic system in the USSR. 2.2.

Operation

during the on-site

operartion

and the transport

on stations

From the operational point of view there were no problems as long as the system was working near stations. All our requirements concerning electrical power, safety, infrastructure, etc. were fullfilled. In Riga also the supply of fuel and the hotel accomodation were satisfying. On all stations an independent communication facility like Inmarsat is necessary. The only local data channnel is a very unreliable telex line. In Simeiz,

the hotel,

food and fuelsituation



A hotel near the station km away.



For fuel a big spare



As there

Additionally

is far under western

tank is necessary,

are no restaurants

most of the people

is very inconvenient:

because

near the station, only speak

russian

standard,

an acceptable

fuel is not always facalities

Transport

During compared

between

than 40

available.

are needed

to prepare

own food

language.

In spite of this problems, the operation of a sir system near fixed stations some preparations are made to facilitate the life of the crew members. 2.2.

one is more

is always

possible,

if

stations

the transport of a system from one station to the operation on stations. channels

to foreign

are not available

in short

becomes

very bad



Communication



Hotels only exist in big cities. Sleeping facilities for all persons in cars (caravans) are strongly recommended because most of the hotels are full without reservation four weeks in advance.

6-2

countries

to the other the situation

time



Food is only available



Security



Fuel is not always availble,



Car repair is possible,

The poeple

is a big problem,

only speak

to help if there 3.

in hotels,

so there are facilities

day and night guards therefor

necessary

are absolutely

big fuel tanks are needed

but takes a lot of time and you should

russian

(or sometimes

german),

to cook in the cars. necessary. in the ears. have all spare

but are very friendly

parts with you.

and will always try

are problems.

Summary In the

second

half of 1991 the

collocate the fixed laser ranging The results of the collocation collocation in To operate regions of the from the local

MTLRS#I

was operating

stations on this places. shows no significant problem

successfully at the Riga

Simeiz is not yet computed. and transport a dedicated geodetic system like MTLRS#I USSR big efforts and preparations are necessary to become infrastructure.

6-3

in Riga fixed

and

Simeiz

slr station.

to The

in in the underdeveloped as independent as possible

REFERENCES

P. Sperber, L. Amberg, Upgrscle, 199t. I. Prochazka,

K.

Instrument&tion P. Sperber,

L. Beyer,

Kamal,

B.

S. 219,

U. Hessels,

R.

G. Blenski,

Sopko,

Proceedings

PAotodiode Motz

U. Hessels,

Bazed

in this

6-4

Detector

Proceedings,

11. Motz,

of

the

seventh

PackaEe The

in this Proceedings,

new

]or

international CentimeLer

MTLRS#1

Remlhj

Workshop Satellite

Receiver

ol the MTLR5#1

on

Ranglug Pscka#e

Laser , 1989.

199t.

Ranging

NUMBER

OF TRACKED

PASSES

14 LIGHT 12

!_

i-i

NIGHT

....n-m

_

ALL

II II

II

.... mm--

-

II

_

]

4

_1

10 8 6

J

1

30

32

34

36 WEEK

Fig.

1

Number

NORMALPOINTS

15

Lageos

40

42

44

46

48

IN 1991 Passes

in

Riga

and

Simeiz

PER PASS

LIGHT

I

of

38

E_J NIGHT

ii

ALL ]

36

38

10

oi i 0

I

i

1

t

30

_

_

32

34

WEEK Fig.

2

Number in Riga

-7-

40

42

44

46

IN 1991

of Normalpoints and Simeiz 6-5

per

Lageos

pass

48

Results

Riga

MTLRS#1

(1884)

15868

27166

Returns

R.M.S.

1 cm

20 cm Aug.

15 - Oct.

16 Simultaneous

Range in cm

Bias

7560-1884

Fig.

Returns (SPAD)

(43

Passes)

- 2 cm

(PMT)

10 Passes

Epoch Bias in microsec

7560-1884

-24 -30 -58 -23 -4 +66

-6.1 -2.5 -5.1 -16.3 -0.4 -8.4

-6.5

(7560)

cm +- 5 cm

3

Results

-12 microsec

of

the

6-6

collocation

in

Riga

+- 30

RANGE 030 .'"'1

....

I ....

RESIDUALS

I""1""1'"'1''"I'"'1

....

I ....

1'"'1

....

I ....

I ....

I'"'1

....

t'"'1

....

t ....

I .... -

r.

025 o

020

!!;,t T= tz rv bJ _J Lg Z H

z >

Ra

MILLIMETER SATELLITE

Figure 3: variables

Diagram used in

6 SATELLITE

of a spherical the analysis.

IMPULSE

RESPONSE

AND

geodetic

TARGET

satellite

the

SPECKLE

An individual retroreflector responds as a point does not spread the laser pulse in time. However, array of retroreflectors, the laser pulse arrives tion center" of each retroreflector at a slightly leading to a broadening of the received location of the "reflection center" for corner reflector is given by [Fitzmaurice

defining

source and hence with a typical at the "reflecdifferent time

pulse [Degnan, an individual et al, 1977;

1985]. The solid cube Arnold,

1978] si AR(Oi.c)=

nL

1-

_,c = nLcosOre/

(6.1)

where AR(0i,c) is measured from the center of the front face of the cube corner to the reflection point, L is the vertex to front face dimension, n is the refractive index of the corner cube material,

0_.c is

the

sponding

refraction

angle angle.

of

incidence,

From 7 -44

Figure

and 3,

it

0re/ is can

be

the

corre-

seen

that

a

cube at relative

an angle to the

station

(0inc =

0i,c to surface

0),

the of

given

incident wave the satellite

Rs

is

the

a time delay, to the ranging

by

At(0inc)=_{R-[Rs-_R(Oinc)]cosOinc}= C s

where

produces closest

l--c0s0in

satellite

radius.

It

should

e

l--EC0S0re

also

be

I

noted

(6.2)

that

the differential delay between target reflection points also introduces a random phase delay between individual reflectors. Thus, if the temporal profiles from multiple cubes overlap at the range receiver, the electric fields will interfere with each other in a random way from shot to shot resulting in target "speckle". On average, however, the return waveform from the satellite should behave as if each of the retroreflectors is an incoherent source. This was an implicit assumption in our derivation of target optical cross-section in Section 5. In the same spirit, the time-averaged satellite impulse response can be estimated by summing the weighted (incoherent) returns from each of the retroreflectors. Using the simple model for a spherical satellite introduced in Section 5, the impulse response can be shown to be

where

the

dO,._ sin

O,..q2 (O,._)6[t

geometric

weighting

(5.3),A/(0_,c)

is

given

by

5[t-_/(0Lnc) ] represents incident function

on is

factor

is

given

by

and

the

delta

function

(6.2),

an

the satellite. nonzero only

(6 3)

- At(O,._)]

infinitely

short

From (6.2), we when the condition

laser see

(2.4)

or

pulse

that

waveform

the

delta

I-T

cosO( , ,n)

(6.4)

= 1 -_+,

1 -6 holds

where

we

have

[co,oc. c,°,]2

defined

the

new

variables

ct 6 < T < T max

T --

2Rs

nL 6 -- R s

1 T max

The variable • is roundtrip transit center cube

and to

the

back,

= I-

C0S0ma

x

i-6

+

COS

a normalized time, expressed time from the surface of the and

satellite

e is

the

radius.

ratio

of

The

minimum

7-45

max

(6.5)

I -_

the

in units satellite

optical and

of the to the

depth

maximum

of

values

the of

are

determined

(6.4) the

and

by

solving

baseline,

From

setting for

is

(5.3)

and

expressed

as

_.

given

equal

The

by

(6.3), a

8

total

of

the

lite

diameters

defined

by

(6 _ 0),

impulse

0('_, 1

is

8ma × respectively

duration,

variables

[

l('_'6'n'Oma×)=°cc-2sinO(_'6'n)L 8(T.6.n)

and

pulse

satellite

N

where

zero

in

measured

at

At=tmax-6.

the

function

to

_,

e,

n,

can

and

now

@max,

be

i.e.

6,n)72

0--m_x

(6.4).

(6.4)

response

In

reduces

(6.6)

J the

to

limit

the

of

simple

large

satel-

form

l

O(_,O,n)=cos-l(1-_) and

(6.7)

(6.6)

becomes

2'(T,0,/'_,0max)

= 0ceNt/2(1

The quantity'e (6.4) can be point,

-_)[

1 -

is typically easily solved

cos-'(1-,)] Omax

2

small and, by iteration

(6.8)

for nonzero values using (6.7) as a

of e, starting

i.e. I-T (6.9)

[COS0(T'6'/_)]i÷I

= ]-6_

until

it

As

illustration,

an

response silica), value use

converges.

of L= 6 =

a

l-nzl--+[[c°se(x'E'n)]J]2--n

value

let

the LAGEOS 1.905 cm,

.093. 0ma

We x

=

us

(6.6)

satellite. and Rs =

recall .75

use

rad.

29.8

that,

to

Substituting cm into

for

Now,

estimate

solid

using

n = (6.5b),

cube

(6.6)

the

and

impulse 1.455 (fused we obtain a

corners, (6.9),

we we

can obtain

the plot of the LAGEOS impulse response shown in Figure 4(a). The profile shows the characteristic fast rise and long tail of the LAGEOS response. Furthermore, if we compute a center-of-mass correction from the centroid of this impulse response profile, we obtain a value of 250.2 mm which is in excellent agreement with the accepted 1977].

LAGEOS

value

of

249

±

1.7

mm

[Fitzmaurice

et

al,

The impulse response in the large satellite limit, given by (6.8), is shown in figure 4(b) for the case of solid quartz and hollow cubes. Note that both the temporal width of the reflected pulse and the target optical crossection is smaller for the hollow cubes than for the solid cubes because of their smaller field of view.

7-46

1.0

I

08

N_

I

I

I

I

I

_

06

0.4 0.2

L_

0

J 100

0

I

200

300

400

500

_d_-600

700

t (PICOSECONDS)

0.06

uJ o9 z O O..

o9

i

I

i

I

i

_OUAR_cuBE_ 0.04 0.03

W

nr" w o9 ....I

0.02 0.01

O..

--

0

i

I_l

0

I

0.1

0.2 Z=_

Figure puted large 7

4: (a) Impulse response by our simple analytical satellite limit for both

FEASIBILITY

OF

MILLIMETER

2R

of the model. hollow

ACCURACY

0.3

LAGEOS satellite as com(b) Impulse response in the and solid quartz cubes. SATELLITES

We will now demonstrate that, in order to achieve high optical cross-section simultaneously with minimal pulse spreading with a spherical satellite, we must increase the satellite diameter and/or retroreflector density and simultaneously restrict the response to retroreflectors within a relatively small solid angle on the satellite surface about the station line o5 sight. 7-47

The total time duration of the sity points) can be determined

At = --('_max=C which,

in

-R'0

the

_

t2sin

limit

of

--+62 small

maximum

Eq.

(5.4)

for

n2

incidence

0%

inten-

1 angles,

(7 .1) reduces

to

the

satellite

optical

cross-

N

O _,,,

(7.3)

As mentioned earlier, the angular recessing the retroreflectors in (7.2) into (7.3) yields OccN

response can their holders.

be restricted Substituting

CAtmax

24 We

to

(7.2)

same limit, reduces to

o'- -_-

0 =

1

COS Omax

(0% e.

e(l+

[-

In the section 0"co

E)

reflected pulse from (6.5), i.

by

(7.4)

R,[1-"_(I+_)]

can

now 8

express

the

total

number

of

retroreflectors

as

2

(7.5)

N=4p_ where sents faces. 0_

_ is a "packing density" ( = .435 for LAGEOS) which reprethe fraction of total surface area occupied by the cube Substituting (7.5) into (7.4) yields our final result [3R_cAt

(7.6)

6Ri The

product

_RscAt

that, if we wish some factor (via we must increase radius get 8

product

in

(7.6)

quantifies

our

earlier

statement

to reduce the amount of pulse-spreading At by reduction of the retroreflector field-of-view), the retroreflector packing density -satellite

(_Rs)

by

the

same

factor

to

retain

a

similar

tar-

cross-section.

CONCLUSION

As mentioned previously, the principal technical challenge in designing a millimeter accuracy satellite to support two color observations at high altitudes is to provide high optical crosssection simultaneously with minimal pulse spreading. Increasing the satellite diameter provides: (i) a larger surface area for additional cube mounting thereby leading to higher crosssections; and (2) makes the satellite surface a better match for

7-48

the incoming Simultaneously by recessing fraction of phasefront For

planar phasefront of the laser beam as in Figure 3. restricting the retroreflector field of view (e.g. it in its holder) limits the target response to the the satellite surface which best matches the optical thereby

near

term

reducing

experiments,

the the

amount small

of

pulse

radius

of

spreading. STARLETTE

makes

it

an attractive target for testing and.evaluating two color systems or for testing atmospheric models. Furthermore, its low altitude (960 Km) and moderate target cross section results in relatively high received signal levels. AJISAI, also in a relatively low 1375 Km orbit, consists of small clusters of retroreflectors separated by large reflecting panels and has comparable signal strength to STARLETTE. Unfortunately, the satellite is quite large and simultaneous returns from several retro clusters results in a complicated satellite signature [Prochazka et al, 1991]. seconds

From FWHM

although, individual certain

Figure 4, LAGEOS spreading is in agreement with [Fitzmaurice

in

excess et al,

of 150 1977]

pico-

with sufficiently short laser pulses ( co

0

0 L.O 0

_D

I

0

o O0

0 0 4.

I::::1

I

0 L._ o

C_ O0 I---

.,:-I CO

LLI .._1

o

!%0

8-14

!

N94-15587 GEOMETRIC

ANALYSIS

B.

OF

CONKLIN,

S.

Bendix

SATELLITE

BUCEY,

Field

LASER

V.

HUSSON,

Engineering

J.

The the

analysis

method

of

presented use

with

of

has

a

potential

In centimeters

_he

and was

community substantially.

Most

shot

data

has

single

During

shot the

satellite

positions

term

for

used

the of

on

that

trilateration analyzing

lower

other

The

major

the

station

determined

from to Our

or

steps

that

starts the

must

fullrate as

written

data the to

data

tape.

simultaneous pass

of

for

4 each

After data

the

more

pass. file. of

points

solution,

has

These

Poorly

The the

polynomials

are

created

fitted

although

for

this

are

rarely

8-15

not

allow

been

removed occurs.

for

can

be

is

also

data. data

The

observations

second

for

from

preprocessing process selects

during

a polynomial

30

need

Software

and

means

to

biases

that

analysis. tape.

have

passes

does

several

continues

at

the range

simultaneous

meteorological

process

as

determination

precision

polynomials, and

way

Trilateration

release

of a

radiation

method

are

found

a

orbit

low

have

The

provides

information

and

data

It

include

the

There

been

that

advantage

separating

other

stations

a

of

improved.

drag,

process.

the

with

determine

such

the

requires

statistics

all

polynomials.

corrupt

or

timeframe

output

that

fullrate

ranging

much

primary

in

in

ratio

prepare

to

orbit.

in

and

stations.

LAGEOS

polynomial an

fact

field

- 15

improved data

were

satellite

difficulty

the

has

and

trilateration

trilateration

to

produce

of

used

determination

ranging

which a

as

signal-to-noise

occur

Once

the and

orbit

now

quality

are to

gravity

a monthly in

timeframe.

such

an low

laser

with

passes

the

method

more

such that

10

ranging

laser

data

The

data

monthly

about

laser

analysis

independent and

was

the

trilateration. is

data,

data

methods

data

disadvantages

height, of

susceptible

is

with

centimeter.

mathematical

models

ranging

extraction

four

geometric

influences

methods.

the

the

outside gravity

from

1970s

1

the

Trilateration

studies

centimeters

below

is The

demonstrate

international

2

1992

positions.

satellite

stations

precision

to

using

to

millimeters.

of

below

positions

and

simultaneous

ranging

technique

pressure,

the

1987

station

laser of

time

using

station

of

investigated

provide.

- 12

precision

LAGEOS

method

and

deformation 3

precision

the

laser

precision

a

regional

that

and

is from

simulated

distribution

Since

expanded

data data

satellites

between

the

global

poor.

has

single

1970s

the

is

these

for

of

rates

data

solutions

Center

laser

baseline

Etalon

Project

Flight

Analysis

coverage

American

Dynamics Space

simultaneous

selected

global

great

LAGEOS

stations

of

simultaneous

additional

DEGNAN

Goddard

trilateration.

DECKER

MD

901/Crustial

NASA

W.

DATA

Corporation

Seabrook,

Code

RANGING

is

the

fit

other

to

information

information, through

same

the

the

are entire

determined, intervals so Any

that time

from

the

they

don't

intervals

within A

the

selected

summary

produced

at

of

this

separated

into

Commonwealth Americas,

of Easter

To

station

velocities

to is

step.

is

the

the data. started.

position position.

with

and

edits.

A

epochs. 4

P

the

is

the

repeated

than

of

the

This

database

observations

in

for

a

solution

by

and

Figure

contains the

European

2.

point

linear

weighting

the

month.

term

solutions.

The

high

stations

of at

rates

are

of

each

of

rates

of

trilateration

this

method are

requirement

for

satellites. high time.

The altitude However,

lends

8-16

of

data

offer to

one

which

a

satellites

month of

is

shown

and is

a

computed

observations

which

satellites Etalon

a

in

during

Figures

to

doing

3-5.

long

stationposition

method

ranging

the

for

by-product

itself

area,

visibility

simultaneous

for

database.

baselines

of

and

of

illustrated

over

Etalon

x,y,z

a

the

number

simultaneous

and

The

combinations

initial

a broad

is

in

of

from

This

are

determined

sample

station

the

determined.

coverage

a

are

O-Cs process

both

A

are

iterations

are

1.

and

analysis

from in

have

is

4

still

stations

turn

great

larger

to

the

reached.

summary

plot,

the

these

number

Figure

also

both

satellite

saved

in

the

with

corrections

for

sample

generated

baseline

In

one

shown

baseline

method

their

is

These

lengths

the

A

sample

geographic

altitude

because

and

distribution

fit

the station velocities under development.

requires

month

stations.

respect

consecutive

and

inverted,

with

average

been

the

times

by

determined

solution

monthly

Some

The

is

of 12

then

O-Cs

iterative

have

baseline

all

(ie.

Again

between

by

multiplied of

This

O-Cs

The

the two

set

iterations

the

then

along of

elements

for

and

the

determined.

position

solution

Baseline

that

six

in

also

new

done

vector

coordinate

of

summary

file

number

is

O-Cs

computed.

a

the

height.

between

least-squares

then

and

of

station

the

is

A

the

generated

are

reiterated.

or

involved

is

O-Cs

O-Cs

percent,

stations

The

and rms

the

is

new

delta

longitude

only

the

the

each

derivatives

a matrix

and

for

P matrix

positions

P matrix

one

change

latitude,

period

the

until

less

and

computed

by

performed

partial

and

the

average editing

summary

This on

applied process

an

6-sigma

position

has

each

and

with

run

station

station

for

the

iterative from each

a

and

at

recomputed.

the

generated.

the offset

transpose of

station

are

multiplied

then of

position.

partials to

positions

total

P

matrix

per

LAGEOS

positions

computed

along

is to

of

position

using

position

the

determined

epoch

is

Africa,

for

station

are

edited

written

each

is

station

resulting

is

A

the

adjusted

applied

effect

procedure

stations).

where

is

each

is

corrected an values (O-Cs)

each

then

observations

to

what

for

are

matrix

offsets

This

rms

of

sensitivity small

has been computed

data

North

section

is

deleted.

also

tracking

stations

corrections

satellite

and

number

determining

for

average mean

each

position

is

Europe,

Apriori each

are

yield

section

satellite

are

for

other

dependent

O-Cs

stations

LAGEOS

is

The

compute

determined

The

the

total

applying and

are

stations

the

to

4 data

remaining

section

software.

