A new cavity pyrheliometer, the active cavity radiometer type IV (ACR IV), has been developed for the .... of the electrical and solar heating is negligible (well.
Active cavity radiometer type V Richard C. Willson
A new cavity pyrheliometer,
the active cavity radiometer type IV (ACR IV), has been developed for the
measurement of total solar optical irradiance. Analysis predicts its ability to measure at the solar constant level with 0.1% uncertainty
in SI units.
In comparison tests ACR IVs have consistently demonstrated 0.3%
higher results than the World Radiometric Reference scale. A prototype has been tested, and a flight instrument has been developed and flown in a sounding rocket experiment to determine the solar constant. ACR IV instrumentation
is being developed for flight experiments on the Spacelab I and Solar Maximum
missions to monitor the total solar output of optical radiation as part of a long-term program to detect variations of climatological significance.
I.
Introduction
The IPS56 error and the high accuracy of the new
Development of pyrheliometric instrumentation for absolute solar radiation measurements has been the
focus of considerable effort at the Jet Propulsion Laboratory for more than a decade.'
Research begun at the
Jet Propulsion Laboratory (JPL) in 1964 by Haley, Plamondon, Willson, and Kendall resulted in the first extraatmospheric observations by a cavity pyrheliometer, made by the Temperature Control Flux Monitors on Mariners 6 and 72 Kendall and Willson produced a succession of increasingly accurate pyrheliometric instruments. Kendall's work was orientated principally toward instruments for defining the absolute
generation of cavity pyrheliometers
were widely ac-
cepted by 1974when the First International Comparison of Absolute Cavity Pyrheliometers was held at Davos, Switzerland. During this comparison and the subsequent International Pyrheliometric Comparisons (IPC IV) held at the same site in 1975,a group of the most advanced of the new cavity pyrheliometers were intercompared.
A new reference scale based on the
results, the World Radiometric Reference (WRR), has now officially replaced the IPS56. Definition of the new scale is to be accomplished by the performance of five
pyrheliometers-ACRs
310 and 311 (Willson, JPL),
radiation scale in laboratory and ground-based solar
PACRAD III (Kendall, JPL), PMO-2 (Frbhlich, Davos),
irradiance experiments. The pyrheliometers developed for these purposes are known as the ACRAD, SACRAD, and PACRAD.1' 3 Willson developed a series of pyrheliometers known as active cavity radiometers (ACRs)
by 2.2%.
whose primary purpose has been the measurement of solar irradiance in high-altitude balloon, sounding rocket, Spacelab, and satellite experiments designed to determine the constancy of the solar constant.14 5 In a series of tests begun in 1968, the solar irradiance
observations of ACRs and PACRADs were compared with those of instrumentation used to define the International Pyrheliometric Scale of 1956 (IPS56) used by the World Meteorological Organization. A 2.2% disagreement between the absolute scale and the IPS56
was discovered4 -8 and subsequently verified independently by others. 9 10
The author is with California Institute of Technology, Jet Propulsion Laboratory, Pasadena, California 91103. Received 11 May 1978. 0003-6935/79/020179-10$00.50/0. ©31979 Optical Society of America.
and the CROM (Crommelynk, Brussels). Measurements on the WRR exceed measurements on the IPS56 The first determinations of the solar constant by absolute radiometers, in which atmospheric attenuation uncertainty was not the dominant source of error, were made by ACR IIs on a high altitude balloon in 1968.7 In 1969, the more accurate ACR III yielded a solar constant of 1369 (0.6%) W/m2 in a 36-km balloon experiment." In June 1976 a two-detector version of the ACR IV was
flown on an Aerobee sounding rocket to measure the total solar irradiance at altitudes above 100 km. The solar constant was determined to be 1368 W/m2 with an 3 uncertainty less than ±0.5%.12"1
II.
Development of the Active Cavity Radiometer
Type IV
A NASA-sponsored design study was carried out by Willson during 1975 to develop a satellite instrumentation system to measure the total solar output of optical radiation and its spectrum. The experience acquired over the past decade in developing the previous active cavity radiometers was applied to the design of 15 January 19i9 / Vol. 18, No. 2 / APPLIEDOPTICS
179
THINFILM BONDED SENSOR TEMPERATURE HEATSINK STRUCTURE CAVITY 1 SECONDARY THIN FILMBONDED TEMPERATURE SENSOR
guarantees that the contribution of the shutter's exitance to the irradiance of the cavity is negligible. Electrical heating provides the necessary power to
SCALE-- 1 cm APPROXITMATE SHUTTER
PRIMARY APERTURE INSULATION ALUMINIZED
CASE SHUTTER
balance the cavity's conductive and radiative losses and
C ELECTRONI g
r..
