Equipment
Adelaide
Davis Station Radar
Mid and low latitude mesospheric temperature estimation using meteor radar and OH rotational temperatures
Adelaide
Iain M. Reid1,3, David A. Holdsworth1,2,3, Jonathan Woithe3, Daniel McIntosh1, Abas Sivjee4, Ray J. Morris2, Damian J. Murphy2, Gary B. Burns2 and W. John R. French2 2Australian Government Antarctic Division, Kingston, Tasmania, Australia Australia 3ATRAD Pty Ltd, Thebarton,, Thebarton
South Australia, Australia
MF Doppler / SA Radar
MF SA Radar
VHF Meteor Radar (31.0 MHz) (also deployed on campaign basis)
VHF Meteor Radar (around 31.0 MHz) from early 2005
Optical
4Embry--Riddle Aeronautical University, Daytona Beach, FL, USA Embry
Davis: ST 55 MHz array
[1] Email:
[email protected]
VHF Meteor / Atmospheric (M)ST Radar (55.0 MHz)
VHF Boundary Layer Radar
Davis
Davis: Single 33 MHz meteor receive antenna
1School of Chemistry and Physics, University of Adelaide, Adelaide, Adelaide, Australia
VHF Meteor / Atmospheric ST radar (55.0 MHz)
Darwin
OH, O2 Spectrometer
BP: Single 55 MHz meteor transmit antenna
(with Embry Riddle) 3FP (558 nm and 730 nm)
CzernyCzerny-Turner Spectrometer (OH band) 3FP (558 and 630 nm)
AllAll-sky airglow Imager (OH and O2)
BP 55 MHz array
(with Aerospace Corp) Rayleigh Lidar / Na Lidar (under development)
JAS007, Response of the atmosphere/ionosphere coupling system to forcing from the Sun and the lower atmosphere
Rayleigh Lidar
IUGG Meeting Perugia, Italy, July 2-13, 2007
Meteor distributions at 55 (BP) and 33 MHz (Darwin) -- typical day
Meteor temperature estimation
Total unambiguous events 33 MHz: 19,153 55 MHz: 11,900
The decay time t of an underdense meteor echo is given by [e.g. McKinley, 1961] 1961]
t= 92 km
λ2
16π 2 D
(1)
where λ is the radar wavelength, and D is the ambipolar diffusion coefficient [e.g. Cervera and Reid, Reid, 2000]
89 km
Daily meteor counts 25000
daily counts BP daily counts Darwin
20000
T2 P
(2)
and K0 is the zero field reduced mobility factor. factor.
15000 10000 5000 0 2006
D = 6.39 × 10 − 2 K 0
2007
Temperature can be estimated from (2) by using D with an assumed pressure climatology [e.g. Hall et al., 2004]: 2004]: hereafter pressure climatology temperatures. temperatures.
Meteor temperature estimation Temperature can also be estimated using a temperature gradient climatology [e.g. Hocking et al., 1999]: 1999]:
T = S(
mg dT + 2 ) log10 e dz k
where S is the slope estimated from the scatter plot d = log10 D versus height, dT/dz is the temperature gradient, m is the mass of a typical atmospheric particle, g is the acceleration due to gravity, and k is Boltzmann's constant: hereafter temperature gradient climatology temperatures.
Buckland Park O2 (top) and OH (bottom) rotational temperatures. Yellow lines are harmonic fits OH: DC component: amplitude = 192 K 180 day period: amplitude = 2.8 K, phase = 60 days 365 day period: amplitude = 7.2 K, phase = 96 days M.G. Shepherd et al jgr, jgr, 109, doi:10.1029/2004jd005054, 2004
35S, 87 km: mean 192+/192+/-8.5; annual amp, 5.8+/5.8+/- 0.3; annual phase, 151+/151+/- 15.7 days; semisemi-annual amp, 1.8+/1.8+/-1.2; phase 118.7+/118.7+/- 8.0.