Once all the data The observed minus

satelllte each

used

angle

than and

geometrically, the

The

that

satellite

of

the

States. Huahine.

and

by

are

fewer

stations

One

data

in

month.

so

criterion

the

read

have

Finally,

sections.

analyze

observation

that

available

Independent Island,

data

of

the

two

ranging

middle

passes

output

requires potential

number had

of a

low

global

ranging

stations during be

at May

used

priority any

and

for

data launch

3

simultaneous

about

65

period

from of

a

velocities. expected biases

the

the

include individual

on

the

we

will

these

4

and

In

that

accurately

and

the

8-17

and 2,

at

an

only

1

scheduled

to

obtain

for a

inclination

obtain

of

simultaneous

satellite.

geometric

Etalon

analysis

2,

spread

baselines

and

geometric

determination

sets.

the'campaign

addition,

SLR

can

analysis.

opportunity

determine to

satellites

stations

to

two

performed

data

simultaneous

Etalonl

than

was

LAGEOS

best

priority

expect

additional data

from

opportunity

a high

modifications the

next

an

more

priorities,

sets

set.

be

LAGEOS2,

well

geometric

simultaneously.

data will

Etalon

by

campaign

how and

data

offers

future,

days,

low

ranged

Etalon

determine

ranged

sets

rarely

research

ranging

LAGEOS, few

to

ranging

there

Future to

from

of

1992,

data

In data

1992

station

degrees,

trans-atlantic

of

of

are intense

Dynamics

October

large

they An

simultaneous

where

in

June

because

no set

so

time.

Crustal

Unfortunately, yielded

one

analysis of

range

over

software and

time

a

are

II

II

tl

II

II

II

II

II

@@

UUUUUUUU

8-18

Z

0._

x

MXXUxX

xI xu I

(.0 0 [.J

xxx

.,-I

'_

(]Z

xxxu

GO GO

mini

[J

Nx Wx Uw Xw XN Xw Xx x M

f

UXMW XN

I

IXXxzxxx

I

_z xz



-:r t._ CO

C_

O) _r q

"*

% *-'4

"_

o

8-19

• /\

x

Lt_

8

Lt_

Q Lt_

II

a--

"'Z r.1

X \ #

i

X X X

x >< 0

0

X X X 0

0

0

0

0

0

d

0

CWW)

8-20

3NIq_S_8

"(._

n_

"'Z

I • t

| II

B

0

.

° 0

0

0 UI

(NNI 8-21

_NI_SUg

8-22

8 th International Session:

Workshop

Mobile

on SLR Instrumentation

System

Upgrades

/ Developments

N94-15588

IMPROVEMENT M. KASSER,

OF SLR ACCURACY,

ESGT,

18 All6e

J. Rostand,

A POSSIBLE BP 77,

NEW

91 002

STEP

EVRY

France

Abstract

s .....

The SLR technology experienced I..... i,, 1970 les, leading now

Presently it apl_ears as the atmospheric

as useless propagation

proposed since many years considerable technological delayed

such

a large number to a millimetric

to increase these delay suffers

of technical improvements instrumental accuracy.

instrumental performances as long its actual imprecision. It has been

to work in multiwavelength difficulties of subptcosecond

mode. but up to now the timing have seriously

an approach.

Then a new possibility is proposed, using a device which is not optimized now for SLR but has already given good results in the lower troposphere for wind measurement: the association of a radar mad a sodar. While waiting for the 2-k methodology, this one could provide an atmospheric propagation delay at the millimetre level during a few years with only little technological investment.

I/ INTRODUCTION It has been pointed out since a long time that all space geodesy techniques have to deal with the same general problem, i. e. the crossing of the atmosphere reach either an artificial satellite, or the moon, or a star, or a quasar, etc... Nevertheless,

if it is the

same

atmosphere

known to be quite different for radiowave ones suffering more than the other ones is corrected classically by two-frequencies transit (because of the atmospheric

for every

techniques,

the

effect

to

is

and for optical methodologies, the first from the crossing of ionosphere (but this methods) and from the troposphere water vapor content which is quite

unpredictable and has a strong impact on propagation of radio waves optical ones). Considering these aspects, the general advantages of radio compared with optical ones in geodesy are their aU-weather capabilities, drawbacks are linked with their poor tropospheric correction quality.

and not on techniques and their

As long as the instrumental accuracies were at a few centimeters level, this meteorological aspects were of secondary concern. This is no longer the case since many years with VLBI, who had to use costly water vapor radiometers to upgrade the tropospheric correction to an excellent level (and since this period, VLBI has got the best positioning precision among all space geodesy techniques). Since a few years

too

it is no

longer

the

case

with

satellite

8-23

laser

ranging,

which

is able

of an

internal instrumental positioning accuracy only problems SLR stations),

precision of a few millimeters, to be compared with the final ! of 2-3 cm. Of course the tropospheric corrections are not the

biasing the results (consider but significant improvements

for example the poor world coverage of of this parameter must be researched.

Many teams have pointed out that multiwavelength methodology required amelioration. Anyway it is quite clear that this will ask

could p_vlde this for a tremendous

technological effort and will probably not be operational since some years. For that reason we have looked for a new solution, able to give better tropospheric correction with up-to-date' techniques, even if probably not as efficient as 2-colour method, but immediately available.

II/THE

and

ATMOSPHERIC

CORRECTION

If n is the Index of refraction, c in the atmosphere). Classically

we define

and

for optical

wavelengths,

N=A_(P.

/T).{

SLR

n = c o / c (c o is the light

the co-index N=

FOR

(nusing

of refraction 1).10

celerity

In vacuum

as

6

for example

Essen's

I ÷ (a-b.t).P.}-B

x(Pv

formula: /T)

where P. is the atmospheric pressure, Pv is the water vapor pressure. T is the absolute temperature and t is the centigrade temperature. Ax and B x are _.dependant parameters, and a and b are constants. With a I0 -7 precision it is acceptable to use the simplified formulation: N=A'_tP.

/T)-B_(p_

/ T)

We define also the geometric length of the optical ray Lg, the optical path Lo, and the geometric distance L between two points s o and s t being one at the ground level and the other beyond the atmosphere, at the satellite level:

$1

$1

,, = fa,

L. = f,(,).d,

so

the

so

If we call AL = L o - L. it is a function of tile angle direction of the satellite. We notice that: AL= L ± _2± i:

:

tL o-Lg)

a between

the

vertical

and

+ (L z-L)

::

As the Curvature

of the ray pa{i_ ]glow.

and

as this

curvature

only on a small range (a few tens of kilometersL the difference considered as negligible. On another hand. AL can be expressed AL = f(a) . AL,,_tl_ t

8-24

IS experienced

Lg - L IS generally as:

And the function f(a) may be found in (Berrada-Baby et al., 1987 or Akhundov et Stoskii, 1992). We have now to deal with AL for _ = 0. h o being the elevation of the SLR station above "sea level",

w

AL = f N(h).dh ko

for these

Considering the very optical wavelengths,

low dependance we will focus

of the result regarding on the "dry" part of this

the water expression

vapor AI d :

m

AL,: f A P--'.dh T ho

And. p(h) being atmospheric pressure h. with R = 287 J.kg/°K

the and

* P. = p(h). * d P(h)

air density. T(h) the absolute temperature, g(h) the acceleration of the pesanteur at the

R.

P(h) the elevation

T(h)

= p(h).

g(h).

dh

so that.

AL d = f

dP(h).A.R

ho g(h)

At this

level

we may

have

different

assumptions:

I 't hyp.

g(h)

= go = constant.

2"

g(h)

- go / ( 1 + 2.h/r

hyp.

aLd-

And

gradient

the difference

AId = A'R'Po

o ) closer

to the

/ go reality.

We find

then:

A.R.P o I

+

ro -p- (go ÷ R. "to

8o

( _ is the

Then

of T ). For Po = I000 between

these

mbar,

two models

- 2.go

we find is below

]

ALd -_ 2,3 I cm

meters

: various

studies

show (Berrada-Baby. 1987) that this discrepancy is around 5 mm. and is quite stable and thus easy to model. On another hand, the differences between the second model and real data deduced from (Cira 1965) campaigns are quite small (I mm

typically).

8-25

So excellent And one

we conclude that it is extremely important to measure I% with an precision (1 mbar of error induces 2.3 mm on the atmospheric correction). must take into account the fact that the function f(a) (that is the air-mass

number) will measurements.

multiply

these

If the temperature is fig. 2, we observe that half the upper layers are quite compare these facts with around 1-2 mm, multiplied it is clear will leave

values

by

numbers

up

to 2.5

in operationnal

SLR

measured in the first 5 kilometers, looking at fig. 1 and the total correction is already acquired (fig. I), and that well defined by the profile in the troposphere. If we the abovementionned value of 5 mm. whose noise is by the air-mass value (I to 2.5), i. e. less than 5 mm,,

that any temperature profile of good a residual error on the atmospheric

precision correction

acquired at the

in the troposphere millimeter level.

1.0C ff Jf

3.gc

J

3.8C

J /

).7C 3.6C 3.50 /

3.40 ).30 ).20 3'.10 ).00 0

10000 5000

20000 15000

30000 25000

Fig. I: Fraction (Y-axls) of the total atmospheric correction for a vertical transit of a laser pulse, from the ground level up to a given altitude (X-axis, in meters). Computed from an observed radio-profile in Greece, 1990.

8-26

Z[_a! J

\

_\

6C

/ /

/

I I I

/ 1963

temperature Avenge standard

20 ¸ II

II

_emperat'ure 1961

1963

BCA-60

!

i I

\

0

2o0

15o -150

-]00

250 -50

3or, 0

T[CK ] 40

T[ c

'

Fig. 2: Typical temperature profiles (from CIRA 1965)• If the possible variations are quite large, it is also noticeable that the profiles have the same topological aspects and are quite "parallel", with no intersection from one to another.

8-27

HI/PROPOSITIONS The first one is obviously to measure Po quite carefully. calibrate often the barometers employed, to measure preferably to the level of the axes intersection of the telescope (it is generally

It is necessary to the pressure close the case). It must

be posslble

< 0.2

to measure

the pressure

with

an absolute

precision

mbar.

Considering the efficiency ofbarometric levelling, whose precision may reacll one meter, we observe that it means that constant pressure surfaces are quite horizontal, so that it is useless to measure the horizontal gradient of Po in order to correct its effect in the direction of the sight, provided the weather is reasonably quiet. Anyway, it seems mostly advisable to improve the correction by a good measurement of T(h) in the direction of the satellite, at the O. I°C level, up to an elevation' of 3 to 10 km. In these

conditions,

one

may

be sure

to get

an atmospheric

than 1 mm. It is easy to notice that such a precision require a 20 times better precision on the differential between the two colors (i. e. 0.5 picosecond), which

correction

better

with 2-wavelength SLR will time-of-flight measurement is quite uneasy to reach.

The solution we propose for an easy measurement of the temperature profile along the llne of sight is a SODAR (Acoustic LIDAR) associated with a RADAR, the acoustic and the radio wave having the same wavelength. The radar is used to track the acoustic wave, so it allows to measure the speed of the sound in the atmosphere, which in turn provides an excellent temperature profile (at the 0. I °C level). Such instruments (e. g. Remtech) are now used close to airports, in order to measure winds and wind shears, and this technology is already available. The common

ranges

are

3 to 5 km,

but

large

8-28

instruments

may

reach

I0 km.

IV/

close that

CONCLUSIONS We propose that an experiment be carried with a radar/sodar equipment to an up-to-date SLR station in order to check the possible improvements such data could provide to the laser data. This solution is probably an

excellent stations.

alternative

allowing

to wait

for

On barometric

another hand. we insist on the equipments used in SLR stations,

important

to perform

a good

atmospheric

the

2-wavelength

necessity as the

generation

to check P, parameter

of SLR

quite carefully is by far the

the most

correction.

BIBLIOGRAPHY

H. BERRADA-BABY.

P. GOLE

& J. LAVERGNAT

Effets de la troposphere sur les mesures de distance Terre-Satelllte. Application au proJet DORIS. CRPE/158 Technical note. June 1987 RPE/ETP, 38-40 rue du G6ndral LECLERC. 92 131 ISSY-LES-MOULINEAUX, France

CIRA

1965

North-Holland

T. A. AKHUNDOV,

Publishing

Company.

Amsterdam

Edition,

1965

A. A. STOTSKII

Zenith angle dependance of path length through the troposphere Proceedings of the symposium "Refraction of transatmospheric Den Haag, 19-22 May 1992, pp 37-41

signals

REMTECH Company, technical papers on acoustic wind profilers. Remtech Inc., 1,040 Townley Circle, Long_ont, CO 80 50 I, USA

8-29

in geodesy".

Operational

Software

Developments

N94-15589 On the Accuracy

8th International

of ERS-1

Workshop

Orbit

Predictions

presented at on Laser Ranging

Annapolis, May

MD

18-22,

Instrumentation

USA 1992

by R. Kt_nig, H. Li, F.-H. Massmann, J.C. Raimondo, Deutsches Geodiitisches Forschungsinstitut, German

Processing

C. Rajasenan, Abt.I (DGFI)

and Archiving Facility MiJnchner Str. 20

D-8031

Ch. Reigber and

(D-PAl)

Oberpfaffenhofen Germany

Summary Since

the launch

cessing

and

of ERS-1

Archiving

(first European

Facility)

provides

Remote

Sensing

regularly

orbit

predictions

The weekly The tuning

distributed procedure,

(Satellite Laser Ranging) tracking network. called tuned IRVs and tuned SAO-elements.

Satellite)

the D-PAF for

the

(German worldwide

orbital elements are designed to improve

accuracy of the recovery of the orbit at the stations, is discussed based on numerical This shows that tuning of elements is essential for ERS-1 with the currently applied

ProSLR sothe

results. tracking

procedures. The the bias

orbital time

piled and

bias

function

Finally

elements

the for

are updated

function increases

quality

about

250

has considerably

by daily

distributed

is explained.

Problems

the prediction

accuracy

assessment days

since

of ERS-1 launch.

and

bias

numerical

functions.

results

The generation

are presented.

The

of time

considerably.

orbit The

iime

predictions

average

improved.

9-1

is described.

accuracy

lies

in the

The

accuracy

range

is com-

of 50-100

ms

1. Background Since

and

the very

Introduction

first days

when

ERS-1

predictions to the SLR community. tablished IRVs and SAO-elements. elements are very compressed the orbit. The trick to achieve luminated Very

from

soon

the numerical

after

launch

forms of complicated this is called tuning

The

it became

clear

extremly such

high

solar

the D-PAF

is sending

orbit

in form of the well es2. Though these orbital

orbit trajectories, they of elements. In chapter

that the ERS-1

orbit predictions the time bias

Meanwhile the time bias function features of the time bias function

events

in space,

optimally represent 3 the problem is il-

point of view.

and geomagnetic activities. The The D-PAF therefore introduced daily. Some

began his mission

The orbit predictions are produced The procedure is given in chapter

activity

as a geomagnetic

is extremely

affected

has become an accepted tool in the SLR and related topics are explained in chapter

in the

storm

trajectory

first

of nearly

by solar

lost accuracy mainly in along-track direction. function which is generated and disseminated

months

of

singular

the

type.

mission

Though

brought

SLR

community. 4. extraordinary

tracking

was

almost

lost for two days, the D-PAF orbit prediction system (and the SLR community in turn) gained at the end. The software system got more backup solutions and the quality checks were largely extended.

Chapter

Finally the orbit becomes obvious the more

prediction from the

detailed

2. Generation In a first

analysis

of Orbit

step

DOGS-OC

5 tells more

the

system course

quality

assessment

has developed of the average

in chapter

and quality

checks.

to a reliable system in SLR tracking. This. accuracy of the orbit predictions shown in

5.

Predictions

orbital

(DGFI

about

Orbit

parameters and

are estimated

Geodetic

parameter

by differential estimation

orbit

System

correction

- Orbit

with

the

Computation)

software. Laser ranges are mainly used as observations for the least squares procedure few cases also RA FD data (Radar Altimeter Fast Delivery) ( Massmann, F.-H. et.al.,

and in 1992 ).

The use launch.

prior

Orbit

of PRARE

data

perturbation

represented by non-gravitational OC adopts!a Earth's high activity

models

(Precise

used

the GRIM4-S2 perturbations

Range

are

And

state

gravity of the

Range

of the

rate Equipment)

art ones.

field model low altitude

The

( Schwintzer, orbit stemm

was

also planned

gravitational

forces

are mainly

P. et.al., 1992 ). The major from surface forces. DOGS-

macro modell of the surface of ERS-1 for albedo, direct solar radiation and drag. atmosphere is represented by the CIRA'86 model where solar and geomagnetic

is needed

as input.

In a second ters derived

step the orbit is integrated forward by the DOGS-OC software with the paramebefore. Adopted are the same perturbation models but now solar and geomagnetic

activity

predicted

that

are

degrades

to

for the respectiv

the accuracy

of ERS-1

time period. orbit

predictions.

And

this indeed

Also

the ERPs

ters) are predicted for the time period of the forward integration. tions is quite good for the lifetime of ERS-1 orbit predictions, neglected. 9-2

is the major (Earth

error

rotation

source parame-

The accuracy of ERP predicthe effect can therefore be

In the third step DOGS-PD (PD for Predictions) compresses the orbit to dedicated forms and formats of orbital elements which can easily be disseminated worldwide via various telecommunication channels. In addition these elements are tuned so that the user will recover the orbit in an optimal (Smithsonian elements

way.

So-called

Astrophysical

to the PRARE

tuned

IRVs

Observatory)

(Inter-Range

are

Vectors)

delivered

to

the

and

tuned

SLR

SAO-elements

community,

PRARE-

system.

3. Tuning The orbit

is recovered

computers.

These

nongravitational starts cover

at the SLR

programs forces

sites by simple

adopt

as drag.

low The

degree

and

IRVINT

of these

despite

severe

of simple

optimum

neglections,

perturbation

is a minimum

program

D-PAF

modelling

difference

tunes

onsite

IRVs

squares

orbit. The IRVINT assumes for IRVINT

GEM10N

up

gravity

should run coefficients.

field

in the

RGO

coefficients version

to degree

(Royal

Fig. 1 shows the along-track error if an un-tuned holds in general. Due to the missing drag forces ca. 1400m ;in along-track the satellite.

position

and

Greenwich

or ca. 200ms

gravity

fields at the

and

and

do

orbit

SAO-elements.

consider of Texas

theory

18. The

orbit

to re-

means

that

way.

The

in an optimal

and AIMLASER a 60s integration

Obse_atory)

not

This

of the recovered

order

capacity

university

analytical

is recovered

sense

perturbation model orbit or reference run then in a defined mode. D-PAF

for use on low

written applies

the orbit

in the least

intended

order

integrator

from the IRVs. The ancient AIMLASER the orbit from the SAO-elements.

Because

programs

to the full

programs step size

have to and the

AIMLASER

program

dedicated

GEM10B.

with

IRV is integrated by IRVINT. The example modelling the error grows during the day to in time

bias

which

would

prevent

tracking

of

Fig. 2 displays the error after tuning. Maxima are now ca. 20ms around midnight and around noon. The remaining along-track errors are similar day by day, the maxima are dependent on the

solar

a day.