.
I~~~~~ b
i
I
I
l
d
,
maintains the constant cavity-heatsink temperature difference. When viewing the sun in the shutter open
.-
(observation) phase, the cavity electrical heating power
-~~~~~~~~~~
/ INSULATING\ SUPPORT A THERMAL IMPEDANCES
EXTENSIONOF / HEATSINK POR VIEWLITITIN '
PRIMARY CAVITY
|
5°VIEW X LIMITER
SUPPORTI INSULATING
THIN FILMBONDED CAVITYHEATER
Schematic drawing of the JPL active cavity radiometer type IV.
Fig. 1.
a new pyrheliometer capable of measuring irradiance at the solar constant level with an uncertainty of ±0.1% in SI units. Prototype ACR IV instrumentation was
constructed and tested to verify various aspects of the design.
is automatically decreased by the electronics in an amount proportional to the absorption of solar irradiance by the cavity. Absolute irradiance measurements are derived from the difference in the electrical power supplied to maintain the constant cavityheatsink temperature difference in the two phases of measurement. From the previous discussion, we would expect the mathematical abstraction of the radiometer's operation to have the form H=K(Per-Peo)+E, where H = measured irradiance,
K = standard detector constant of proportionality, Per = cavity electrical heater power in the reference phase of the measurement,
Description of the ACR IV
A.
The design concept of the new cavity pyrheliometer, designated the active cavity radiometer type IV (ACR IV), is shown in Fig. 1. Two right circular, conical cavity
detectors are thermally connected to the heatsink through their respective thermal impedances. All four parts are electrodeposited 99.99%pure silver. The interior of the cavities are coated with a specular black
paint. A low temperature coefficient heater winding is bonded to the back of each cavity in the region cor-
responding to solar irradiance of the cavity interior. Resistance temperature sensors are bonded to the top of each thermal impedance near the cavity apertures. The primary cavity is irradiated through a precisely machined and accurately measured primary aperture. The detector's field of view is defined by the primary
cavity's field of view through the secondary aperture at the top of an extension of the heatsink. The heatsink assembly is insulated from the outer case. Operation
B.
of the ACR IV
(1)
Peo = cavity electrical power in the observation
phase of the measurement, and E = sum of many small terms due to small departures from instrument equilibrium and due to uncertainties in instrument parameters relative to SI units. The constant of proportionality is the term that facilitates the absolute measurement of radiation by a standard detector. It contains instrumental parameters (e.g., the detector area, absorptance, and heater resistance) whose specifications determine the interaction between the radiometer and irradiant fluxes in SI units. C.
Discussion
of the ACR IV Design
While the starting point in designing the ACR IV was its predecessor the ACR III,4 there are some important
new features. The most significant of these are (1) the
The dissipation of radiant or electrical heating power
in either cavity will produce a temperature difference across its thermal impedance. This difference, transduced by the resistance temperature sensors, is used by an electronic servo-system in the active cavity radi-
ometers to maintain automatically constant power
use of a specular cavity absorbing surface, (2) a dual-
cavity detection scheme, (3) an isothermally confined field of view, (4) thin film bonding of cavity heater and temperature sensors, (5) improved electronics, and (6) electrodeposited cavity and thermal impedance structures.
dissipation in the primary cavity by controlling a dc
The ACR IV uses a right circular cavity with a wire-
The primary
wound, low-temperature coefficient, resistance heater winding attached to the outside by a thin coating of an electrically insulating, thermally conductive bonding agent. Additional coatings of the bonding agent are used to immerse the windings in the bonding medium. The windings are then covered with two back-to-back aluminized kapton shields to prevent radiative, conductive, or convective heat transport to the cavity's rearward field of view. The specular black absorber is
voltage supplied to the cavity heater.
cavity detector of the ACR IV is thereby accurately
maintained at a slightly higher (1 0C) temperature than the heatsink at all times. The secondary cavity is not controlled but allowed to drift passively in temperature
in equilibrium with its heatsink environment. The ACR IV operates in a differential
mode.