1
Davis Temperature climatology
Davis Pressure climatology CIRA (blue line) MSIS (green line)
Temperature and temperature gradient climatologies at the peak height height of the meteor height distribution (red dashed line) for Davis Davis OH spectrometer (black pluses) Meteor height distribution weighted estimates based on the 69o N rocket climatologies* of Lübken and von Zahn [1991] (blue triangles) and Lübken [1999] (orange line) and Syowa lidar [Kawahara et al., 2002] (green squares). Temperature gradient climatology* derived for 69o N by Hocking et. al [2004] (shifted by 182.5 days) (purple line)
Davis 55 MHz radar: 55-day temperatures
Temperature estimates made using the temperature gradient technique technique (left), the pressure climatology technique (middle). The right plot indicates the pressure climatology temperatures obtained obtained using CIRA (green) and MSIS (blue) pressure climatologies. The red line in all plots indicate the climatology temperature (red (red line).
Comparison of the same meteors detected at 33 and 55 MHz (April 2005)
Climatology (red line) derived using Northern hemisphere rocket climatologies (shifted by 182.5 days) of Lübken and von Zahn [1991] (squares) and Lübken [1999] (diamonds).
Determination of pressure climatologies at the peak height of the meteor height distribution climatology for the Davis MST radar
Davis 33 MHz radar daily temperatures
Temperature estimates made using the temperature gradient climatology climatology technique (left), the pressure climatology technique (right). The red line indicates indicates the temperature climatology. Temperatures underestimated with temperature gradient climatology; overestimated with pressure climatology To investigate this result, we examine the results for meteor trails trails simultaneously observed by both 33 and 55 MHz radars (that is for the same meteors meteors detected by the two radars)
Simultaneously observed 33 and 55 MHz echoes (April 2005) 55 MHz echoes observed at greater heights (0.36 +/+/- 1.44 km). . Similar result obtained comparing Buckland Park (Australia) 31 and 55 MHz meteor radar diffusion coefficients, which use the same (ATRAD) meteor analysis software as the Davis radars. Similar result also obtained comparing Andenes (Norway) 32.5 and 53.5 MHz meteor radar diffusion coefficients, which use different (SkiYMet) SkiYMet) meteor analysis software to the Davis radars.
Scatter plots of height, (left) and diffusion coefficient (middle): (middle): xx-axis 55 MHz, yy-axis 33 MHz. Grey line indicates y = x.
Have compared individual echoes to investigate this apparent contradiction and find:
55 MHz echoes observed at greater heights (0.36 +/+/- 1.44 km). 33 MHz diffusion coefficients larger than 55 MHz diffusion coefficients, coefficients, thus explaining larger 33 MHz pressure climatology temperatures.
2
Simultaneously observed 33 and 55 MHz echoes (April 2005)
Possible reasons
33 MHz echoes amplitude variation show smaller decay times (in t/λ2 units) and hence larger diffusion coefficients. Echo phase variations show good agreement. The diffusion coefficient difference is clearly inconsistent with with theory, which assumes the diffusion coefficient is independent of frequency. Parameterizing the difference in the 33 and 55 MHz diffusion coefficients we observe that the difference: 1) decreases with increasing SNR. 2) maximizes in midmid-summer at around 82 km. km. 3) appears independent of all other measurable parameters (local time, angle of arrival, range, radial velocity), and radar parameters (e.g. transmit power, transmit polarization, etc).