The

remaining

activity. more errors

They

could

proper

way

are in a range

be accounted of course

for by a function

would

be to model

that can be accepted

9-3

giving

drag

for tracking.

the error

in the integrator.

for 24 hours Anyway

of the

[ m ]

ALONG-TRACK ERROR

1200

1oo0

800

6O0

400

200

DGFI 2

4

6

Figure

8

10

12

1 Along-track

"

'14

"

1.8

1.8

20

error from un-tuned IRVs

ALONG-TRACK ERROR

[m]

[h]

22

J

150

1oo

5O

-5O

-DGFI -lO0 o

2'

4

6

8

I0

12

Figure 2 Along-track

9-4

14

1"6

18

error from tuned IRVs

20

2"2

[h]

The accuracy

of tuning,

that is the difference

of the recovered

orbit to the reference

orbit, can

be seen in Table I where the typical differences are given as RMS-values in meters for both types of prediction elements. Again the values are largely dependent on the solar activity.

Table

IRYs

SAO-elements

ld

lOd

radial

3-5

10-20

across-track

3

10

along-track

20-60

20-100

1: Tuning

The

accuracy

SAO-elements

(in m)

show

less

accurate

recovery

period (normally 10d) whereas the IRVs

because

they

represent

the whole

prediction

are given each day. All values are within a range

thatcan be accepted for tracking. 4

4. Time

Bias

D-PAF Due

generates

to the

curacy

and disseminates

sensitivity

mainly

bias provides Therefore Look)

D-PAF

prediction

to variations

direction.

and simple

generates

ranges.

relative

and

For every to the

update

pass

IRVs

is also rises

known

the predictions

to the

too early.

stations

The

function.

over

A typical

example

removing

the outliers

The time

bias

of time

function

is capable

of predicting

The

(European

EDC

problem. Currently in real-time

The function

bias

daily a time all stations

set.

A third

fits to the time

values

bias function

having order

Data

Center)

available

to other

values

from

observed function

bias values

in Fig.3. check

the time bias

the along-track

actime

of the time

Fig.4

new

ERS-1, in time

Q/L

a time is fitted

(Quick bias

is

by LS

the time bias values are ,leading all in all to a best within

shows

1ms

for up to four

the fitted

function

after

of the parameters. of the last few days.

More

important

it

error. provides

Stations can store the observed Graz, Potsdam, RGO, Santiago

own computations to more confidence

is given

and after the significance represents

loose

as so-called

determination

(Least Squares) to these time bias values. In the LS procedure checked for outliers and the parameters are checked for significance approximating days.

and SAO-elements.

of the orbit predictions.

disseminates

last prediction

sets containing

of the high atmosphere

This

rises too late or the satellite

a quick

Laser

computed

of ERS-1

in along-track

bias i.e. the satellite

weekly

stations.

another

important

time bias immediately de Cuba and WettzeU The

stations

can

and with the D-PAF time bias function. in the time bias that is going to be used

9-5

then

tool

to handle

the

time

bias

after the pass in the data base. do so. This time bias is nearly compare

the

results

This all this together for the next upcoming

with should pass.

their yield

Time Bias

[ms] .0

-5O

-100

-150

DGFI 0.0

0,5

1.0

1.5

Figure

2.0

3 Time

2.5

3.0

3.5

4.o[_]

bias values

Time Bias

[ms]

-50

-IOO

-150

-200

0.0

0.5

1.0

1.5

Figure

2.0

4 Time bias

9-6

2.5

function

3.0

3.5

4.0[ d]

5. Quality The quality of the ERS-1 prediction products is indicated to the users by a quality flag. In case of orbit predictions the quality flag is computed from the fit of observations in the orbit analysis (step 1 in chapter 2), from the comparison of overlapping arcs and from the tuning accuracy. The quality of the time bias function is assessed from the approximation error of the function and the dam coverage. Of course the quality flag gives only an idea on how the quality of the products can be. By nature predictions go into the future and nobody can determine what will be. Particularly the accuracy of ERS-1 orbit predictions is heavily influenced by unpredictable solar activity. Anyway the quality flag shows the current knowledge of the quality. It convinces the generator that he did a good job and it gives a first indication to the user in case he has problems. All prediction products are subject to permanent quality checks. Dally the time bias of newly incoming Q/L Laser passes are determined relative to the actual prediction orbit which leads to the time bias function. The prediction orbit is also compared to the MMCC orbit (Mission and Management Control Center). The MMCC orbit is generated by an independent group in ESOC (European Space Operation Center), Darmstadt, from S-band observations and radar altimeter fast delivery ranges. The comparison with such a completely external source provides more confidence in the results. On a weekly basis when new orbit predictions are generated the new prediction orbit is compared to the old one. The differences in along-track direction should exactly reflect the old time bias function. The differences in radial and across-track direction normally stay within a few meters. The internal quality checks proceed in a closed loop from prediction set to prediction set. The external checks give an independent judgement. So the reliability of the orbit prediction system has reached a quite high level. Fig.5 displays the along-track errors of ERS-1 orbit predictions since launch. The first 100d are clearly affected by large solar activities. In addition a manoeuvre was missed. Meanwhile these cases are taken care of. In the right half of Fig.5 it can be seen that the mean accuracy steadily improves, but is clearly degraded by the frequent manoeuvres.

9-7

[ms]

Along-track

Errors

(7 days

moving

!t!

4OO

mean)

i!

i !i! i!

",i n°'uvi

so,or _u_i



! i!i!i ! i !i! ! i!i!i i, ii! i! ! •

ii!

i

.3OO

i'

i •

i

200

°

i DOFI

°

i !i!

i!

i !i!

i!

i

i !i!

i!

i

,!;i

;!

i_

i!

.

ii!

_i

iii

iil

! i!i!i

100

ii

0

i

i •

50

0

100

Figure

°

150

5 Accuracy

of ERS-1

.

200

orbit

[d]

250

predictions

6. Conclusions It has been ly time

bias

shown

that

functions

ERS-1 prediction manently checked since launch.

tuning

of orbit predictions

that account

products internally

is essential

for the along-track

for ERS-1.

D-PAF

error of the predictions.

provides The

is assessed and indicated to the users. The predictions and externally. The average accuracy has considerably

dai-

quality

of

are perimproved

7. Literature Massmann, Laser

F.-H.,

Reigber,

Ranging

Ch., Li, H., Ktinig,

Network

Performance

R., Raimondo, and

Routine

J.C.,

Orbit

Rajasenan,

Determination

C. and M. Vei: at D-PAF,

this

proceedings Schwintzer,

P., et.

sungswesen

14 al.:

GRIM4

117, pp. 227-247,

- Globale 1992

9-8

Erdschwerefeldmodelle,

Zeitschrift

£dr Vermes-

N94-15590

COMPENSATION FOR PRED%CTIONS

THE DUE

DISTORTION TO VARYING

Matti Finnish

IN SATELLITE PULSE TRAVEL

LASER TIMES

RANGE

Paunonen

Geodetic Ilmalankatu

SF-00240 Helsinki, Telefax:358-0-264995

Institute IA Finland

ABSTRACT A method for compensating for the effect of the varying travel time of a transmitted laser pulse to a satellite is described. The "observed minus predicted" range differences then appear to be linear, which makes data screening or use in range gating more effective.

I.

INTRODUCTION

Accurate range predictions are necessary in satellite laser range measurements when the operation takes place in daylight. Then the range gate, where the return pulse detector is active, can be set very narrow to effectively discriminate against noise. Data screening of the observations is often done using ',observed minus predicted" (O-C) range differences. The predicted satellite times are generally equally spaced, as are the transmit times. But the true hitting times are not equally spaced because of the varying pulse travel times. This leads to distortion of the O-C differences which can be as high as 10-20 m. A polynomial of suitable degree is used in screening. It is well known that the polynomial should be of low degree so as to avoid an artificially good fit or end effects. If the travel time distortion is removed, the O-C deviations are nearly constant with time for high quality orbit predictions. This paper describes and tests a simple correction method.

9-9

2.

METHOD

The

OF

range

RANGE

from

CORRECTION

the

laser

station

to

the

satellite

varies

approximately parabolically with time, Fig. i. When the satellite is passing the closest point, a small time change does not affect the range much. At the far end of the pass the range changes several kilometers per second.

\

P /

\ \

%

\

A Ri/_'

TRUE POSITION

Rmin

Ri_/

Ri

o

CALCULATED ....

/ _/

Fig.l

LASER STATION

Conceptual scheme for the time distortion in ranges.

origin

ot

the

travel

This rate , vi , can be calculated together with ranges, or it can be approximated from the change the successive predicted ranges Ri_ ] and R i vi = where The

T

is

travel

Rmi n

, and

the

time

step

rection

c

difference

the

instantaneous

is

the

_R i to

speed be AR

Ri_] ) / T

(often

time

Zit i = where

( Ri -

between range

the

Minimum

_

is

light.

Then

to

the

predicted

i = vi

*

At i



range

(2)

,

added

9-10

(1)

,

1 s).

(R i - Rmi n ) / c of

the between

the

travel range

time

cor-

is

(3)

3.

RESULTS

AND

A test involved

with

program /i/. A

used 220 ms

DISCUSSION

real data relatively

is few

shown in Fig. observations

highly accurate time correction

long-term was used

12 EIO

o

"_°.o. 8

o

a)

\

o

@

°

6

,.ul

o o

,,,

4

"_

2

IRV predictions the calculation.

-o. o

F--

_

in

METSAHOVI 1990-12-4 LAGEOS O-C DEVIATIONS

_..

z

2. This LAGEOS pass (21). The prediction

o\

,_

KEPLER-ORBIT 0

_

0 !

I

|

I

|

I

I

0

2

4

6

8

10

12

TIME(min)

z o

_-

8

o

6

>

'-'

METS_HOVI 1990-12-4 LAGEOS TRAVEL-TIME CORRECTED O-C DEVIATIONS

b)

E

I

16

FROM 15h57mo0 s

lO v

i

14

°o °'-°°°

--o __._°o o-

4

o

0 !

OO

(D

LI-NORM

0

I

I

i

I

I

I

I

I

2

4

6

8

lO

12

14

16

TIME(min)

Fig.

2.

FROM 15h57mo0 s

a) 'Observed-predicted' measured LAGEOS pass. b) O-C linear

FIT

deviations median fit

A

(O-C) Kepler

after the is shown.

9-11

deviations orbit fit travel

time

of a is also correction.

shown. A

The time error varies slowly known from earlier observations.

with

time, and is normally The O-C deviations are

shown in Fig. 2a. A simple Kepler orbit vations quite well, and only one outlier In polynomial fitting determination of easy. The first point would be rejected

/2/ is the in

fits the obserindicated. order is not a linear fit.

The situation becomes considerably clearer after application of the travel time correction method described (Fig.2b). The fit using the linear Ll-norm /3/ eliminated the last point as well as the known outlier. Because the model is now linear, it is not necessary to use higher order polynomials. This example possible in to 7.5 m of implementing As has been

shows that a very narrow range window is ranging to LAGEOS. A 50 ns window corresponds range. This would help considerably in daylight satellite laser range-finding. seen, the method can also help resolve difficult

data are

screening essential

tasks. Good orbit for this method.

elements

and

Note the O-C

that recalculation of the orbit after satellite hit times, produces approximately deviations as the method described.

the

a

good

orbit

pass, using the same

REFERENCES

.

The

Texas

University,

long

term

IRV-

predictions

for

LAGEOS satellite; a satellite prediction program developed at the Royal Greenwich Observatory, Herstmonceaux, England. .

.

J_ Kakkuri, O. of the Finnish Reports of the Helsinki 1978,

Ojanen, satellite Finnish ii p.

M. Paunonen, Use range observations cosmos Symposium Observations 1989, Cracow, of Artificial Polish Academy

ORBIT

M.

Paunonen, Ranging precision laser range finder. Geodetic Institute 78:8,

of

the Ll-norm to satellites. on the Use of

in

screening laser Proc. 18th InterArtificial Satellite

for Geodesy and Geophysics, June Poland (Ed. WI. Goral, Observations satellites of of Sciences,

9-12

12-17,

the Earth, No 27, 1989, Warsaw 1990),pp.359-362.

94-15591 Timebias

corrections Roger

Wood

and

to Philip

predictions Gibbs

Satellite Laser IL_nger Group Herstmonceux Castle, Ha_lsham East Sussex, BN27 IILP, UK

Abstract This talk aims to highlight the importance to an SLR observer of an accurate knowledge of the timebias "corrections to predicted orbits, especially for low satellites. Sources of timebias values and the optimum strategy for extrapolation are discussed from the viewpoint of the observer wishing to maximise the chances of getting returns from the next pass. What is said may be seen as a commercial data centres for

encouraging wider mutually beneficial

and speedier use of existing exchange of timebias data.

Introduction The real behaviour of low satellites is never in exact agreement The principal reasons axe that modelling the effect of atmospheric

with even the best predictions. drag is difficult, and the pertur-

bations due to solar radiation pressure are not predictable in advance. EI'tS-1 is much larger and somewhat lower than either Staxlctte or Ajisai and consequently presents the biggest headaches in making_predictions. Dcspitc the problcms, dctailed elscwhcre in these proceedings, the prediction teams do an excellent job in providing orbits and thereby chance of tracking the satellites succcssfully. Fortunately, orbit appears as an along-track crror and can be corrected predictions (as has been done 'very successfully for Lageos set of predictions observing stations can, for each satellite,

give observing stations the best possible most of the deviation from the predicted by applying a suitable timebias to the for many years now). For a particular model the history of the behaviour of

the timebias with time and extrapolate to predict the timebias at the time of observation of the next pass. At Herstmonceux we have been working to improve the accuracy of our extrapolation of timebias values. This talk outlines our findings, indicates the benefits to be gained from increased use of the data centres and shows how helpful good timebiases can be if you have English weather and are trying to keep us with Grasse! Why

do

we

We see three

need distinct

accurate

timebiases?

advantages:

Ease of satellite acquisition. This is always important, at whatever stage of the pass the satellite is first acquired. But it is especially important at the very beginning of a pass, necessarily at low elevation_ whcn accurate pointing helps to minimise the time spent returns. Ease of acquisition is also importaz_t when there has been a break

in searching for of a few days in

the scquencc of observations, because of cloud or technical downtime, where accurate knowledge of the timebias correction can lead to immediate recovery of the satellite. This second point emphasises the usefullness of having access to continuous records of timebias data in order to base any extrapolation on the best available data set.

9-13

Improved

telescope

tracking.

If an incorrect

timebias

is used

to compute

the telescope

position during a pass, the offsets from the reference time and the observer will have to insert ever-changing

position on the detector will depend on manual corrections in order to keep the

satellite

i_ [|ttle

image.

pointing Better

By contrast

a correct

model for the telescope noise

elimination.

timebias

results

or no guiding,

provided

that the

is weU-determined.

When

the timebias

is exactly

right the returns

from the satellite

always appear at the same place in the range window. This means that the window can be made very small and so the amount of noise reaching the detector can be drastically reduced. This improvement The second

in the ratio of signal

and

third advantages

to noise is particularly

can still be gained

valuable

for daylight

(even in cases where

ranging.

an accurate

timebias

is

not available beforehand) if the first few good returns from the satellite are used in real time to convert a range offset into a timebias correction. This technique is routinely applied for multiphoton detection systems where the identity of true returns is not in doubt. But for single photon detection in daylight, or using an intrinsically noisy detector such as a SPAD, it is not always easy for the software to pick out the track, and some form of manual intervention may be required. Where

can

Originally

we

get

timebiases?

we used only our own history

of passes to make extrapolations

to current

behaviour:

Tl_is

was quite satisfactory during periods of continuous good weather, but was often very frustrating when trying to recover satellites _ter a break of two or three days due to cloud or instrumental problems. Then, for Ajisai and Starlette, we tried using Bendix's timebias data from many stations in addition to our own. We immediately found a great improvement in the quality of our extrapolation and could use narrower range windows right at the start of the pass. We also found that the change from using elements to using IRVs gave better consistency in the timebias histories. When EILS-1 was launched we, like everyone else, had mixed success in finding

it.

When we

had a reasonable history and the solar activity was not too wild, we got good passes. But at other times, once we had failed to get passes for a day or two, it became increasingly difficult to recover the satellite. We were very pleased when DGFI _reed to use the pass data that they receive from many stations to produce timebias functions for EILS-1. This gave, and continues to give, valuable additional

weight

to timebias

trends,

,il,001

v

[-.

87 g_

I

!

I

i

I

_

[]

RGO

x

Potsdam



Santiago

0

de

Cuba

Graz

o o I

I

i

48720

i

i 48722 Julian

48724 Date

48728

Figure 1: ERS-I timebiasesfrom EDC

9-14

for IRV setA920,_07 and symbol key

Since

then, following

the

establishment

of the European

Data

Centre

(EDC),

a small

stations have begun to make regular deposits of their timebias data for ERS-1 observations. We often get EDC data from Graz and Potsdam--and occasionally

number

of

soon after the from %Vettzell

and Santiago de Cuba. Figure 1 shows a plot of the data deposited for IR.V set A920407 at the beginning of 1992 April. The interagreement between the values from different stations is striking and gives great confidence that all timebiases have been reduced to a common system. But, with all the stations close together in Europe, it is clear that the data are clumped together in time: indeed, sometimes they are all clouded out simultaneously! The addition of more data points from other stations

at other longitudes

How

can

we

best

use

around

the world would give mucll better

timebias

"round

the clock" coverage.

histories?

1

EB.S-1 timebias

Over the last two or three months we have been investigating how best to predict future values from whatever timebias history we ha_l available on a particular day. This usually

consisted of a small number of values covering a few days. First we looked at the timebiases shown in Figure 1 which cover a period of about a week and are all referred to the same IRV set from DGFI. Our approach was to see, in retrospect, how well the observed values available up to a certain day could be used as predictors for the points which followed. The results of our trial fits are shown in Figure 2. First we fitted to data for the first two days only; next we used data for the first three days; then the first four days and

__'-_A.'

O

'

'

I '

'

' /II'

II[I_

O O

O

O O

•....

I

:S

7 $ 7 $ T

,

E O I

-,

I ,

,

48720

I,

,

,,

48722

°,

_ tXl

48724

48726

I,

For each

O

O

E

v

so on.

i

"'-

, I , 48720

, , , , , I , 48722 48724 Julian Date

Julian Date

°

48726

°:,

r_.l

, '

,

I

' '

' I

'

'

'

I ' -

O

o

7

'_,.

:

I

i-r.