A
shutter alternately blocks solar radiation from, and admits it to, the primary cavity. In the shutter closed (reference) phase of the measurement, the ACR IV views its own heatsink.
The high-reflectance
low-
emittance shutter surface seen by the primary cavity 180
APPLIEDOPTICS/ Vol. 18, No. 2 / 15 January 1979
applied to the inside surface of the cone in the minimum
required thickness for coverage. Care is taken to prevent the formation of a significant meniscus at the apex
of the cone. The temperature differences across the black absorber and the cavity heater's bonding layer were modeled using measurements of the thicknesses of these coatings and-their manufacturer's published values for thermal conductances. For equal power dissipation of solar constant magnitude at the external edge of each surface, the temperature differences were of the order of magnitude of a few microdegrees Kelvin.
At this level the contribution to the uncertainty of solar irradiance observations due to power nonequivalence of the electrical and solar heating is negligible (well below the 0.01% level). By far the largest single improvement in the accuracy of the ACR IV relative to previous ACRs arises from the
use of a specular black coating on the cavity surface. (Diffuse blacks were used in previous ACRs.) The ACR IV cavity has a 300 cone angle, which causes six internal
interactions with the cavity walls for an axial incident ray before reflection out the aperture. The effective cavity reflectance is then
rium with the incident radiant power, the electrical heating power provided by the electronics, and the thermal environment of the ACR heatsink. The secondary cavity is in equilibrium with the thermal environment of the heatsink. Using its temperature sensor in a bridge with that of the primary cavity, its response to heatsink temperature drift nullifies that of the primary cavity to within the limit of accuracy in matching their thermal responses. The use of electrodeposited silver cavity and thermal impedance structures improves the mechanical and thermal homogeneity of the ACR IV detectors relative to previous detectors, which were fabricated from sheet silver. Detector time response is decreased, and uniformity of detector sensitivity to radiant and electrical heating is improved. The ACR IV's 5 deg field of view is defined by an extension of the heatsink. The isothermally defined field of view assures constant radiative transfer at all times between the cavity and the remainder of its field of view. The ACR IV field of view can be varied from
Pc = p,
(2)
where ps = surface reflectance (Ps = 1 - a5 ), a, = surface absorptance, Pc = effective cavity reflectance, and 6 = number of reflections for axial rays.
In contrast to cavities with diffuse surfaces, the requirement for high surface absorptance with small un-
a minimum just larger than the radiation source to be measured to nearly 2-xsr. The desirability of using a small field of view arises from uncertainty introduced into measurements by the difference in effective radiative temperature of the field of view in the shutter open and shutter closed phases of
measurement. This source of uncertainty propagates as the square of the field of view. The uncertainty contribution for a 5 deg field of view is small and, when
certainty is considerably relaxed: a surface absorptance
analytically accounted for in data analysis, becomes
of 0.9 with an uncertainty of ±5% yields a theoretical cavity absorptance of 0.999999with an uncertainty of
negligible. The choice of 5 deg, with a 1-deg tolerance
±0.0003%!
The advantage of the specular cavity relative to the diffuse types used in previous ACRs is apparent from
for pointing error, corresponds closely to that used in current instrumentation designed for direct solar irradiance observations at the earth's surface, facilitating direct intercomparisons without uncertainties intro-
the following. For the ACR III, 3M velvet diffuse black was employed as the cavity surface material. Its surface absorptance for solar flux was 0.95 with an uncertainty of ±0.02. The shaded ACR III cavity, which has nearly
duced by differing fields of view. III.
Analysis of the ACR IV
the same geometry as the ACR IV cavity, produced an effective cavity absorptance of 0.997 i 0.002. The
A.
Basic Considerations
uncertainty in effective cavity absorptance for solar flux,
respect to fundamental physical concepts can be carried
one of the most important parameters of absolute pyrheliometry, is decreased from 0.2%to 0.0003% by using a specular absorber, even though its surface absorptance
uncertainty is 2.5 times larger than that of the diffuse type. While the above comparison applies to perfectly formed cavities, which cannot be realized, considerable
departure from the ideal can be tolerated with the specular cavity before its absorptance uncertainty reaches the 0.01% level.