Singer has recently examined a number of meteor radars operating at both high northern and southern latitudes and also at a low latitude southern hemisphere site He concludes that there is a difference between the diffusion coefficients determined from strong and weak meteor echoes, and hence a difference in the temperatures determined from them He attributes this effect to the presence of meteoritic dust, which which retards the diffusion and increases the diffusion coefficient This effect would also appear as a frequency dependent effect and so provides a possible explanation
Daily mean temperatures
Buckland park daily mean temperatures Daily M ean Te mpe ratures
O2 temperatures (95 km) Harmonic fit O2 temperatures
Harmonic fit O2 temperatures
55 MHz meteor temperatures (89 km)
31 MHz meteor temperatures (92 km)
MSIS90e 94 km temperatures
240 220
MSIS90e 94 km temperatures
240
200 Temperature / K
Temperature / K
55 MHz meteor temperatures (89 km)
260
SABER 2006 temp
180 160 140
31 MHz
120 100 2001
2002
55 MHz
2003
2004
2005
SABER 2006 temp
220 200 180 160
55 MHz
140
2006
120
2007
100 2005
Year
Daily M ean Te mpe ratures
2006
2007
Year
OH temperatures (87 km) Harmonic fit OH temperatures 31 MHz meteor temperatures (92 km)
260
Daily Me an Temperature s
OH temperatures (87 km) Harmonic fit OH temperatures 31 MHz meteor temperatures (92 km) 55 MHz meteor temperatures (89 km) MSIS90e 87 km temperature SABER 2006 temp AURA_MET_estimate 90 KM
55 MHz meteor temperatures (89 km)
240
MSIS90e 87 km temperature 260
SABER 2006 temp
220
240
200 220
180
Temperature / K
Temperature / K
O2 temperatures (95 km)
Daily Me an Temperature s
31 MHz meteor temperatures (92 km) 260
160 140
200 180 160
120 140
100 2001
2002
2003
2004
2005
2006
2007
Year
The meteor results are pressure model temperatures, obtained using using P_msis * x 1.2. 1.2. The 1.2 factor is derived from the mean "non"non-summer" correction required to get Davis 55 MHz temperatures in agreement with cocolocated OH temperatures. The fact that T_33 > T_55 is consistent with Davis results, and is related to the fact that that D_33 > D_55
Buckland Park Temperatures AURA-MLS 01LT
TOH
A ltitu d e / k m
SABRE data Jan-Feb 2006
40
55 MHz meteor temperature Feb 2006
20 31 MHz meteor temperature Jan Feb 2003 55 MHz Jan 2006 aura pressure model
0 150
170
190
210
230
250
270
290
Temperature /K Parameter and study
Mean
SAO
AO
O2 temperature (long term average)
201 K
7.4 K day 78
1.9 K day 359
OH temperature (long term average)
192 K
2.8 K day 60
7.2 K day 96
WINDII OH temperature (Shepherd (Shepherd et al., 2004)
192 K
1.8 K day 119
5.8 K day 151
2006
2006.5
2007
Year
Temperature estimates made using the Davis MST (55 MHz) radar show show very good agreement with temperature climatology.
In the figure to the left, rotational temperatures are nightly averages for all clear nights in January and February 2006. The error bars indicate the variation of temperature over that observing period. AURAMLS observations are for two local times: 01LT and 13LT and for overpasses within 500 km averaged for the same period.
This result is attributable to larger diffusion coefficients estimated estimated by the 33 MHz system
TO2
AURA-MLS 13LT
2005.5
There is good agreement between the mean TOH and WINDII satellite observations (see table below)
80
60
100 2005
Summary and conclusions
TO2, TOH, Meteor, AURA-MLS and SABER Temperatures Jan-Feb 2006
100
120
The AURA-MLS temperature results are preliminary and are recognized to be underestimates in the MLT region (see Froidevaux et al., IEEE Trans. Geoscience and Remote Sensing, 44, 1106-1120, 2006)
Applying the same temperature estimation technique using the Davis Davis meteor (33 MHz) radar yields overestimated pressure climatology temperatures, and underestimated underestimated temperature gradient climatology temperatures. The magnitude of the difference between the 33 and 55 MHz diffusion diffusion coefficients is dependent on SNR, height and time of year. TOH and TO2 rotational temperature observations have been made at the Buckland Park site since 2001. The BP TOH temperatures agree well with WINDII temperatures AURA-MLS temperatures appear to underestimate the BP TOH and TO2 temperatures. Meteor derived temperatures show similar behaviour to TO2 and TOH at BP, but there is an outstanding question on the assumptions used to derive temperatures from the meteor trail diffusion coefficients Acknowledgements Funding for the 55 MHz meteor radar addadd-on: ASAC project 2529, ATRAD, AGAD and ARC Grant 20049700 Funding for the Davis ST radar: ASAC project 2325 and AGAD Funding for the 33 MHz meteor radar: ASAC project 2668 and AGAD Funding for the 55 MHz meteor radar: ARC Grant 20049700, ATRAD, ASAC project 2325 (loan of VTXVTX-6 Transmitter and high power feeder cables), ASAC Project 2529 (development of new 55 MHz meteor transmit antenna and antenna switching intended for Davis MST radar)
3