,;._.

i

-

{W I O O iv) I O O

O I

O

-,

I , 48720

, i I , _ I , I , 48722 48724 48726 Julian Date

Figure

2: Successive

fits to ERS-I 9-15

I

t

48720

timcbiases

48722 48724 Julian Date

from

EDC

_,.

kxj 48726

--

trial we fitted

the data

with functions

of three different

orders--linear

(represented

by a solid line

in Figure 2), quadratic (dotted line) and cubic (dashed line). Only points to the left of the vertical arrow at the bottom of each plot wcre included in the fit. We then looked at the next few points to the right of the arrow to see how well, or how badly, each of the attempted data which followcd. From this examination we drew a number of preliminary • Use the simplest fit)--it

function

will generally

• Use only linear

wlfich gives

be better

a reasonable

fit to the data

for extrapolationthan

fits for the first 2 or 3 days--there

a higher

(even

order

are too few data

fits represented conclusions:

the

if it is not the

best

function;

points

to justify

anything

fancy; • It would coverage.

be helpful This

to have

quick

o fill the gaps in the Europema o add data o provide

timebias

deposits

by all stations

points

longitude

actual

timebias

the fit;

values

closer to the time of observation

in the Bendix

timebiases

use in revealing a week in early

than

are avMlable

now.

based on normal point data for EKS-1 passes extracted to give timebiases. [We had hoped to use directly the

Bendix values, but there are currently unexplained discrepancies between tlmebias and ours, so we have worked entirely from the normal point data. prevent their which covers

better

data;

to strengthen

More recently we tried a new experiment from the CDDIS database and reduced

to exist

and

would:

for Ajisai

and

Starlette,

but

their derived values of This effect also appears

to a lesser extent

which

does not

long-term trends]. In this case we took the data set shown in Figure 3 May and is referred to IRV set A920505 from DGFI. It is clear that _---]-_--iiiiiiiII1'""'1111_ O O B

t_

E

B

v

oo E 0 O

48748

Pigure the longitude is a drawback

3:ERS-1

48750 48752 Julian Date

timebiases

from

CDDIS

48754

normal

points

coverage is more uniform and the daily density of points is higher. However, there as far as immediacy is concerned in that the turn round time for data input can

be longer than for EDC, so procedure that we used for well the extrapolated curves Figure 4 again show that the

that some of the data are relatively old. We then followed the same the EDC data and made trial fits to subsets of the data to see how acted as predictors for what was subsequently observed. The plots in higher order fits are wildly ineffective for the first few days, and only

begin to make more sense later. But around day 48752 the data take a wholly unexpected turn towards large positive values and the trends evident in the early points are no longer relcvant. This sudden change of behaviour is probably due to solar activity, andin such circumstances it is wise to abandon the early points of the data set and start the fit again from part way through.

9-16

'

oo

I '_

(w

'

'

I

i

.

I _ ° ° I ' J ' '-[] I in

-4 []

-I

I

, I, g

oo

I

,

,

,

I

,

[]

, j 11

I

I A

I

E

!

\

o_

o

I

I

[]

o

48748

48750 Julian

48752 Date

48754

48748

4{]750 Ju]ian

48752 Date

48754

o O O3

.--.

E

_-' I

'

'

'

I

'

'

'_/'

'

' '[]I

_

v

J

[]

/--

/

o-_

/

4

/

E []

[] / /

.. 13

o

./'"

0

48748

48750

48752

...*''"

48754

48748

Julian Date

Figure

Our

_:dditional

4: Successive

fits

as a result

of this

conclusions



For El%S-1 it is necessary to examine pected trends in good time;



Better



Access to the any functions!

Ajisai

and

longitude

coverage raw

data

Starlette

is essential is very

Generally

exercise the

data

from

48752 Date

48754

CDDIS

were: functions

warning

for making

prediction

timcbiases

and

for early

useful

the

to ERS-1

48750 Julian

value

d'aily

in order

of departures

from

judgements

of timebiases

for

on

these

the

to pick

previous quality

satellites

up

unex-

behaviour; of the

is much

fit of

more

straightforward than for EllS-1 since they are relatively unaffected by atmospheric drag and solar radiation pressure. Indeed orbit predictions can run quite satisfactorily for several weeks so that a long timebias history can be established for one II%V set. Over such an interval timebias corrections usually to one What

behave sensibly and, even when the corrections get quite large, a good set of IllVs allows meaningful extrapolation to get accurate values. can

We think trends in possible

be

done

to

that all stations timebiases for after

the

pass

with

respect

improve?

can EllS-1

has

history

share the clear by depositing

ended.

Then,

benefits to be gained their timebias values

by retrieving

9117

a comprehensive

from with

the daily examination data centres as soon set

of timebiases

of as

obtained

==

worldwide, and using itas the basisforextrapolation, the "hitrate" can be significantly improved. Alternatively, but probably with greaterdelay,stationscan retrievenormal point data and derive timebiasesfrom .them. Itwould be very worthwhile to extend the system to includeAjisaiand Starlette; and to make sure that provisionismade forrapid feedbackof Topex/Poseidon data. Finallythe proper commercial: we think that timebiasesarejustlikewashing powdcr--use the rightone and your whole lifeis transformedforthe better!

Figure

5: "Net# improved

Timebias _

Acknowledgements We

wish to thank the staffof the predictionand data archiving centresfor advice and data.

Specialthanks go to Brion Conklin of Bendix forhis speedy and cheerfulresponsesto our many requests.We greatlyappreciateassistancefrom Rolf Koenig of DGFI and Wolfgang SeemiiUcr of the European Data Centre.

9-18

N94-15592 Formation

of on-site

Normal

Points

G.M. Appleby and A.T. Sinclair Royal Greenwich Observatory Madingley Cambridge

Road

CB3

0EZ,

UK

Abstract. We

describe

from

our

methods

a predicted

directions.

orbit,

The

predicted

orbits

analytically

of fittin 9 a smoothin by deducing

method

has

computed

from

been

9 function

corrections used

on

elements.

The

to the

orbit

observations

both by numerical

orbital

to observational in the

of a variety

integration

along-track

differences

and

along-track

radial

of satellites,

usin 9 [RVs

corrections

range

and

as startin

to the

using

9 values

and

orbit

have

predicted

been succesfully used in the form of time biases to improve subsequent predictions, and a statistical test has been devised to ensure that the range residuals may be used to form unbiased quicklook

normal

points.

1. Introduction. It was

recommended

at the

Fifth

International

Laser

Ranging

Instrumentation

Workshop

held at Herstmonceux, UK, that laser range normal points should be formed by the stations shortly after each satellite pass and transmitted as quicklook data (Gaignebet, 1985). The final

version

of the

the CSTG The

quicklook

and published

process

of forming

normal

to the observational order to form a flat

range track

mean

accepted

values

smoothing

of the function

normal

point

in the April

and

points

differences of residuals,

consists

predicted

orbit

of the bin are then added points are then virtually

of the

orbit.

In this report we detail for rejecting noise and (these 2.

by the

Ranging (a)

SLR

Subcommission

Newsletter fitting

spread

at the

through epoch

to the mean independent

the

(Schutz,

a smoothing

pass.

of an actual

of 1990).

function

the function and forming The

values

observation

of the nearest

value of the residuals in the bin. of both the smoothing function

the processes adopted to carry forming unbiased normal point

in the

The and

out stage (a) of the process; methods bin means are developed by Sinclair

proceedings)

Development

of the

During the observation for all events detected appearing example gross

Laser

of 2 stages;

in bins

the mean epoch resulting normal predicted

was agreed

from a predicted orbit and subtracting and (b) rejecting noise and outliers

residuals

the

format

1990 Satellite

as a track of such

noise

events

Smoothing

Function.

of a satellite pass, differences from within the range gate and displayed I

of correlated

a plot

is given

is carried

out

values in Figure visually

among

the

1, which using 9-19 '

the

the predicted in realtime,

randomly shows plot,

distributed

a pass and

range are computed true satellite returns

the

of ERS-1. subsequent

noise

events.

The

rejection

set

An

of satellite

of

range

measurements

At an early to use

the

stage

and

events

are passed

development

of the

observations

to s61ve

the

along-track

to orbits

similarly

to the

preprocessing for

error,

next

stage

software,

corrections

in order

of processing.

the

to the

to improve

decision

was

predicted

orbit,

subsequent

correction

computed

corrections

orbit

during

can be expressed

by numerical

to the

IRVs

integration

would

not

the

solution

as an orbital

iteration

and

using

be applicable

time-bias

process,

computed and stored. For each observational tion to compute the instantaneous position

and

for improving

the prediction process an orbit in the form and velocities at 1-minute time intervals

generated

by numerical

respect to the true analytically-derived equinox station the

equator orbit,

IRVs

as starting

to orbits

computed

the

derived

subsequent

of geocentric and in units

along-track predictions.

rectangular of Mm and

satellite Mm/day,

epoch, we use 8th order Lagrangean r and velocity v of the satellite. Let

predicted

topocentric

predicted range of the observed

predicted

orbit and

2.2

Derivatives.

Partial

(a) Let Then

radial

the along-track the

displaced

using

IRVs,

the

coordinates

partial

components

of this range,

interpolar = (x,y,z),

are

given

=OR v_(x

+ v_AT-

ZT = z -- Zs

,

topocentric vector. We thus form and in order to solve for corrections

of range

with

respect

of the predicted

to the

along-track,

orbit.

be AT.

by

xs) 2 + (y + v_AT-

xs) + v_(y

or approximately

9-20

with

epoch of the IRVs. For an to the mean equator and

ys) 2 + (z + vzAT--

Zs) 2

Then

R.._-_-_

The for

are

displacement

or time-bias

R is given

R 2 = (x + vzAT-

satellite

derivatives of the

cois

we compute the coordinates of the observing station coordinates be s = (zs, y,, z,). Then

YT = Y - Ys,

is the magnitude from the predicted

displacement range

of the

= X '--:-Xs,

we require

across-track

and

epoch, Let the

coordinates XT

the

integration

and mean equinox at 0hrs UT of the the coordinates are given with respect

of date. At the observational in the same reference frame.

and the difference

and

prediction

and v = (v,, vu, vz), and let the magnitudes of these vectors be r and v respectively. reference frame used is different depending upon the source of the orbit being considered; an orbit

taken

Thus it was decided to express orbital corrections as along-track, across-track displa_:ements to the predicted position. Such corlections are then readily applied

geocentric

During ordinates

noise

the possible use of an arbitrary polynomial for the smoothing function was However, corrections to the predicted orbital elements of the satellite would

to apply

analytically. and radial to the

of the

to monitor

accuracy. Thus not considered. values,

some

pass-by-pass

in particular

be difficult

and

+ vyAT-

y,) + v_(z

+ vzAT-

z,),

the to

OR OAT (b) Let The

the

across-track

across-track

= (V=ZT + vyyT

displacement

direction

+ VzZT)/R

be A Q

is rXV

= (l,m,n)

, say.

r-v

So_

l = (yvz- zv_)/r_

n = (xv v - yvx)/rv So the

displaced

range

R is given

n 2 = (x + IAQ

-

by

z,) 2 + (y + mZlQ

-

y,)2

+ (z + nAQ

-

z,) 2

So_

OR RcOAQ

= l(x + lAQ-

z,)

+ m(y

+ ,nAQ

- y,) + n(z

+ nAQ

-

or approximately OR OAQ (c) Let

the

radial

displacement

So the

displaced

coordinates

So the

displaced

range

-

(l XT + m YT A- n ZT)/R

be Ar of the

R is given

satellite

are

x + zAr/r,

etc.

by zAr

R2 = (_ + _____5_ _ x_)_+ (y + yAr __ _ y,12+ (z + --r r

Then

as before

we calculate

r

the

approximate

partial

derivative

OR

o3r = (__r + yyr + z zr)/(_R) 9-21

-

z,) 2

z,)),

2.3

Solutions.

In practice the radial not

it was found that the across-track correction was always highly correlated correction, and in many cases the solution was indeterminate. Hence both

be solved-for,

for a radial found

and

that

simple

in general

decided

which

to suppress

would

along-track

all the error

derivatives

parameters

only,

constant

absorb

for a particular pass time rates of change partial

so it was

correction

and

of range

with

and

solution the

corrections that

to the

rates

of the constant

to the predicted

of change

terms

necessary

were

to solve

required

for only

to remove

4 of the

all trends

from

possible

the

Initially a scheme was devised to automatically among the 4 unknowns _h, _,/_,/_, and to supress determinate

solution,

consistent

with obtaining

their The

of these

by t and

but

t 2, where

occasionally

all six

residuals.

check for very high (>0.999) correlations any one or two of them in order to obtain a flat

track

of residuals.

accelerations. parameters,

imposed a-priori standard errors of these the program to solve for all six unknowns.

In order satellite

did not

accelerations

six parameters,

determined for 2h and

We thus allowed

also

of the pass. Such a definition the unknowns. It was usually

showed that the values order of 0.1 ms/minute and

orbit

displacements.and T, T, T, R, R,/_.

and

solve

It was

to be determined

by mutiplication

observational

and

error.

the set of parameters

time t is the epoch of each observation relative to the mid-tlme for the origin of t is optimum in reducing correlations between found

for across-track,

across-track

from along-track and radial We denote these parameters

respect

from those

the

absorb

radial

in the orbits,

should be selected and accelerations.

were formed

thus

with could

for these 4 parameters 1.0 cm/minute for R,

However

were always and of similar

a

experience

quite small, of the magnitude for the

magnitudes

to iterate the solution, at each stage we replace the predicted by the displaced coordinates as determined from the previous

upon

coordinates solution.

the

4

of the So for

example

x = x + vzAT and

similarly

outliers

for y and

of magnitude

A selection

of some

z.

It was

greater plots

than

check

a single-factor

that

the

analysis

that

these

track

removed

residuals

residuals

test

is indeed

on

the

were

at each

is given

are readily

of residuals

of variance

+ xAr/r

4 or 5 iterations

3 x rms were

of observational

all trends have been removed; the Herstmonceux algorithm. As a final

found

+ IAQ

usually

2, where

to form

statistically

residuals.

This

where

stage.

in Figure used

sufficient,

flat, test

it is clear

that

points

using

normal

we have checks

introduced

for significant

differences between the means of residuals grouped into normal point bins; any differences indicate that not all trends have been removed from the residuals, and a warning is given that

normal

anticipated during

points that

this

should test

not may

be formed prove

useful

until

the

for the

a pass.

9-22

cause detection

of the

problem

of system

is traced.

calibration

It is changes

3. The

Conclusion. scheme

was

found

to be determinate

for a wide

range

of satellites

and

pass

durations

and quality of predictions, and has been adopted at Herstmonceux. An earlier version of the software is also in use at several European SLR stations. The deduced values of time bias are used has been

to good effect for improving subsequent used to generate time bias values from

stations,

as described

4.

by Wood

and

Gibbs

(these

predictions, the quicklook

and additionally the software observations from other SLR

proceedings).

References.

Gaignebet, on Laser Schutz, University

J. (Ed) Ranging B.E.

(Ed)

of Texas

1985.

Recommendation

Instrumentation. 1990,

84A,

CERGA,

April

Satellite

at Austin,

USA.

Laser

9-23

Normal

Points,

in Proc.

5th Int.

Workshop

Center

for Space

Research,

France. Ranging

Newsletter,

-

0 0

la0 °,,q

!

°



0 0 0

0 0

(su)

0

I_?npTsa_ 9-24

0 O.

a_ue_

i

0 0 O_ I

.4

.2 _° o" .

o

.

° o.

"..

.

° .





....



o



.

o

".•



°* • °. °



.

°

.

• ° .:

°.



• .,'._

"t......°

• ::.-...._.-., ... • .......

o : ...



• °

.-._ .'...

• ...

". ° .° o ° °° %. ; ; ° ..

.

...



° .o

.



.

:'." . .'_ ., ...; ...

.

• °

. o.." •

0 .. - .....

,..°_

": _ '.:. -'°.."..'.°.. • " "." :: • '

w

""



o_

.o

.L•

. •

.."

.'...._

-"'.A,

t*

.:-_ .°



-°.':.

.° •

D •

'

"."

" °"

'T"x;

".'" o"._ _ .,_''_." q : "_'.. "."::°._ ": ."....'" .'...,."t ..... "' " " ". " • "" '" ." -" %'.-"F" :_



"



"







..o:

' "'" .:" •

°

i

"

-.2

-.4

I

I

,

I

I

I

3550

I

I

I

3600

,

,

I

,

3650 (secs) ERA

Time

,

,

,

1 i

j

,

281.

I ,._____

3750

' I''

r-

4

_ I

3'700 (SPAD+HP)

'-]_--_-_-_j

i

7/

i

2

_, -r.

-..

=

..

°

.

,.



. •-

,



°o

°



.

°°



. o.,:..-'.:.'o:::.-.[...:..¢ ;.... • :; _._,.._; • ...I.._...: -..-,..



:.....

* o

.

,



......-.. ."....

..

_ ,'" ". "".-.:_'",'>'.."_"_,(,:. ':,;')!.'.'._",'L_ e"7"" _._ ;'.:.:.

I

°



,

_%.e**• •• . * •.

_

•°

"_

*" • "° . • .. •

•.'-";'": :,.4 *."

_0 •

.

;

t

"

°..



w,.,

.,...

s,

*t:..'_*-.%.:.'e?_._

.-;_..2.._._. "'" * _l"

.. ."

"%*¢'*

;_¢_:.*..o::_,..*-.

.,. r.. ,':>_.,-,_,._:._j,-_-.'_.._?; -,__,._.#T_, _._'£_':."",_,_".;:.:."_ L: .,".-_,• • .._.::,." "*'-(': .V :"' ":". _:'._-;",,.:;_'. • • ' "2_

|

-_

__.

-.4

_.o_,.... .,

i

,



• °







,,. ."

* .,..' . _

,

s

,.-.o.,

*



.

"

,r

°

.%



.

,T



°

m!



'.L"..,,

_' • _

:* "" • "...

;'_.. " _. "

. "

",._ ":

""

• _ _.:; ;.,

.'.V:2.• " ".



*,

' "

: i ij

4200

Figure

4400

2. _ange

i ,_'_ 4600

residuals

4800 Time

from

: ,I ; ,. _ i ____1 , 5000 5200 5400 (s) L-_G 236

a pass 9-25

of ERS-1

l

I

"

(b) :.L_J

t

••" •

(a)

_ i 5600

,. ; ;

,_,.__

5800

a x/d Lageos

6000

(b).

Poisson

Filtering

of

Laser

Ranging

Data

by

McDonald

R. L Ricklefs and P. J. Shelus Observatory, University of Texas, Austin, TX 78712

The filtering of data in a high noise, low signal strengthenvironment is a situation encountered routinely in lunar laser ranging (LLR) and, to a lesser extent, in artificial satellite laser ranging (SLR). The use of Poisson statistics as one of the tools for filteringLLR data is described first in a historical context. The more recent application of this statistical technique to noisy SLR data is also described.

Introduction Routine LLR operations began at McDonald Observatory in August, 1969 as a part of the NASA Apollo Lunar Ranging Experiment (LURE), using the newly-commissioned 2.7-m astronomical telescope to range to the retro-reflector array placed upon the lunar surface by the Apollo 11 astronauts. During the next several years, additional arrays were left on the moon during the Apollo 14 and Apollo 15 missions, and by two Soviet unmanned soft landing vehicles. Routine LLR operations continued on the 2.7-m telescope until the mid-1980's when these operations were transferred to.the dedicated 0.76-m aperture McDonald Laser Ranging Station (MLRS). It is important to realize that the return signal strength ratio, neglecting all parameters except distance, for a near-Earth artificial satellite and the Moon is something like 3 x 1012. Thus, it is more than a trillion times more difficult to range to the Moon than it is to range to, say, Ajisai. In the extremely low signal strength environment of LLR it was (and continues to be) necessary to operate largely at the single photon-electron level. However, this means that a very large number of "volunteer" photons from any number of background noise sources may trigger the detection electronics, even though very narrow spatial, spectral and temporal f'flters are routinely employed in the receive package. Since all incoming photons are indistinguishable from one another, identifying the valid returns from the moon has been a difficult task. From this experiment's inception, the reliance on statistical f'fltering methods, in addition to the physical ones, had been always assumed. The application of Poisson statistics to the LLR data filtering problem provided an effective tool for dealing with high-noise, low signal data. The technique has now been expanded and adapted to handle marginal SLR data as well and has proven to be especially effective when applied to data from single or low-multiple photo-electron systems during daylight passes and data from all stations when they are ranging more difficult targets like Etalon- 1, Etalon-2, and MP-2. Figures 1 and 2 show samples of noisy data, displayed by MLRS onsite software, for which the Poisson technique is required.