The dual-cavity configuration minimizes ACR sensitivity to heatsink temperature drift. The secondary cavity views the heatsink and has thermal properties nearly identical to the primary detecting cavity. The cavity-heatsink temperature difference is monitored between the top of both cavity-thermal impedance structures in the ACR IV as contrasted with the top and bottom of the single cavity-thermal impedance structure of the ACR III. The primary cavity is in equilib-
The definition of the absolute radiation scale with out by employing either standard detectors (pyrheliometers) or standard sources. Standard sources are chiefly employed in calibrating low-level total irradiance
measurements and as standards of spectral radiance or irradiance. Pyrheliometers provide the most accurate means of defining the absolute radiation scale at the solar constant level in the International System of Units. Pyrheliometers have been constructed in various forms, but they all have certain features in common. They are calorimeters in which the heating effect of unknown irradiant flux on a detector is compared with that of an electrical current passed through a heating element placed in intimate thermal contact with the detector. An accurate knowledge of the effective absorptance of the detector for the irradiant flux, the area over which the detector is illuminated, and the electrical heating power facilitates the accurate measurement of 15 January 1979 / Vol. 18, No. 2 / APPLIEDOPTICS
181
P = power conducted from cavity to heatsink through the thermal impedance, 'C = power required to maintain cavity's thermodynamic state, ' = rate of cavity (and heatsink) temperature drift,
A
irradiant fluxes on an absolute basis in SI units.
description of the analytical approach to pyrheliometry used in the development of ACRs follows.
2
Derivation of the ACR IV Quasi-Equilibrium
B.
C = cavity heat capacity, j Prcj = sum of ir power outputs radiated by
the cavity to its various fields of
Equation
view,
The starting point in the following analysis of the
2 j Pacj = sum of power outputs by air conE
ACR physics is the concept of viewing the ACR IV primary cavity as an independent thermodynamic device. We account for all the cavity power sources and sinks
in the reference (shutter closed) and observation (shutter open) phases of measurement. In the reference phase, the sums of cavity power in-
2;k Pick
=
duction to various parts of the cavity field of view, sum of cavity power outputs by con-
duction through electrical leads,
AcH(ac + PPc) =
solar irradiancepowerinput to cavity,
puts and outputs are
Ac = area of detector's primary aperture,
H = solar irradiance,
(3)
SPin=Pr + E Pri.
ac =
effective cavity absorptance for solar
irradiance, XPout = PC+ 1'C + E (P.cj + Pacj) + E Pick. k j
(4)
Pc = cavity reflectance
In the observation phase, the sums of cavity power L1ilputs and outputs are
initially absorbed by the cavity, back
to the cavity. The The quasi-equilibrium condition implies that
EP'n = Peo+ AcH(ac + Ppc)+ E P, Plck (P;Cj+ Pacj)+ E lPyut = PC+ T"C+ E k j
in the reference phase FPin _ FPout,
(6)
and in the observation phase
where Per, Peo = electrical reference and observation
phase cavity heating powers, Zj Prj = sum of ir power inputs radiated to the cavity by its various fields of view,
ACR IV quasi-equilibrium
equation
H = [Ac(ac+ ppc)- (Per- Peo)+
[Ac(ac+ppc)]-' [C( - V) + A, (Prcj -
Prcj) -
E (Prj - Pj)+ (Pc - Pc) +
+ E (Pac- Pacj) E (Pck -P'lck)J
_Pin
ZPout.
(7) (8)
Equating Eq. (3) to Eq. (4) and Eq. (5) to Eq. (6), solving
for the electrical heating powers, and putting the resulting expressions into the form of Eq. (1) yield the quasi-equilibrium equation for the ACR IV:
Term description (electrical power)
(cavity heat capacity) (radiation from cavity to surroundings)
(radiation to cavity from surroundings) (conductance to heat sink by thermal impedance) (air conduction)
(electrical lead conduction).
kI
182
for solar irra-
diance, p = rereflectance of solar radiation not
APPLIEDOPTICS/ Vol. 18, No. 2 / 15 January 1979
(9)
1.0
of the cavity's field of view p (excluding the field of view
KEY: VR - Cavity heatervoltage, referece phae V
ac.Cavity 0.5
through the 5-deg view limiter) are shown. The term
_ Cavity heater voltageobservation phase
for the difference of effective radiative temperature seen
sIar abs ptnce
Ac -Cavity
by the cavity through its view limiting aperture (fieldof-view 1) in the shutter open and closed phases is (Prl - P'ri). The combination of a small field of view and
prieary perture area
TOTAL