Background Let us consider the mathematical concept of the Poisson distribution, which gives the probability that a certain integer number, k, of unit rate random events from a population, n, will occur in a given interval of time, x, within some total interval of time. The operative equation is: 9-26

P(k) -

x[k] e-X [k]!

where [k] is the largest integer less than k. For x large enough, this function has a bell shape comprised of many rectangular steps. Multiplying the computed probability by the number of bins in the entire sample time gives the number of bins which contain exactly k events. In a descriptivesense,Poisson statistics statesthatffone has "n" observationsof some independent variabledefined over some range of "m" bins of uniform width, and ifthose observationsarc random, one can compute, in a statistical sense,how many of those bins will contain exactly k observations. For example, for some specificexperiment of,say, 16 independent observationsof some value,one might compute from theabove equation that 3 of the bins will be empty, 4 of them will contain exactly 1 observation,3 of them willcontain exactly 2 observations,2 of them willcontain exactly3 observations,and no bins will contain 4 or more observations. Then, ifone looks at the actualhistogram of measurements and findsthatone of the bins contains8 of the 16 observations,one might safelyassume thatthose observationsare indeed not from a random distribution and that something systematic has occurred. In the case of LLR or SLR observations,where the measurement is an (o-c)range residual,itis assumed thatthe bin (or bins) containing a statistical excess of observationsarc probably valid returnsfrom the targetor theresultof some systematiceffectin theranging equipment. As another example, figure3 represents the statistics from the lunar data in figure I. The predictionsindicatethatthere should be no bins with more than 21 points. One bins has 22 points, which is probably not significant, but thereis another bin with 47 points,which issignificant. That bin contains the lunardata. This technique is routinely applied to thc filtering of McDonald Observatory LLR data and is fully described in "Laser observations of the Moon: Identification and construction of normal points for 1969-1971 by Abbot, Shclus, Mulholland and Silverberg (Astron. J., Vol. 78, No. 8, pp 784-793, October, 1973).

Current

state

of

affairs

In the routine application of the Poisson technique to data filterin.g it must be noted that, in order to have the best possible chance of isolating signal from norse, one should use the narrowest bin width possible, keeping in mind at all times, of course, the quantity being measured. Narrow bins will maximize the signal to noise ratio since, in a narrow-bin environment, a relatively small number of noise photons will be found in any one bin. Further, it is absolutely required that the data be flattened before a histogram is formed, i.e., trends must be removed from the data as much as is possible, so that signal will not "spill over" into many adjacent bins and, again, be lost in the "noise". This problem has been addressed in a straightforward way in the current data screening implementation at the MI_S, as shown in figure 4. All bins are scanned sequentially through a predetermined range of slopes, the plan being that the signal will appear at a significant peak at a slope that matches the inherent slope of the data. This technique is especially important for LLR data when signal strength is low and for SLR data when orbits arc not well known. Of course, in altering the slope of the data, one must take care to exclude, or compensate for, those bins which will be incomplete due to clipping. In its present implementation at the MLRS, the Poisson screening software also allows multiple segments of a pass, or a run, to be treated separately. This handles the natural change in residual slope over an extended observation session. 9-27

Although personnel atthe MLRS arc now applying the Poisson filtering technique for both lunar and artificial satellite data,the LLR and SLR screeningsystems are somewhat different. Both systems provide user interactionto allow the selectionof the parameters of the Poisson fit,placement of the time bins, maximum and minimum slopes, and a maximum 8nd minimum residualfor data to be accepted as "real'_.For lunar data, wc employ fairlystrict limitson acceptableslopesand allowable residuals.Earth rotationand polar motion are major contributorsto slope and residualfor lunar data, so the earth orientationpredictionsarekept as freshas possible.Further,thereisconsiderablelatitude in selectingtime bins to reflecttheobserved clumping of NR,RS LLR data. Ultimately,the Poisson filteris reliedon to provide the LLR data compression, i.e.,normal pointing, softwarewith a setof acceptabledata. For the SLR case, the Poisson algorithms serve as a preliminary filter and suggest to the polynomial fitting algorithms in the normal pointing packages a set of data with which to start. The polynomial fit, in its job, can then return previously rejected data back into the fit or remove data suggested to it by the Poisson filter, if either of these tasks arc warranted. Also, the polynomial fit, working on the entire pass, can pull in data that might be missed or misinterpreted when the Poisson filter was working on only shorter segments of a pass. The independent time segments, into which an SLR pass is broken, allow the software to follow changes in slope during a pass and minimize the effect of anomalous distributions of data in short segments. We are examining the desirability of having several software checks that require the slopes and average O-C residuals to be smoothly varying over the pass. However, in the cases where there are breaks in the data and other unusual distributions, this may not be practical.

Results In the case of LLR data,the presentlyimplemented Poisson filtering technique works well,especiallywith a bitof interactivehelp from the user,although,in marginal cases,it would be preferableto use allruns taken over a lunar pass to aid in the filtering.This particularextension isnow being examined. As time permits we arc alsopursuing more elaboratefiltering techniques such as jack-knifingand bootstrappingto see ifthey hold merit. In the case of SLR, the technique has been used successfully at MLRS for about a year now. In that time our experience has been that the combination of the Poisson filter and the polynomial fit produces good results virtually all of the time. The marginal cases can usually be satisfactorily processed, after the fact, by altering some of the parameter of the Poisson fit or by eliminating from consideration extremely noisy parts of the pass. Of course, most of the anomalies usually occur with weak passes, especially Etalon- 1, Etalon2 and ERS-1. The Poisson filter algorithms are being upgraded so that each satellite uses a parameter set tailored to it. Tests indicate that this will allow successful batch processing of virtually all passes that previously required special handling. Also, an interactive tool similar to the Lunar Data Editor is being developed to help the observing crews to quickly recover passes that were improperly processed. The Laser Tracking Network operated by the Bendix Field Engineering Corporation will soon install the MLRS "batch" Poisson code into their processing software.

9-28

Conclusion The Poisson filtering technique has proven itself to be an extremely valuable tool for both LLR and SLR operations at the MLRS. This technique is being optimized for each ranging target, and in the future provision will be made for routine, near-real-time observing crew interaction with the filter. In addition, new filtering techniques such as jack-knifing and bootstrapping will be evaluated as to suitability for laser ranging data filtering.

9-29

rd.p File

Edit

Graph

Iot

Enlarge

Help

OC Residuals O-C

(rmc>x

x

x

1,4_ x

x 1._

x

x

1._.

x

X

i_

X X

1.0_

X

x _x x_ xX Xx

x x

x

1.1'.

T_ x

>:_

)4,x x

_x

_< xx

x

X x

X

)4z_xX

X xx

.

x

_< x x

x

_x_

x

x

x

_

x

x x

X

0._

x

X

X X x

x

x_x x x

X

XKX

X

x

x

x XX

XXx

_x

X

x

x

X X

xx

X X

x

X

X

X

xx

x

x

x

x

xx

X

X xx

xxxXx

x xX

0.R

X

XX

X>

X

X

X

X

X_x

_

x

x

X xKX

x>:

x

XX

x X

X

X

XX

X

0.4_

X _

XXxX

x

0._

x

X

x x

_

x

x x

x

xX

x

0.05 -0._

>o


10

W

L..

I1)

E = z

5

0 0

10

Figure 1: Histogram of the 'semitrain 1991. The session lasted 15 minutes

20

30

Range

Residuals

ranges'

to APOLLO

10-16

40

50

60

the night

of Nov. 25th

[ns]

15 during

Fixed Station Upgrades/Developments

N94"15601 MATERA

LASER

RANGING

OBSERVATORY

Thomas

Varghese,

Winfield

(MLRO),

Decker,

AN

Henry

OVERVIEW

Crooks

Allied-Signal Aerospace Company Bendix/CDSLR Network Seabrook,

Maryland U.S.A.

Giuseppe Agenzia

Spaziale

20706

Bianco

Italiana/Centro Matera,

di Geodesia

Spaziale

Italy

Abstracl;: The

Agenzia

Bendix Company Centro

Italiana

Engineering

(ASAC)

telescope

future,

with

with

a specialized

on a 1.5

multiple

dome

which

The

fixed

locating

the

following:

and the operations concrete

for

pad for

minimizing power

the

laser

fitting

The

is currently

scaling

or changes

system

with

system.

Provisions

with

Signal

a variety

for the

calls

for

the

astronomical

of experiments

Three

the

Aerospace

observatory

contract

Cassegrain

ranging.

console. added of

The

EMI,

transmit

focal

planes,

for viz.

the

be highly

It is conceived specifications

also made

to adapt

course

for

areas

for

telescope/dome and mount

Provisions maximum

providing

rest on will be in

cleanliness

an ergonomic

laboratory.

modular

and

adaptable

to be a highly to any currently

and accommodate

of design

ii-i

tables

obtaining and

to be provided

into four

stability.

multipurpose to

facilities optics,

optical

for

optics,

designed

performance are

building

be partitioned

mechanical

in technology.

during

will

transmit/receive

to a state-of-the-art

superior

are of significance

facility

effects

and

environment system

to support

on laser

laser,

a common high

The

afocal

will be part of the

observatory

enclosure, place

negotiation

Allied

laser ranging Italy.

meter

ports

emphasis

under

of the

and Nasmyth will be available for these experiments. system will be protected from dust and turbulence using

by ASI.

for

(BFEC)

in Matera,

based

primary

Cassegrain, Coude, The open,telescope

is currently

a state-of-the-art

Spaziale,

of a system

quality

(ASI)

Corporation

to build

di Geodesia

delivery the

Spaziale

Field

and integration.

for

automated operational

changes

that

MLRO Objective:

:An Overview

Build a state-of-the-art, multi-purpose, observatory in Matera, Italy

laser ranging

General Features I

Day and night automated

ranging

capability

on all CCR-equipped

satellites

(> 400 km) and the moon •

Application

of state-of-the-art



Real time calibration



Multicolor

ranging



Advanced

computing



Computerized

documentation,

expert

and hypermedia

system

technology

and interleaved

in all sub-systems

multi-satellite

tracking

and data analysis

tools

capability environment

with features

such as relational

text

System Specifications LAGEOS

ranging" Single

shot precision

• --5 mm

.Normal point precision Low orbit satellite

(ERS-1,

Single Lunar ranging

• Single shot precision

Single

i

Range

Starlette,

shot precision

Real time calibration/ground

accuracy:

• < 1 mm

• 3 - 5 mm

• -1 cm

ranging

shot precision

etc.) •



• =10 Hz

Computer

controlled

Modularity

seeding

for energy

Laser firing stability

beam divergence scaling

/ multi-wavelength

: better than 20 nsec

11-3

generation

MLRO

:An Overview Control

D

Software

D

Automatic

recording

Automated

star calibration,

in high level languages

Computer-assisted t

Distributed

D

Advanced

System

Automated

of all operational

optical

parameters

mount modelling alignment

data acquisition,

and verification

processing,

computing/control

environment

system/sub-system

simulation

Transmit/Receive D

Detection"

and debugging

Electronics

High Q.E, high speed MCP-PMT/APD

Signal processing

:

Speed

: 1 - 5 GHz

Dynamic Jitter Event

and control

range

: 20

:

ca

1 ,//

/ !

co t'i__J l--

t..la

o 0

.la o

._I ,-4 u c _0 o

......

k

O*J u_

12-23

Decrease

of MTLR,C #1

r.m.s

r.m.s in ns

r.m.s

in cm 10

0,5

[]

r,m.s in n8

I

9

r.m.s in cm

I

8

0,4

7 6

0,3

5 4

0,2 []

3

[]

2

0,1

1 I

old

I

new Laser

Fig. 4: Decrease

R.M.S, r.m,_,

new Start

of the

single

0

I

Receiverpa(-k

shot

r.m.s,

of MTLRS#1

during

SPAD

upgrade

Components

ps

In

I I 400

300

100 200

65 i

3O

lO0 20 0

0

I

La=er

Fig. 5. Single

START

shot

.....

_

I

Recelverpack.

r.m.s

before 12-24

STOP

and after

Counter

upgrading

typical

Lageos

nighttime

pass

(PMT)

before

upgrading

_.._

"T

_-- -.q8._._4

ns

Pre Calibration

Post Calibration •

I_V_.._._.

_"

typical

--

Fig.

....

• • ._

:.

-,

Lageos

.._

_:,_:_,.

':_,-. - _ L,-_.

_.'.v.,_-:,-

_ "...:

.



:

day/nighttime

_

_:_]

,.

.

pass

.,..,

Lageos

Pass

before

and

after 12-25

_._:_.-.

(SPAD)

-e@eS@@r,s

-

..

upgrading

after

upgrading

N94-15611 The

new P.

MTLRS_I

SPERBER,

INSTITUT

Receiving

U.

FI_R

HESSELS,

R.

ANGEWANDTE

MOTZ,

GEODASIE

FUNDAMENTALSTATION DER

System

WETTZELL

FORSCHUNGSGRUPPE

SATELLITENGEODASIE

D-8493

K()TZTING

H.-M. GI.iTHER HOECHST AG, MATERIALFORSCHUNG D-6230 FRANKFURT AM MAIN 80

Abstract. In this

report

Transportable

we are giving Laser

and

a Single

and

night

Photon

ranging

a detailed

Ranging

System

Avalanche

capability

description

Diode to all

of the

MTLRS_I (SPAD)

satellltes

new receiving

consisting aa single

with

cm

system

of a spectral photon

of the

time

detector.

and

The

German field

system

Modular

of view gives

filter

full

day

accuracy.

1. Introduction In order to improve the ranging accuracy and the sensitivity of the German Modular Transportable Laser Ranging System MTLRS-1 a major system upgrade was scheduled and performed from Dec. 1990 to July 1991 (P. Sperber et.al•). mitting part (Laser and Start Diode) (P. Sperber was the main point of the upgrade. The receiving

2.

system



Filter

(Frequency,



Single

Photon

Receiving To extract

mainly

has two parts, Range,

Together with the exchange of the whole transet.al.) the installation of a new receiving system

which will be shown

in detail

Field of View)

Detector.

System reflected

in daylight,

This filtering

photons

fr6m the satellite

from noise photons

which are entering

filter eqmpment

has to be installed

in SLR systems.

a powerful

in modern

systems

is realized

the telescope

in three steps



Range (time interval) filter: A rotating shutter is commonly used to supress straylight during laser firing and to open the detector at the time, the echo from the satellite is expected.



Field of View: A pinhole ensures, telescope axis can hit the detector.



Spectral just

Filter:

The most

effective

light of the well defined

that

only light

filtering

laserwavelength

process

coming

from a small

is the filtering

angle

in frequency

the

which allows

to propagate• Typeset

12-26

around

by .AA45-TEX

The new receiver package of MTLRS#1 requests, deliverd by Technisch Physische The setup is shown

is a very compact, field qualified realization Dienste (TPD), Delft, Netherlands.

of this three

in Fig. 1.

The light from the telescope is focussed to a field of view pinhole, which a motor driven micrometer from 0 to about 30 millidegrees. Connected there is a rotating shutter, synchronized to the predicted system 10 times per second for about 3 ms and mainly straylight during laser firing. The spectral filtering is done by an echelle

grating

is remote adjustable to the field of view

with filter

range window, which opens the receiving protects the single photon detector from

(efficiency:

55 %) used in 7 th order,

which

gives a bandwidth of 0.2 - 0.65 nm depending on the opening of the filed of view pinhole. The grating has a surface of almost 100 mm top-top in air, therefore the range difference of the beams over the grating surface is not neglegible. To compensate this range difference the grating is used a second time in opposite order. At the end of the receiving package a lens is generating a 8 mm diameter, 3.

Single

parallel

beam

with excellent

optical

quality.

PhotonDetector

To see the full advantage

of the short

photon detector with high accuracy receivers are used in the moment: •

Microchannel



Avalanche

laser pulses

is necessary.

and of the errorfree

In the world

of SLR

receiving two types

package of high

a single accuracy

Plates Diodes.

We decided to install a single photon avalanche by the Czech Technical University in our system. •

High ruggedness

under



High Quantum



Low jitter



Low operating

voltage



High dynamic

range and

neglectable



No additional

electronics

(CFD)

diode (SPAD) (I. Prochazka The reasons for this decision

et al.) were:

developed

field conditions,

Efficiency

(20 %)

(35 ps) (25 V) influence

between

of signal

detector

strength

to the range

measurement

and counter.

!

The housing

for the diode



Adjustable



Possihlity



Protection



Flexible

has to meet some demands:

in three dimensions to cool the diode against

humidity

in the pm

below

range.

-10 ° C.

(condensation).

for usage of different

types

of diodes.

The cooling of the diode is necessary because of the high termal noise (some 100 kHz) at room temperature. The temperature of-30°C, which is possible with this setup will decrease the noise far below 100 kHz (Fig. 2) and therefor make the search for satellites at night more easy. To meet •

all this points,

adjustement

we designed

accuracy

a new housing

in each dimension:

12-27

(Fig.

5 - 10pro

3) with the following

specifications:



input

beam:

5-10 mm diameter,



optics:



field of view in our system



copling:



good

effective

focal length:

to -30°C

mechanical

with stability

parallel

11 mm

(100 pm diode):

peltier under

The housing is made from aluminium, of the diode to the housing is DELRIN.

cooler

and

30 mrad water

field conditions

(5°C) (temperature

the inner moving

part

change,

which

makes

humidity) the

termal

isolation

For the fixing of the diode a very special material with high mechanical stability, high thermal conductivity and good electrical isolation was necessary. To get a thermal equilibrium even at high outside temperatures we decided not only to cool the diode, but the whole fixing plate with powerful peltiers (the hot side of the peltier is water cooled). As some diodes (mainly the diode we use) needs an electrical isolator around it, a material with the mentioned specifications was necessary. In cooperation with the Material Research D,epartement of Hoechst AG, a new ceramic materialAluminium Nitride- was selected. The important properties of this are its high thermal conductivity of 170 W/ink (comparable to pure Aluminium) and its dielectric constant of about 8.5 (good insulator). The powder, together with a sintering aid, are prepared for pressing. The cold isostatic pressed parts, in the form of rough discs, are then turned to improve their finish and a hole is bored in the centre. The binder is then burned out. The samples are sintered at 1840°C for 3 h under a nitrogen atmosphere in a special graphite crucible. After the heat treatment the ceramic parts were finished. They were grond with diamond wheels in all directions to get the high dimensional accuracy and surface smoothness for installation of the measuring diode and for the mounting of the peltier elements. The specifications of the material are given in table 1. Due to the flexibility and the good experience now also used at the new Wettzeli Laser Ranging 3.

with this housing in the last field campaign it is System (WLRS) for the tests of different diodes.

Summary

The new receiving system of MTLRS# 1, consisting of a filter package for field of view, time and spectral filtering and a single photon avalanche diode as detector gives the possibility of SLR with cm accuracy during night and day. The whole system is build very modular and flexible, so easy exchange or usage in other SLR systems is possible. All parts are designed and changes in field an adjustement In the last field campaign

manufactured in a way, is not necessary.

the new receiver

12-28

proved

that

even

to work very

under

reliable

extreme

environmental

and satisfactory.

REFERENCES

P. Sperber, Upgrade,

L. Amberg, 199_.

L. Beyer,

G. Blensid,

U. Hessels,

P. Sperber,

U.

R.

Motz,

W. Beck,

in this

B. Sopko,

S. 219,

Pl_otodlode

I. Prochazka,

Hessels, K.

Kamal,

12-29

R. Motz,

Proceedings, Based

in this

The

Detector

Proceedings,

_ew P*ekage

MTLR5 for

Res_lta

TranJmltting Centimeter

o] the

MTl,

,.qyetem, Sstellite

qstltl

1992.

Ran#iBg,.

@

C_ CO L_ r_

(1) >

U _J ri-

ft _.I l--

r

4_ 0

I--

o_ z

0 _0 _,J

_0 0 C 0 .C 0

O_ LL

12-30

SPAD Dark Count Rate versus Temperature Dark

Count

Rate

[kHz]

1000

3O

Centigrades Fig.

-

Density Termal Specific Flexural Electrical Dielectric

Table

1

2

SPAD Dark Temperature

Count (2.5

Rate versus V above Break)

3,3 glcm 3 170 W/m" K 738 Jlkg K 300 - 400 MPa > 1014 Ohm cm

Conductivity Heat Strength Resistivity Constant

Specification

8,5

of

Aluminium

12-31

Nitride

-

9

tl

'r.-,: , C"

i,.

i r i

'

I t_

_t

_"0 0 -r-I tm

.C 0 C

C 0 0 rO-

C_O C 09

if

.C _a 4-I 0 O% C \

"\

"c

T_I

0 mr-

q

CO

_, =

12-32

0% -H LL

N94-15612 The new MTLRS

Transmitting

System

P. SPERBER, U. HESSELS, R. MOTZ INSTITUT FOR ANGEWANDTE GEOD)_SIE FUNDAMENTALSTATION WETTZELL D-8493 K6TZTING GERMANY W. BEEK, J.W. OFFIERSKI, C. VAN Es DELFT UNIVERSITY OF TECHNOLOGY KOOTWIJK OBSERVATORY FOR SATELLITE GEODESY POSTBOX 581, 7300 AN APELDOORN THE NETHERLANDS

Abstract This paper presents a detailed description about the new transmitting system of the Modular Transportable Laser Ranging Systems MTLRS-1/2. A simplified theory of the Self Filtering Unstable Resonator (SFUR) is explaned. Laser design details are discussed concerning the extreme environmental conditions in which these mobile systems are operating. Details are given concerning the detector are necessary

new avalanche parts in order

START detector. The new SFUR laser and START to bring both mobile systems towards 1 cm ranging

accuracy. 1. Introduction Since

1984

the

two

European

Modular

Transportable

Laser

Ranging

Systems

MTLRS-1,

operated by the Institute for Applied Geodesy (IfAG, Germany), and MTLRS-2, operated by the Delft University of Technology (DUT, The Netherlands), have supplied the international network with laser ranging data from sites in Europe, CIS (former USSR) and North America. To improve the carded out a major upgrade transmitting system also 2. The

ranging accuracy from 5 cm to the 1 cm level, both systems in 1991. The main part of this upgrade was the exchange of the

system consisting of the Laser and upgraded the receiver package [1].

the

Start

Detector

(The

German

MTLRS-1

Laser

The accuracy and the number of returns produced by a satellite laser ranging system is mainly influenced by the duration and energy of the emitted laser pulses. To improve these two points, the original Nd:YAP laser (370 ps FWHM pulse-duration, 10 mJ pulse-energy at 539 nm) was exchanged for a new Nd:YAG laser (30 ps FWHM pulse-duration, 30 mJ pulse-energy at 532 nm). To realize this high demands with a simple and reliable optical configuration, a new resonator setup, the Self Filtering Unstable Resonator (SFUR) [2,3] was used. Together with a simple optical setup, this resonator has the advantage to be less sensitive to optical adjustments and operates minimizing the danger of destroying optical levels. 12-33

with a very components

homogeneous spatial pulse-shape, at high resonator energy output

2.1 Simplified SFUR Theory Fig.

1 shows

the principle

[4]

of the SFUR

M2

resonator.

S.F.U.R.

RESONATOR

aperture

Fth

R2-"

M1 R1

d L

L1

/.2 LO

Fig. Two

mirrors

1

M 1 and M2 are

forming

an instable,

confocal

resonator:

R l + R 2 = 2 "Lo

RI and

R 2 are

the

(concave)

radii

of the

mirrors

and

Lo the

optical'

resonator

length.

This

resonator configuration generates a spot in the resonator, which usually destroys optical elements. If a small pinhole is placed in this focal point this problem disappears and some major advantages can be seen. The diameter d of the aperture has to be chosen in a way, that only the central lobe of the Airy pattern (generated by mirror M2) propagates to M t.

d=

(/2."focal The

length

mode,

Gaussian

of M 2, X: laser

generated

shape,

with

by zero

x/ 2.44

"f

maximum

of

" _.

wavelength)

the

central

intensity

at radius

the

Airy

pattern

has

a smooth,

nearly

ro where

R 1

ro = -

M

• d-

(M

2

= -

--

)

,R2

If the parameters of the resonator are correct, the mode is exactly filling the whole of the active medium. Therefore nearly all the energy stored in the rod is converted laser pulse. configurations.

The outcoupled The aperture

volume into the

pulse-energy is much higher compared to usual resonator (typical diameter d< 1 mm) gives a strong spatial filtering at

12-34

each round trip smoothing the energy distribution in the beam and avoiding exact analysis of the SFUR configuration the influence of the thermal lens in the laser rod has to be taken into account.

hot spots. In an (7"u,: focal length)

Lo = L l +/,2

gl

Rl L0

LI

&=l

g

= 2 "gl

M

(L 1 and

L 2 are

optimum diameter

the

distances

= g

of M1 and

position Lp (distance D of the laser beam

I

-

1

M 2 to the virtual

_

Lp

D

_[g2

-

position

from M_) of the aperture, are given by:

1

d

-

"g2

:

--- 1.5

4.88

.M

1

÷

/?2

• Lp

_/g2

the

-

2 "gl

• _. "(

The laser electronics,

lens).

The

d of the aperture

and

1

1

R'z )

"d" .

Technical

diameter

thermal

"Lo

2

2.2

of the

u

.

--

-i -L"

Realization

system consists of four separate units, the laser an external dye-cooling unit and the control panel.

12-35

head,

the

power

supply

with

A schematicdrawing

of

the

laser

head

is shown

Resonator MI - M2 is folded over a prism PR. acousto-optic modulator AML and a bleachable (PML). The intense 30 ps long pulse is cavity dumped by the pockelscell LASER PC and two polarizers PO. Because of the high energy of the output pulse of a SFUR resonator a single stage amplification (AMPL) is sufficient to get 100 mJ per pulse in the infrared. The second Harmonic Generator SHG converts about 40%

E E

Picosecond dye (Kodak

To

HEAD OPTICAL

place,

the

SFUR

are generated with an in 1,2 Dichloroethane)

CONFIGURATION

I

*

the dust of the - Camera. controlled,

save

pulses 9740

1200

of

POWER

2.

I

0 0

of the energy into green light at 532 nm. The two wavelengths are separated by the dichroic mirror M_. The excessive use of diagnostic diodes and motor driven micrometers allows the fine adjustment laser without opening the cover. Even the spatial shape pulse is monitored by a CCD head is built on a temperature

in Fig.

- MOTOR

mm

MIKES

PCT

- POCKELS CELL TRIGGER

SD

- START

DETECTOR Fig.

For maximum mechanical reinforced invar plate.

SUPPLY

stability,

the

whole

2

laser

CONFIGURATION HIGH

COOUNG

U an < O

ELECTRONICS

VOLTAGE

I I lip08 ,. IIIII ]

[--

]

Fig.

0

.£1_1"1

0

3 3

SMPS

1100

",4

mm

3

Fig. 3 shows an inside picture of the power supply. The power supply and control electronics are also built in an isolated, temperature stabilized housing. The laser switches automatically between 380 V/50 Hz and 460V/60 Hz input, and the 3 phase power consumption is 5 KVA. Two independent control panels for the laser can be used; one for the Cabin, in use during ranging, and another one in the Cart for maintenance purposes.

12-36

Fig. 4 shows the laser temperature built in the Cart which is located

control system. The telescope together with the laser are in full sunshine in the middle of the concrete platform

(Pad). During satellite tracking this Cart is open and TEMPERATURE CONTROL SYSTEM because of extreme environ• BEAM mental conditions (-20°C to +40°C), considerable attention is given to the stability and reliability of the laser (komacel-isolation of the head Ko_,c_ ,so.so. __ _fi, -II and power supply, excellent \ _o_. soPP.v I ..,oo I mechanical stability of all optical components). The heat generated by the flashtubes in \ ELEGTRONIGS _J the secondary cooling system is exchanged to the primary cooling system by means of an internal water to water heat exchanger connected to an external water flow with Fig. 4 temperatures between 5-16°C. The laser was first tested under field conditions during the campaign of MTLRS-I in CIS (Common wealth of lndependant States - the former USSR). The experiences were excellent. /

t ,.,,.cL

The specifications

of the laser are summarized

LASER

in Fig. 5.

SPECIFICATIONS 532 nm 30 mJ 30 ps 20 pps (max) 0.4 mrad

Wavelength Pulse energy Pulse width Rep. rate Divergence Mode locking SHG Beam diameter Environment Power Voltage

active + passive KD*P 7 mm -20 °C to +40 °C 5 kVA 380/4605Nph.

Fig. 5

12-37

3. Start

Detector

To start the time interval counter with high accuracy, the optical output signal of the laser has to be converted with minimum jitter to an electrical pulse, which is used in the start channel of the counter. Until 1990 this problem was solved in the MTLRS in the conventional way: a part .of the green pulse was sent electrical pulse with an amplitude corresponding fluctuating from shot fraction discriminator. this setup To

caused

minimize

to shot. These amplitude Changes in the temporal

errors

these

in the start-channel

errors,

an optically

to a fast photo diode to the energy of the

fluctuations pulse shape

which generates laser pulse which

were compensated and strong energy

an is

by a constant fluctuations in

of the counter. triggered

avalanche

diode

[5,6]

was

the system. This diode delivers an output signal with a constant amplitude the energy of the laser pulse. The output signal is directly used as startsignal a constant fraction generator is superfluous. Furthermore, to introduce high

integrated

into

independent on for the counter, stability and the

possibility to vary the pulse energy during tracking, the start diode is placed before the amplifier (behind mirror M3 in Fig. 2) to detect the infrared light. Fig. 6 shows the output signal of the start diode which has a risetime of 300 - 400 ps and the jitter between optical and electrical pulse is less than 20 ps.

START

DETECTOR

OUTPUT

SIGNAL

+ 7.8 V_

\

_

=.=....._

5 nsidiv

Fig. In order

to get

stable

results

from

the avalanche

start

fibre together with the electronics have to be built ture of the optical bench is stabilized at 25°C.

detector

inside

the

under

laser

6 field

head

where

conditions, the

the

tempera-

4. Summary The upgrade of the transmitting part accuracy of the systems significantly systems worldwide. First tracking' results upgrade

of the receiver

[7] show, package

of MTLRS-1/2 formes and makes them again

that the system's which

is already

12-38

single

the basis to improve the ranging some of the most advanced SLR

shot accuracy

completed

is about

in MTLRS-1.

1 cm

after

the

References: [1]

The new MTLRS-1

Receiving

System,

1992

by P. Sperber,

U. Hessels,

R. Motz,

in

this Proceedings. [2]

Paper 701,

1989

by A. Bianchi,

A. Ferrario,

P.G.

Gobbi,

C. Malvicini,

G.C.

Reali,

SPIE

p. 132.

[3]

Paper WE4.

1987

[4]

Paper 1989 Proce_lings

by K.

Hamal,

V. Hraskova,

H. Jelinkova,

by C. Malvicini, A. Ferrario, G. of the seventh international Workshop

CLEO'85

Proceedings

p.82,

Gabetta, G.P. Banff, G.C. Reali, on Laser Ranging Instrumentation

p. 181. [5]

Paper

1990

Subcentimeter

by I. Prochazka, [6]

Paper 1989 international

[7]

Results

K. Hamal,

Photon

Ranging

All Solid

U. Hessels,

State

Detection

Package,

B. Sopko.

by I. Prochazka, K. Hamal, B. Sopko, Workshop on Laser Ranging Instrumentation

of the MTLRS-1

W. Etling,

Single

Upgrade R. Motz,

by P. Sperber, in this Proceedings.

12-39

Proceedings p. 219.

L. Amberg,

of

L. Beyer,

the

seventh

G. Blenski,

N94-15613 Transputer

Based

Control

System

for MTLRS

E. Vermaat, J.W. Offierski, K.H. Otten, W. Beek, C. van Es Kootwijk Observatory for Satellite Geodesy Delft University of Technology P.O. Box 581 7300 AN APELDOORN THE NETHERLANDS P. Sperber Institut ffir Angewandte Geod_ie Fundamental Station Wettzell 8493 KtTZTING GERMANY Introduction The Modular Transportable Laser Ranging Systems (MTLRS-1 and MTLRS-2) have been designed in the early eighties and have been in operation very successfully since 1984. The original design of the electronic control system was based on the philosophy of parallel processing, but these ideas could at that time only be implemented to a very limited extent. This present system utilizes two MOTOROLA 6800 8-bit processors slaved to a HP A-600 micro-computer. These processors support the telescope tracking system and the data-acquisition/formatting respectively. Nevertheless the overall design still is largely hardware oriented. Because the system is now some nine years old, aging of components increases the risk of malfunctioning and some components or units are outdated and not available anymore. The control system for MTLRS is now being re-designed completely, based on the original philosophy of parallel processing, making use of contemporary advanced electronics and processor technology [Beek, 1989]. The new design aims at the requirements for, Satellite Laser Ranging (SLR) in the nineties, making use of the extensive operational experience obtained with the two transportable systems. Design

goals

Major considerations followed ensure full capability for all modifications and alternative approach. I. Optimize

in the design are to bring down present and future SLR satellites applications through a highly

cost of operations

Deploying SLR systems is quite expensive, design aiming at minimizing these costs, time-between-failure (MTBF) and facilitate a. Minimize

the operational costs, to and to facilitate future structured and modular

crew

in particular in transportable mode. must minimize crew size, increase trouble shooting and repair.

A modern the mean-

size

The default mode of operation will be fully automatic. This means that the system will initiate _d pei'form all basic tasks, e.g. maintaining satellite alert information, ephemeris calculation, initiate and perform calibration and tracking procedures, data screening, Normal Point calculation and data mailing. The routine task for the operator (if any) will be to monitor system performance and to respond to problems. Manual mode of operation is always possible through operator intervention. 12-40

b. Reliability Increasing the MTBF will be primarily accomplished through defining a maximum of all functions in software, running in a multiprocessor parallel architecture. The remaining hardware will be built from high quality, highly integrated components, including Programmable Gate Array (PGA) logic. Integrated components also allow a great deal of miniaturization. c. Self-diagnostics capability Diagnostic tasks running concurrently and localization of problems. Parallel functions.

with the control tasks will enable rapid detection processors are very convenient for hosting these

d. Modular hardware design The hardware is designed with an optimum number of printed circuit (PC) boards. Same processor-functions at different locations use identical PC-boards. Different application-functions are realized with a minimum number of different PC-boards. Interface-functions are accomplished by separate PC-boards in order to optimize hardware portability. Most spares will be available at PC-board level which enables rapid replacement, reducing system down-time under field conditions. II. All-satellite a. Satellite

capability

altitude

This decade a variety of high and low SLR satellites are or will become available. Typical extreme altitudes are ARISTOTELES (200 km) and METEOSAT (geostationary). A design goal is to eliminate any logical restriction to the satellite range in the control system (e.g. by accommodating time interval measurement and event timing). b. Satellite interleaving Because the system will be designed for automatic ranging, the implementation of decision schemes in software for interleaving observations to different satellites will be relatively easy. III. Flexible'design The design of the control system must simplify future modifications in view of new or modified requirements as well as implementation of more advanced technologies. This calls for a highly structured design in software and hardware, which may also enhance the portability of the design to other SLR stations, with different hardware environments. a. Adaptation to modifications Hardware, The general electronic design functions: 1. interfaces, 2. applications and physically separated in PC-boards which are uses PGA logic, which can be easily modified Software The RT-software system is strictly 1. global (functional), 2. device dependent, modifications can be implemented locally in affecting other layers. Because of this high level of structurization SLR systems which are quite different from 12-41

is separated into three major types of 3. processors. These different types are individually exchangeable. The hardware (by software). separated into three independent layers: 3. interface dependent. In this way, the related module at one layer, without

it is feasible to implement this design the MTLRS concept, with a minimum

in of

modifications, primarily in software. b. Expanding processor capability If new requirements dictate additional computer power (e.g. RT-filtering, Frame Grabbing) this capability can be easily implemented in the structured design by adding standard processor boards hosting additional software tasks. To enable implementation of different types of (parallel) processors than the transputer which has been selected in the present design, all software is written in ANSI based high-level languages and no RT-program modifications have to be made because the 3L-compiler [3L] supports different

processors.

c. Highly,manufacturer independent Custom made hardware usually increases the dependence on particular manufacturers. This is avoided by full restriction to standard bought-out components and self-designed PC boards. The parallel p_rocessor set-up eliminates the necessity of selecting a vendor dependent RT operating system and thus ensures the flexibility of adopting any brand of (future) parallel processor. In the current design the INMOS transputer (see below) has been selected because of its early availability and its capabilities. Any future transition to another parallel processor is possible and will basically only require the exchange of the T805 credit card size processor board. Lay-out

of the design

I. Hardware, In SLR the great progress in (opto-)electronics of the last decades has been applied to aspects such as detection, laser technology and timing, but seems to have largely bypassed the issue of Control System design. Here state-of-the-art electronics is introduced in the MTLRS Control System, resulting in a modular design with a high level of reliability,

miniaturization

and flexibility.

Probably the most dominant feature in the design is the application of a parallel processor architecture. This choice supports most of the design goals in a unique way, in particular the issues of modularity and of adaptivity to future requirements for extended computing power. The

INMOS

IM

transputer

I

Parallel-processing nov '88] can be architectures;

computers [BYTE divided into two basic the shared-memory

multiprocessor (fig. 1) and the multicomputer type (fig. 2). All existing microprocessors can be used in the first architecture which will be limited in practice to about 4 processors due so called "von Neumann" bottleneck. Fig 1

to the This

bottleneck arises when one processor to the shared-memory while the other processors have to wait during this time till the "sharedFig 2 memory bus" is free! Only few processors like the INMOS TS00 transputer family ['INMOS 1989] and the Texas Instruments TMS320C40 [TMS User guide] can make an optimal use of the multi-computer architecture because these processors possess high 12-42 reads

or

writes

speed ports

communication ports besides their standard Oinks) the processors communicate with

address and data busses. Through each other in a network where

these each

processor possesses its own memory and thus eliminating the yon Neumann bottleneck. In a practical multicomputer network the number of processors can be almost unlimited. The INMOS transputer family consists of T2.., T4.., T8.. and "1"9... processors supported by link-crossbarswitches and periferal link-adapters. All INMOS processors are equipped with 4 links. Each link consists of one serial transmitting wire and one serial receiving wire both having an unidirectional speed of 20 Mbaud. If a link sends and receives data at the same time, the total data transport will be limited to 23.5 Mbaud/link. The T9000 will be available at the end of 1992 and has a peak performance of 200 MIPS and 25 MFLOPS with link speeds up to 100 Mbaud.

T 805 64 bit Floating

Peak performance 25 MIPS 3.6 MFLOPS

point unit

25 MHz (Internal) Reset "//////t'__ CPU

--

Boot

--

5 MHz clock

from

EPROM

11_7_Etherne --_==_-I

4 Kb SRAM

Memory ....

_

i HAM

I SCSI

2O Mbau_-_d

GPIB

t

¢_

Interface .................

:::::32:::::::::::::::::::::::::::::::

20 Mbaud it=

up to 4 Km ::: ::::::::1

Boot II 4Sbytell RS232/IEEE EPROMI I DRAM I ICENTRONICS Fig 3

For the MTLRS Control System T805 has been selected and

the is

running at an internal clock speed of 25 Mhz giving 25 MIPS and 3.6 MFLOPS peak performance. The external 5 Mhz clock is internally multiplied by 5. Fig. 3 illustrates how the processor can be interfaced to the outside world in a flexible way. The present Control System is designed around a minimum of 5 transputers T805 which are interconnected through their links. A link is connected to a server-program running on a host-PC and through this link all the different programs are distributed through the network and loaded to a particular transputer. Each program is started automatically after it is loaded onto the destination

processor. The design makes no use of the boot-EPROM facility but this makes the processor also very powerful for embedded designs. The word TRAM is an abbreviation for TRAnsputer Module which is an off-theshelf interface board delivered by several manufacturers. For parallel processor networks the software can task can be cutted into pieces which run in parallel can run in two or more processors so that each multitasking in one processor can be programmed

be programmed in various ways; one at different processors, different tasks processor runs a different task. Also in a partly parallel way because the

T805 can perform processes like cpu- operation, link-communication, timing and memoryIO concurrently. This means that while e.g. one task is busy with link-communication, the cpu is free for other tasks, scheduled in a sequential way following task priority. When a task consists of a number of processes, running concurrently in different processors, process-synchronization can be accomplished by the use of channel I/O (inter process data exchange) and semaphores. Part of the internal CASH of 4 Kb SRAM is used to hold all registers for up to 99 different tasks and internal processes. If more than 99 modules are running concurrently, external DRAM is also used to hold registers. In contrast to conventional processors where the cpu has to share time to all processes, the T805 is doing this partly parallel which leads to a much higher performance. 12-43

USER INTERFACE

REAL TIME PC

. MAIN PC [

1,nk

1

]

MAIN CONTROL

lin_

UNIT interface

interface I

I

interface

interface

MOUNT

I/O ]IEEE I IEEE

I II

I

NET

i TIME

1 RS2a__k_2 tlJ :':'.

DATA

ACQUISITION

:::::::

UNIT

interlace

I/O [ IEEE IEEE RS232 _

MAINS

l

ma, lpUtll

POWER

SUPPLY

Mainsflltedng Mainsdisffibutlon S_fetyalarmbutton

Fig 4 12-44

_

timer

_

relays

_

The hardware

lay-out

The hardware system can be separated into four major units, the user interface, the main control unit, the data acquisition unit and the mains power supply (fig. 4). It consists of several printed circuit boards, which largely demonstrate the modularity of the design and make it flexible to future modifications. The basic board is the processor board 1"805. On this creditcard size board the T805 transputer is placed, which is dedicated to parallel processing (see above). In addition this processor board contains 4 MByte memory, a special reset system for checking that the processor is still alive (EMI protection) and four 20 MBaud serial link drivers. Other creditcard size boards in the system are IEEE-488 and RS-232 interfaces for connection to standard devices. These creditcard size boards are daughter boards to the extended eurocard size boards of the Main Control Unit and the Data Acquisition Unit.

MAIN CONTROL

lit

inte[[_ce

IEEE

in

el: ace

MOUNT

J

11 1tJ

RS--

[i!:TsOs:!:i:!:!: I

i:!!:i:!:i:i:i:i:i:i:i?:

8 * RS232 2 * IEEE-488

The

_ce

TIME

I/0

fig

er

UNIT

Clock Synchronization Timer Range Gate Window

2 - axis mount controller limit-& safety switches

Computation network

5

Main I)

Control

the I/0

Unit (fig. 5) is made

board

for connecting

up of four extended

the processor

to IEEE,

eurocard RS-232

size

boards:

devices.

2) the TIME board for all time dependent tasks: 4 channel event counter (low accuracy 20 ns), range gate programmable generator (400 ps one shot accuracy), window programmable generator (20 ns resolution), The

programmable generator/synchronizer for the shutter, independent from window, RT-clock with battery backup, programmable synchronization for all required signals (e.g. laser firing). TIME

board

requires

10 Mhz and 12-45

1 pps input signals.

3)

the MOUNT board for driving a 2-axis telescope. The specifications for both axes are: 32 bit position and 16 bit velocity and acceleration controlling, three 16 bit coefficient programming for PID filter, DC and DC-Brushless motors driving, position and velocity mode of operation, interface for quadrature incremental encoder with index pulse, absolute encoder up to 32 bit, limit-, safety switches, etc. ports, joystick connection, nonvolatile RAM for storing position, offset, etc. during power off, without battery backup.

4)

the NET board with a minimum of one processor board with extending computer power up to five processor boards. Additional be implemented for expanding the network even further.

Instead of one big backplane interface, four individual backplane designed. The design of each small backplane is dedicated to application, e.g. dependent on the type of encoders etc. Each backplane can easily be modified to different input signal positive, NIM, "VI'L, 50 fl, ....

DATA

ACQUISITION

the possibility NET boards

of can

interface boards are the typical hardware levels

e.g.

negative,

UNIT

interface

lie IEEE IEEE RS232 RS232 T805!:!:!:!:!:!:. -

2"777

DAC' s

inputs _

ADC's

outputs

i/// i/// r/// _-///

_-//J LL/-Z

opto counter / _..._.... couplers timer _///_ relays

':';':-:-;-:-:-:-:-:-:-2;!

8 ° RS232 2 * IEEE-488 Remote Control On/off, Standby Temp. control etc.

Fig 6

The Data Acquisition Unit (fig. 6) will be used for controlling additional hardware for the purpose of automation e.g. DC-motors, stepper-motors, temperature control, contacts checking and on/off-switching of equipment. It consists of an I/O extended eurocard size board which is identical to the one in the main control unit, together with several Data 12-46

Acquisition -

boards like:

12 bit D/A-converter offset. 12 bit A/D-converter and auto-calibration.

(multiplexed (multiplexed

for for

8 outputs) 8 inputs)

with with

programmable

programmable

range range,

and offset

24 parallel input/output ports. Counter/timer for controlling counting/timing functions. Galvanic separation by opto-couplers and solid state relays.

The User Interface (fig. 4) consists of two identical 486 Personal Computers (PC) which are linked to the transputer network by means of 20 Mbaud link extension boards located in each PC. At this level the system can also be connected to Local and Wide Area Networks protocols. Finally filtering

by

adding

standard

interface

boards

to

the

the Mains Power Supply (fig. 4) is designed mains. It also features an alarm button for safety.

PC-ISA

for

bus,

distribution,

utilizing

standard

switching

and

II. Software General aspects The computer configuration includes a transputer network and two identical PC486AT's, the Main PC and the Real-Time PC, each connected to the transputer network by a so called 'link'(fig. 4). On both PC's the operating system MS-DOS is running together with the highly efficient and friendly Microsoft WINDOWS which is a widely accepted industry standard graphical environment for PC's. In the transputer network the RT-software is running without any operating system. On both PC's the WINDOWS TRANSPUTER FILE SERVER is running. The Main PC hosts the offline _oftware and is used for data storage. Tasks like: site installation, preparing predictions, real time processes, data screening, normal point calculations and data mailing are initiated from or performed by the Main PC. The Real-Time PC interacts with the ranging process as a terminal through the on-line Real-Time (RT) windows transputer file server program, supporting graphics capabilities for the transputer network during all real time processes like: satellite ranging, target calibration, telescope alignment and diagnostics. Through the Real-Time PC the operator interacts with all ranging related processes. The transputer network is built around several transputers where the actual RT-software is running in order to directly control the hardware. Modification of (default) input parameters for the various modules by the operator is based on WINDOWS dialogue boxes and menus which are on both PC's supported by extensive help information. In case of malfunctioning of one of the PC's, the software is designed to run on one PC only, with minimum inconvenience to the operator. The

Thus

each PC is a back-up

unit for the other

one.

off-line software This software is largely based on the original MTLRS system and will be updated for compatibility with the PC-hardware platform under MS-DOS. The software will be embedded in the WINDOWS graphical environment. Error handling is supported by supplying whenever whereas

help information feasible. Primarily individual existing

programmes. Both standard to ensure

and by allowing the new software modules are kept

operator intervention in run-time package has been re-written in C, in FORTRAN-77, callable from C-

the C- and the FORTRAN coding are based on to the ANSImaximum portability of the software. All data files have been 12-47

defined format.

in ASCII-format,

unless

specific

requirements

dictated

the

use

of a binary

The RT-software The transputer software system consists of a set of communicating processes, logically combined in functions. To make the software usable for other ranging systems and to facilitate future developments, the software has been structured into three layers: a. b. c.

the GLOBAL, SLR SYSTEM independent software the DEVICE or data format dependent DRIVER software the INTERFACE type dependent DRIVER software

Diagnostic capabilities are integrated in all three running concurrently by default or can be activated

layers of the RT-software to run concurrently.

and

are

a. Global system The global system includes all functions of the present MTLRS RT-software. The current MTLRS limitations concerning the observation range, the correction domain and the fixed firing epoch have been eliminated. Interleaving satellite ranging is made possib!e. To optimize this, the coarse pass prediction integration and the fine prediction interpolation are moved into the RT-software, This also allows firing epoch dependent range residual calculation to speed up data filtering. The meteo data and the UTC-GPS time are made available to the RT-time system. b. Device

driver

A device driver prepares the standard information from the global system to meet the requirements of the device and vice versa_ It executes a device function by preparing a device command according to the operational device protocol and by calling interface drivers for data transfer. Device drivers are available for all external devices to be accessed c. Interface

by the global

software.

driver

An interface interface'to

driver transfers the data from the transputer network through each device and vice versa. No modifications are applied

a hardware to the data.

Interface drivers are available for each type of interface and protocol used by the external devices. Event interrupt processing is supported for devices which are capable of asserting the transputer event signal. International

design

standard

Because of the structured approach this activity will contribute to the control systems.

to the design establishment

of this control system it is hoped that of an international standard for SLR

References -

Beek,

W. International OCA/CERGA,

and K.H. Otten, "MTLRS-2 Workshop On Laser Ranging Grasse, France.

Upgrade", Proceedings Instrumentation, Matera,

Seventh 1989,

INMOS, "The Transputer Databook", Second Edition, document No.72 TRN 01, 1989, Redwood Burn Ltd, Trowbridge. TMS User guide, Texas Instruments Data Book: TMS320C40 User guide. BYTE: November 1988, Parallel Processing in depth, TS00 and counting. 3L Ltd. 'Peel House, Ladywell, Livingston EH54 6AG SCOTLAND, Telefax: -506-415944. 12-48

203

UK

Airborne

and Spaceborne Systems

N94-15614 AIRBORNE

2 COLOR

RANGING

EXPERIMENT

Pamela S. Millar James B. Abshire Jan F. McGarry÷ Thomas W. Zagwodzki* Linda K. Pacini*

Experimental Instrumentation Branch, Code 924 NASMGoddard Space Flight Center Greenbelt, Maryland 20771 +Crustal Dynamics Project, Code 901 *Photonics Branch, Code 715

ABSTRACT Horizontal

variations

in the atmospheric

refractivity

are a limiting

error source for many

precise laser and radio space geodetic techniques t-7. This experiment was designed to directly measure horizontal variations in atmospheric refractivity, for the first time, by using 2 color laser ranging_me_urements to an air_craf_t_. The^2.collor!a_r syste_k2t_e a T-39 aircraft. Circular patterns wmcn extenaeo irom mv _,u _, D.C. Beltway to the southern edge of Baltimore, MD were flown counter clockwise around Greenbelt, MD. Successful acquisition, tracking and ranging for 21 circular paths were achieved on three flights in August 1992, resulting in over 20,000 two color ranging measurements.

INTRODUCTION Atmospheric refractivity is a limiting error source in precise satellite laser ranging (SLR). Several models have been developed to correct SLR measurements for the increased optical path length caused by atmospheric refractionl-3'8: 9. The first three fomr_Ul_c_S_et_ at atmospheric refraction is spherically symmetric, ano require surtace metero oga laser ranging site. A higher order model developed by Gardner 8,9 can, in principle compensate for spatially varying atmospheric refraction by utilizing pressure and temperature measurements at many locations near the laser ranging site. However, this model depends on the surface measurements being strongly correlated to atmospheric density at higher altitudes. 13-1

A more accurate method for determining the atmospheric correction is to directly measure the atmospheric dispersion with a two color laser ranging system4-7. In the two color ranging approach, the single color range and the differential delay between the two colors is measured. Due to atmospheric dispersion, the differential delay is proportional to the pathintegrated atmospheric density, and therefore it can be used to estimate the atmospheric correction. In 1980, Abshire proposed to measure the magnitude of the horizontal gradients in atmospheric refraction by making dual color ranging measurements to an aircraft equipped with a retroreflector. We have developed and successfully carried out such an airborne two color ranging experiment. We expect that data from this experiment will improve our knowledge of gradients in the refractivity, and could contribute towards improving atmospheric models.

APPROACH

Our experiment utilized Goddard's 2 color ranging system at the 1.2 meter (m) telescope facility with a cooperative aser target package on a T-39 aircraft. A simplified diagram of the experiment is shown in Figure 1. The ground-based laser system tracked and ranged to the aircraft as it circled Greenbelt, MD. With this circular flight pattern, the aircraft passed partially over two major cities, Washington D.C. and Baltimore, and land near the Chesapeake bay. It is likely that variations in the terrain temperatures will cause horizontal gradients in the air density directly above, which could be measured by the experiment. The initial step in the experiment is to acquire the aircraft's approximate location. A realtime GPS data relay link was developed for this purpose. A GPS receiver on the T-39 relayed the aircraft position every 3 seconds through an RF digital packet link to the ground telescope. The telescope's tracking program then determined the pointing angles to the aircraft by extrapolating the aircraft's position from the GPS data. Once acquired, we closed an automatic tracking loop by using a CCD camera and frame grabber which commanded the 1.2 m telescope to automatically track the aircraft's 810 nm laser diode beacon. The aircraft target package was manually pointed to the ground from inside the aircraft by viewing the ground beacon which was modulated at 5 Hz. To determine the optical path length we measured the range 20 psec resolution laser ranging system. For the differential streak camera-based receiver which recorded the reflected 1.2 nanosecond (nsec) time window.

AIRCRAFT

to the aircraft at 355 nm with a delay measurement, we used a 355 & 532 nm pulses within a

INSTRUMENTATION

Figure 2 shows the T-39 aircraft based at (Wallops Flight Faciiity) _ in this experiment, along with the flight crew. The T-39 is a two engine, 4 passenger, jet aircraft capable of maintaining 12 kilometer (km) altitude. The three science flights were limited to 6 - 7 km altitudes due to high cirrus clouds near 7 km. The target package was located beneath the nadir port of the aircraft just aft of the air brake. On the aircraft we used a combination of laser and electro-optic instrumentation for pointing the target package with two radio transceivers for data and voice communications to the ground laser facility. A photograph of the target configuration is shown in Figure 3. The target package consisted of a cube comer array (CCA), a laser diode beacon and a CCD camera with a narrow bandpass filter. The package was secured to a commercial pan and 13-2

tilt camera mount and was controlled with a joystick inside the aircraft. The entire package was protected by an aluminum enclosure which incorporated a 1/4" thick quartz window to transmit the optical signals at 355, 532 and 810 nm. This window was tilted -3 °, with respect to the aircraft nadir, so that the beams would transmit through at right angles to the window. We used baffles on both the laser diode beacon and the CCD camera to reduce scattering of the laser beacon close to the enclosure window. Table 1.

from the window The target package

into the camera. The baffles extended components and parameters are listed in

We used several techniques to control the target package's temperature. To avoid condensation on optical surfaces, tape heaters were attached to the inner walls of the CCA housing and around the filter adaptor tube on the CCD camera. Resistive heaters were installed on the back mounting plate of the target package. An AD590 temperature sensor was placed on the filter adapter of the CCD camera. The sensor indicated that at 7 km altitude the target package stayed at 10°C for the flight duration. Dry nitrogen was flushed into the target enclosure before takeoff and during ascent, descent and landing to displace water vapor, which could condense on cold optical surfaces. The target package could be pointed in azimuth :£5 ° from port, parallel to the left wing and in elevation +5 ° to -15 ° from horizontal. The operator who moved the target package via joystick could instruct the pilots to adjust aircraft attitude if the ground beacon was moving outside of the CCD camera's i0 ° field of view (FOV). This was necessary particularly in azimuth during flights with strong cross winds. The flight crew consisted of two pilots and two science team members. One person pointed the target and communicated with the pilots, while the other operated the GPS control software and maintained radio communication with the ground ranging site. The GPS receiver was used to receive the aircraft's lattitude, longitude, altitude and time. The GPS receiver relayed its information to the flight computer via a RS232 port, as well as to the packet radio. The packet information was transmitted in bursts every 3 seconds. The flight computer allowed the operator to verify that at least four satellites were in the receivers FOV and that the GPS data being sent was correct. An aircraft LORAN unit was used as a backup to initialize the GPS receiver in the event of a computer crash. The ground packet radio receiver relayed the aircraft's position to the DEC PDP 11/24 computer. This mount control computer pointed the 1.2 m telescope to the aircraft. While flying the circular pattern, the pilots used a TACAN transceiver which gave them slant range to GORF where another TACAN transceiver was located.

GROUND-BASED

RANGING

& TRACKING

INSTRUMENTATION

The airborne two color laser ranging experiment was one of several projects which have used Goddard's 1.2 m telescope ranging & tracking facility. This facility was developed for satellite laser ranging. We tried to design the aircraft experiment to minimize the impact to the ongoing satellite ranging program, so that only slight system modifications were made to convert from satellite mode to aircraft mode. A simplified block diagram of the system is shown in Figure 4, with system parameters given in Table 2. For the aircraft experiments the Nd:YAG laser was operated at 10 mJ at 355 nm and 0.3 mJ at 532 nm to avoid damaging the telescope optics. The laser was housed about 10 meters from the base of the telescope. Figure 5 shows the laser transmitter and two color optics. The experimenters area is located at the base of the telescope near the telescope focal plane. This area contains the range receiver electronics, aft optics and streak camera. The aft optics 13-3

for the aircraft configuration is shown in Figure 6. The laser output is directed hole in the 45 ° bifurcated mirror through the telescope and to the aircraft.

through

a

The ranging operation can be described by referring to Figure 4. A start diode samples the outgoing laser pulse to initiate the time interval measurement sequence. The aircraft return pulses are then reflected at the 45 ° birfurcated mirror and into the aft optics receiver. In the receiver a small percentage of the return is split and detected by the MCP photomultiplier tube. This generates the stop signal for the time-of-flight measurement as well as the pretrigger for the streak camera. The majority of the received signal was then separated into 532 nm and 355 nm paths by a dichroic mirror and bandpass filters. Each color was focussed into a fiber optic delay line. The optical delay was required to allow pre-triggering the streak camera with respect to the optical pulse's arrival time. The delay fiber bundle was composed of 27 fused silica fibers with 100 i.tm core for each color which were round at the input side and formed into a single row at the output side. The signals were then imaged onto the streak camera photocathode by the streak camera's input lens. The streak camera's 1.2 ns sweep speed setting was calibrated and used to achieve -4 psec/pixel resolution. During pre-flight tests the detection threshold for optical pulses transmitted through the fiber optic cable to the streak camera was measured to be 6,600 photoelectrons at 355 nm and 4,200 photoelectrons at 532 nm. At the transmitter, an optical delay was added to the 532 nm pulses to allow both 532 and 355 nm received pulses to be recorded within the 1.2 ns sweep window. The 532 nm pulse was delayed with respect to the 355 nm pulse by using a total internal reflection (TIR) cube corner mounted on a computer controlled linear motor stage. This time offset was used to compensate for the additional -2.7 nsec round trip atmospheric delay at 355 nm. Each streak camera waveform pair was digitized to 256 time pixels by 8 bits amplitude and was recorded on hard disk. The streak camera waveforms were stored on the PDP LSI 11/23 at a 2 Hz rate. Each file consisted of the optical waveform pair, the 355 nm time-offlight measurement, the transmitter 532 nm delay setting, telescope azimuth and elevation angles, and, other relevant tracking information. The aircraft was acquired by using the GPS measured aircraft positions which were relayed to the tracking computer. Every 20 msec, the computer tracking program extrapolated the aircraft's pre_ent angular position from the three most recent GPS data points and directed the mount to that angular location. GPS tracking was usually sufficiently accurate to keep the aircraft position within the CCD camera's -.0.3 ° FOV. The tracking loop was closed by reading an X-Y digitizer which determined the angular offset of the aircraft's 810 nm laser diode beacon within the COD FOV. In this mode the mount was driven so that the beacon's angular position stayed within the center of the CCD FOV. Optical background from bright stars, the moon and other aircraft were suppressed with a 5 nm FWHM bandpass tilter in front of the tracking camera. The telescope was able to track the aircraft to a precision of-200 microradians, and successfully tracked the aircraft across the face of a full moon.

STREAK

CAMERA

CALIBRATION

When making two color ranging measurements at 532 and 355 nm, the differential delay is -1/12 the 532 nm delay. The differential delay must be measured to better than 5 psec for i cm single calibrated

color corrections 10,11. to permit this accuracy.

The

streak

13-4

camera's

sweep

speed

had to be carefully

The first stepinvolved finding the zerodelaysettingin thetransmitter.A cubecomerwas placed in the optical pathin front of thetelescopeto avoid anyatmosphericeffects. Then the transmitter's532nm delaywasadjustedto a positionwherethe 532and355nm pulses arrived atthe streakcamerasimultaneously.This wasdenotedthezeroreferenceposition. Next anoptical delayof 100picosecondswassetat the transmitter.Eachwaveformpair was smoothedandpeakpositionswerecomputedalongwith the differencebetweenthe peaks.The 100 psecoptical delay setting was then divided by the peak difference (in pixels) to yield aninversevelocity measurement.This dt/dx value (psec/pixel)wasthen plottedat thex positionwhich wasatthe midpointof the two peaks. This processwasrepeatedseveralhundredtimesasthe pulsepairsweremovedacrossthe streakcamerawindow. Thecoefficientsfor the sweepspeedwerecomputedasthe bestfit to the inverse velocity versus midpoint plot as shown in Figure 7. To check the calibration, and to process the flight data, the time difference between the 355 and 532 nm pulses were computed by integrating the inverse velocity profile 5. The calibration was checked by displacing the TIR cube comer to other known values for comparison with calculated optical delays. The system's calibration was tested either before or after each aircraft experiment, and agreement was typically -4 psec.

SUMMARY: On the evenings of August 5, 6 and 7, 1992 we conducted a two color ranging experiment to an aircraft: Each ranging experiment lasted about 2.5 hours as the aircraft circled the ranging system at an altitude of 6 to 7 km and a slant range of 20 to 25 km. This resulted in 7 to 8 full rotations around the ranging site per flight. About 6,000 to 8,000 two color laser ranging measurements were recorded on each flight. In total 30 Mbytes of calibrated 2 color ranging data were collected. An example of a waveform pair from 8/05/92 is illustrated in Figure 8. Both smoothed and raw data for the green and UV are shown. The waveforms are smoothed for analysis by convolving with a raised cosine pulse. Most of the waveforms had pulse widths wider than the transmitter. Some pulse broadening was caused by modal dispersion in the multimode fiber delay lines. Additional broadening on some signals may be due to saturation of the streak camera. However, in most cases, the sharp leading edges of the optical pulses were preserved in the waveforms. Our work to date shows that timing to the pulse's leading edges appears to have the least timing jitter. I

We are currently in the process of normalizing the data for varied range and elevation angle as well as optimizing our method for calculating the differential delay between the 2 color pulses. These results will be compared to atmospheric models3,8, 9 and published upon completion. ACKNOWLEDGEMENTS The authors wish to thank NASA Headquarters Code SEP, the Crustal Dynamics Project and the EOS GLRS instrument project for supporting this work. We are indebted to Amie Abbott, Jonathan Rail, Richard Chabot, Jack Cheek, Kent Christian, Jimmie Fitzgerald, Dan Hopf, Bill Krabill, Earl Frederick, John Riley, Robert Gidge, George Postell, Virgil Rabine, Dave Pierce, Evan Webb and John Degnan for their many contributions towards the completion of this project.

13-5

REFERENCES 1. Bender,P. L. and J. C. Owens,"Correctionof Optical DistanceMeasurementsfor the FluctuatingAtmosphericIndexof Refraction," J. Geophys.Res.,70(10) pp. 2461-2462,May 1965. 2. Marini, J. W.,

and C. W. Murray, Jr., "Corrections Atmospheric Refractions at Elevations above Nov. 1973.

.

Herring; T. A., "Marini and Murray Atmospheric Range Correction to the CSTG SLR Subcommission meeting NASACDP Principal meeting, Munich, October 1988.

4. Abshire, J. B., "Plan for Investigation Systems," NASA X-723-80-10, .

6.

o

of Laser Range Tracking Data for 10 Degrees," NASA X-591-73-351,

Atmospheric Errors January 1980.

Abshire, J. B., and C. S. Gardner, "Atmospheric Laser Ranging," IEEE Trans Geoscience 425, 1985.

in Satellite

Formula," Report Investigators

Laser

Ranging

Refractivity Corrections in Satellite & Remote Sensing, GE-23(4) pp. 414-

Degnan, J., J., "Satellite Laser Ranging: Current Status Trans. Geoscience & Remote Sensing GE-23(4)pp.

and Future 398-413,

Prospects," 1985.

IEEE

Bender, P. L., "Atmospheric Refraction and Satellite Laser Ranging," submitted to: Syrup. on Refraction of Transatmospheric Signals in Geodesy: The Hague, May 19-22, 1992.

8.

Gardner, C, S., and B. E. Hendrickson, "Correction of Laser Tracking Data Effects of Horizontal Refractivity Gradients," Radio Research Ld_ratory University of Illinois, Urbana Illinois, December 1976.

9.

Gardner, C. S., "Correction Refractivity Gradients,"

for the 478,

of Laser Tracking Data for the Effects of Horizontal Appl. Optics 16(9) pp. 2427-2432, September 1977.

10. Im, K. E., C. S. Gardner, J. B. Abshire, and J. F. McGarry, "Experimental evaluation of the performance of pulsed two-color laser-ranging systems," Soc. Am. A 4(5) pp. 820-833, 1987.

J. Opt.

11. Zagwodzld, T. W., J. F. McGarry, A. Abbot, J. W. Cheek, R. S. Chabot, J. D. Fitzgerald and D. A. Grolemund, "Streak Camera Returns from the Relay Mirror Experiment (RME) Satellite at Goddard space Flight Center's 1.2m Telescope Facility," submitted to: American Geophysical Union Monograph, Space and Geodesy and Geodynamics, Fall 1992.

13-6

Table1. 1.CubeComerArray

2. Laser

Diode

Target

Beacon

AircraftInstrumentationParameters 23 element array, 2.5 cm cube comers, 5 arc second accuracy, UV coated, R=78% @355 rim. SDL-2460 A1GaAs Array, 780 mW at 810 nm, 7 ° azimuth by 8 ° elevation divergence.

3. CCD Camera

Philips model 56471 camera, 25 mm lens, f/0.7, 10 ° FOV, Bandpass Filter: 810 + 2.5 nm.

4. Pan & Tilt Camera

Mount

5. GPS Receiver

VICON model V353AtrI'V variable drive system and tilt controller, 8" by 12.75" mounting plate Motorola EagleVIII, 4-channel, simultaneous L1 C/A code carder

O

O

07 O6

oo oo @ o°o oOO o_ooo_ _ oo c_

o °

05 O4 O3

0

0

O_

02 01 00

0

I

I

22

t

I

,

44 Streak

Figure

7.

I

I

66 Camera

I

_

88

1

110

Window

z

I

132 Location,

t

i1

I

154

I

176

L

t

198

X (pixels)

Streak camera sweep speed calibration for 2nsec sweep using data taken 3/24/92. The linear fit using 100 data points is displayed in the upper right hand corner of the plot. The units of inverse velocity are psec/pixel.

13-15

FLIGHT 260

I

I

DATE I

I

8/05/9Z I

I

'

,"

I

I" '"

i

!

240 220 200

1BO

,6e X

_' 14o

I-

,-,

100

_.1 O.

=: _C

BO 60 40 20 0

I

!

0

20

40

_

60

I

,

i

GREEN 260

--T

_----

F'-----I ---_

I

,

1



I

i

100

BO

i

I

!

I

120

140

160

1B0

i

200

I

220

=

I

i

240

250

WAVE POSITION[P]XELS] ]

1

_

I

,

1

_

I

,

i

.

i

I

i

240 220 20ff 180 O3 --I

"' x

160

_'14o 120 I,-

'--' 100 _I

.,_ BO 60 40 20 =

0 0

i

i

40

20

Figure

8.

60

,

80

100 120 t 40 160 U-V ',/AVE POSIT[ON [P]×EL5]

100

I

200

=

I

220

_

240

Two color waveform pair recorded by Hammamatsu model C1370 Streak Camera. Each waveform is shown in both raw and smoothed form. The horizontal scale factor is about 4.2 ps/pixel determined by calibration. For this pulse pair, the differential delay (including the optical delay set in the 532 nm path) was 2.667 nsec. 13-16

'60

N94-15615 GLRS- R 2-COLOUR

RETROREFLECTOR TARGET PREDICTED PERFORMANCE

Glenn

DESIGN

AND

LUND

Optical

Department

AEROSPATIALE100 Boulevard 06322

Space

& Defense

Division

du Midi

CANNES

LA BOCCA

FRANCE

ABSTRACT

This paper reports on for the GLRS_R spaceborne

the retroreflector dual-wavelength

ground-target laser ranging

design system.

The described passive design flows down from the requirements of high station autonomy, high global FOV (up to 60 ° zenith angle), little or no multiple pulse returns, and adequate optical crosssection for most ranging geometries. The proposed solution makes use of 5 hollow cube-corner retroreflectors of which one points to the zenith and the remaining four are inclined from the vertical at uniform azimuthal spacings. The

need

for

fairly

(within turbulence diffraction lobes, vectorial accuracy compromise solution of the retroreflector

large

(~

I0

cm)

retroreflectors

is

expected

limitations) to generate quite narrow thus placing non-trivial requirements on the of velocity aberration corrections. A good is found by appropriately spoiling just one dihedral angles from 90 ° , thus generating

two symmetrically oriented The required spoil angles ground target latitude.

diffraction are found

to

lobes have

in the return little dependance

Various link budget analyses are presented, influence of such factors as point-ahead turbulence, ranging angle, atmospheric visibility target thermal deformations.

13-17

beam. on

showing the optimisation, and ground-

i.

BASIC

GROUND

TARGET

REQUIREMENTS

During the various study phases investigated of the GLRS project, the Ground Target established to be of most significant following

:

• Choice

of

a

multiple-retroreflector,

• As nearly full zenith angles • Avoidance

of

ambiguous

• Adequate photon under most clear • Moderate

cost

high

of of

2.

following and as a

BASELINE

The

space practical

local

in

keeping

ranging

with

the

accuracies

GT

design leads to the choice (RR) concept where full summation of several contributing

approach diagram

the gradual fall-off inter-RR crosstalk

TARGET

of

of a sky RR

result firstly a static RR,

of this diagram (i.e. multiple

achieve requirements be achieved

an

-

from and from pulse

acceptable and the need under most

Although

applications)

GT

in

between of RRs. i00 mm, minimise

good with size

is is

CONCEPT

5-retroreflector

limits

to

returns.

sub-centimetric conditions.

in this energy

provides a good compromise overlap and minimal number have a useful diameter of construction in order to deformations.

up

concept.

sections the proposed GLRS-R GT design its numerically simulated performance function of several important system variables.

GLRS-R

baseline

possible,

pulse

The adopted solution must therefore compromise between somewhat conflicting for accurate range measurements to conditions. In the described, illustrated

target

reliability.

difficulties reflec[ed

secondly - because the requirement return) avoidance.

as

correction,

budget for atmospheric

and

sky

(multiple)

The requirement of a passive multiple fixed Retroreflector coverage is achieved by the FOVs. inherent non-uniform

passive

coverage of the of at least 60 ° .

• Adequate velocity aberration link budget requirements.

The the

during the course (GT) requirements importance were the

thermal small

and

design

performance solid

temperature

13-18

illustrated

in

Fig.

1

full sky coverage, minimal FOV The individual retroreflectors and will need to be of hollow thermally induced wavefront cube-corners, excursion

is

achieved there beyond

(in are which

the resulting re fract iv_._-ind'ex unacceptable wavefront distortions.

gradients

will

generate

SOUTtt

Figure

1.

Proposed

baseline

As shown in 9 10p, wavefront distortions

5-

cube

corner

ground-target

thermo-mechanical effects in hollow retroreflectors,

can

also although

are expected to be an order of magnitude smaller induced under comparable conditions in a refractive reasons of resistance to environmental influences, reflectors will need to be covered by a protective quality)

lead

to they

than those medium. For the hollow (optical

window.

Preliminary thermo-mechanical analysis has shown that the GT support structure can be made of common materials which would give rise to reflector location stabilities of ~ 3 mm, for temperature excursions of ± 50 ° C (± 90 ° F). If required, partial correction for these excursions could probably be made using epoch and climatic data together with an appropriate thermomechanical model for support structure deformations.

13-19

As explained in §3 and §4, velocity aberration achieved by designing the RR far-field diffraction exhibit a symmetrical twin-lobe structure. Each be oriented (about its local with respect to the overhead achieve appropriate alignment

correction patterns

of particular tracks, lobes.

normal) in a spacecraft of the reflected

the in

is to

RRS must direction order to

Although a 45 ° inclination of the peripheral reflectors provides good overall FOV coverage and has been assumed in the following analyses, the choice of this value is somewhat arbitrary. Parametric analysis could reveal a more favourable inclination, depending on the criteria used to trade link budget performance at high zenith-angles against FOV overlap limitations. Variants involving more than 4 peripheral reflectors could also be considered, although they would incur an increase in the extent of

FOV

overlap,

It is inhibit crosstalk

assumed operation effects

any allowed the return relates the

3.

and

higher

INFLUENCES

Two-way

IN

atmospheric

conditions,

mean

zenith

long-exposure



dihedral

• Detection use

of

camera.

and

will where for

angle

receiving

PERFORMANCE

depending

and

GT

energy

depending

influences Those which are listed

on

visibility

altitude.

characterised

by (

}

Cn2

local and on

wavelength

and

the

(h).dh).

zenith range.

ranging

angle,

local

geometry

and

velocity.

characteristics

techniques, a

BUDGET

as determined by satellite azimuth

aberration, height

Retroreflector and

angle

turbulent

• Ranging geometry, RR incidence angle,

satellite

ranging strategy of geometries and determines,

several quite different link budget performance. to in the "present paper

turbulence,

• Velocity

LINK

transmission,

local

• Atmospheric

costs.

operation, which of the RRs must be providing An appropriate deterministic correction then RR range to a common GT reference point.

AS shown in Fig.2, there which can affect the system are considered or referred below : •

GT

that the spacecraft (S/C) for the small percentage are expected to be strong

ranging signal. measured

PRINCIPAL

overall

such

as

the

case

size,

optical

quality

spoiling. which telescope,

in

transfer

of

GLRS-R

optics

and

imply a

the

streak

I SPACI_ORNE

_ACEBORNE

El_ SSION

-. _,

+ Wavebngths + Lasere_ergy &Pdsewid_h + Be_mPointing& Divergence

| J

6.5"1010 surface elevation measurements with the selected horizontal resolution (grid size) and an image of Earth's surface reflectance at 1 _m wavelength with a similar level of detail. These products will be developed with minimal ground processing, stored on compact optical disks, and made available to the scientific community in a timely fashion. Analysis and use of the topographic data will be done by tl_e end user in the scientific community. REFERENCE: Gardner, C.S., "Generalized Applied Optics, 1992.

Target

Signature

Analysis

for Laser Altimeters",

ACKNOWLEDGMENTS: The definition of the MBLA instrument was made possible through the efforts of a number of individuals in the Laboratory for Terrestrial Physics (LTP) at the Goddard Space Flight Center and' in the aerospace industry. Notable among these efforts were those of David Rabine in the University Research Foundation of the Maryland Advanced Development Laboratory, Nita Walsh of Science Systems and Applications, Inc., Ying Hsu of the Optical Corporation of America, David Thompson of Spectrum Astro, Inc., and Tom Baer, Spectra Physics, Inc.

13-81

MULTI-BEAM LASER ALTIMETER MISSION CONCEPT

EARTH PROBE SPACECRAFT SUNLIGHT 400 km SUN-SYNCHRONOUS 6am/6pm POLAR ORBIT

LASER

PULSES

(30)

1 _m WAVELENGTH

SENSOR FOOTPRINT 30 m diam.

NADIR TRACK

ACROSS TRACK DIRECTION

9OO m SWATH WIDTH

Fig.

13-82

1

1.1_

v

(_

E oi

F=

n=

0 t-_r:r_'_ =t

m

(_

o a= n,' w )== w

) ( (

_.o

I

-b °_

==J

t--

8_

n,= W

i

=J

E_

_=

13-83

13-84

i

_J

g

13-85

LO U-

nW W m

-J

] q)

W rn .J

13-86

E o

Poster Presentations

N94-156 1 SATELLITE

M (_ech, Faculty

of Nuclear

phone

/ fax

K.Hamal,

Science Brehova

+42

LASER STATION STATUS 1992

H.Jelfnkov_l,

HELWAN

A.Novotn)_,

I.Proch_ka

and Physical Engineering, Czech Technical 7, 115 19 Prague 1, Czechoslovakia

2 848840,

telex

B.B.Baghos,

121254

Y. Helali,

fjfi c, E-mail

University

[email protected]

M.J.Tawadros

National Research Institute of Astronomy and Geophysics, Helwan, Egypt phone / fax +20 2 782683, telex 93070 hiag un, E-mail [email protected]

GENERAL

The Institute

Satellite

of Astronomy

Laser and

Station

Helwan

Geophysics

has been

in Helwan,

ope'r.ated jointly Egypt

and the

by the Czech

National

Technical

Research University

in Prague, Czechoslovakia (see Proceedings of the 7th International Workshop on Laser Ranging Instrumentation, Matera, 1989). The station components have been carefully tuned to increase the system overall stability and reliability critical for the remote location. The mount correction model based blind satellite implemented, 1 satellite

has

on the Gaussian smoothing (Kabel:_, 1990) has been implemented to simplify the acquisition and tracking. The on-site normal points generation algorithm has been the station has been connected to the international information network. The ERSbeen

included

into the tracking

schedule.

verified by experimental' Etalon 1 ranging by April centimeters is obtained when ranging to ERS-1, Starlette

The station the Smithonian

operation

Astrophysical

has been cosponsored Observatory

contract,

14-1

The

range

capability

1992. The ranging precision and Lageos satellites.

by the DGFI whose

station

contracts

support

ERS-1/7831/91,92

in acknowledged.

has been of

2-3

and

N94-1562 '

C 0 om

L-

0 L.

0

0 e" im

¢D m

0') C I.U m m

u_ X em

era

14-2

14-3

m

"_t0_ im

,m

m

o

0 Z

0 Z