The Measurements of Pollution in the Troposphere (MOPITT) is an eight-channel gas correlation radiometer to be launched on the EOS Terra spacecraft in 1999. The goal of the experiment is to support studies of the oxidizing capacity of the lower atmosphere on large scales by measuring the global distributions of carbon monoxide (CO) and methane (CH4) and, thus, will represent a significant advancement in the application of space-based remote sensing to global tropospheric chemistry research. The primary measurement objectives of MOPITT are: (1) to obtain CO profiles with a resolution of 22 km by 22 km horizontally, 3-4 km vertically and with an accuracy of 10% throughout the troposphere; (2) to obtain total CO column amount measurements with a horizontal resolution of 22 km by 22 km and an accuracy of 10%; (3) to measure total CH(CH4 column to an accuracy of 1%, with a horizontal resolution similar to that of the CO measurement (Drummond 1992).
MOPITT Level 1 data products include: (1) eight calibrated and geo-located instrument difference radiances for each stare (~ 400 ms); and (2) eight calibrated and geo-located instrument average radiances for each stare (~ 400 ms). MOPITT Level 2 data products include: (1) tropospheric CO profiles, which are currently defined as average mixing ratios of five tropospheric layers (1000 — 700 mb, 700 — 500 mb, 500 — 400 mb, 400 — 300 mb, 300 — 200 mb) for each nominal 22-km-by-22-km pixel; (2) total CO for each atmospheric column over a nominal area of 22-km-by-22-km; (3) total CH(CH4) for each atmospheric column over a nominal area of 22-km-by-22-km. The column amount of CH(CH4) will only be available on the sunlit side of the orbit as a standard Level 2 MOPITT product.
In any remote-sensing experiment, validation of algorithms and data products is essential to ensure the quality of the data products for archiving and use by scientific communities. To complement the validation activities of each Terra instrument team, the EOS Project Science Office developed a NASA Research Announcement (NRA) that was issued in March 1997. As a result of this NRA, a number of investigators using different instruments and techniques were selected to provide correlative measurements for MOPITT data validation (Wang et al. 1998; http://eospso.gsfc.nasa.gov/validation/valinfo.html).
It is important to have confidence in the correlative data and associated data-processing algorithms before they are used to validate the MOPITT data products. It is also useful to test the intercomparison techniques that are to be used in post-launch MOPITT data validation. The Pre-launch MOPITT Validation Exercise (Pre-MOVE) was a validation campaign at the Southern Great Plains (SGP) Cloud and Radiation Testbed (CART) site of the Department of Energy Atmospheric Radiation Measurement (DOE/ARM) program (Stokes and Schwartz 1994) in Lamont, Oklahoma, March 2-6, 1998. The primary reasons for conducting Pre-MOVE at the CART site include: (1) it is a well instrumented site resulting in good characterization of the surface and the atmosphere column; (2) CART facility instruments SORTI (Solar Radiance Transmission Interferometer) (http://www.arm.gov/docs/instruments/static/sorti.html) and AERI (Atmospheric Emitted Radiance Interferometer) (Revercomb et al. 1995), lidars, and radiosondes are available to us; and (3) excellent logistic support at the ARM CART site.
The primary goals of Pre-MOVE were:
(1) validate correlative measurement data-processing algorithms by comparing retrieved CO columns and tropospheric profiles from ground-based interferometers and spectrometers with in situ CO profile measurements by the NOAA/CMDL; and (2) test the MOPITT Airborne Test Radiometer (MATR) (Smith et al. 1998) and associated data-processing algorithm by comparing the retrieved CO profiles from MATR observations with aircraft in situ CO profile measurements by the NOAA/CMDL flask system; and (3) test intercomparison techniques and protocols and prepare for future validation experiments after the MOPITT launch in 1999.
All instruments or instrument types to be used for post-launch MOPITT data validation were part of the Pre-MOVE at the CART site, March 2-6, 1998. Each instrument and its measurements during Pre-MOVE are described in Table 1.
Table 1. Summary of Pre-MOVE Activities.
Although the Pre-MOVE data analysis by the MOPITT correlative team is still in progress, preliminary results are very encouraging. Figure 1 shows the CO and CH4 profiles measured by NOAA/CMDL using their aircraft sampling system on March 6, 1998 at the ARM site in Lamont, Oklahoma. The NOAA/CMDL aircraft sampling unit is an automated package that collects samples of air using a small pump and 20 glass flasks. The samples are then returned to the NOAA/CMDL laboratory in Boulder, Colorado for analysis. Two successful sets of profiles for CO, CH4, and CO2 were obtained from 1-8 km with a vertical resolution of ~0.3 km. As shown in Figure 1, there are clearly elevated CO and CH4 levels around 3 km in both morning and afternoon measurements, possibly as a result of convection. CO measurements made using the CMDL automated flask system have a typical precision of 1 ppbv (Novelli et al. 1994).
Figure 1. Aircraft in situ profiles of CO mixing ratio (top panel) and CH4 mixing ratio (bottom panel). Triangles represent morning (~ 10:00 AM local time) profiles, and squares represent afternoon (~1:00 PM local time) profiles. All data were obtained on March 6, 1998 (Paul Navello and Brad Gore).
Ground-based remote sensing measurements were made with three instruments: AERI, SORTI, and a grating spectrometer from UT. Figure 2 shows the retrieved CO profile using ground-based solar absorption FTIR measurements (SORTI) on March 3, 1998 with a spectral resolution of 0.013 cm. The retrieval was carried out by N. Jones of NIWA and N. Pougatchev of CNU (Pougatchev and Rinsland 1995). The agreement with the airborne in situ measurement is within about 10% in the middle troposphere, but the agreement is not satisfactory in the lower and upper troposphere. This could be due to the lack of coincidence (difference of three days) of both location and time for the interferometer observations and the in situ measurements. Therefore, the comparison here is more qualitative rather than quantitative. It could also mean that we need to further improve the algorithm and intercom-parison protocols, which is one of the goals of Pre-MOVE. All these issues are still being investigated.
Figure 2. Retrieved CO profile using ground-based solar absorption FTIR measurement (SORTI) on March 3, 1998. Black circles (squares) are the in situ CO profile obtained during the morning (afternoon) of March 6, 1998. Open circles, squares, and triangles are the retrieved CO profile using SORTI data (Nikita Pougatchev and Nicholas Jones)
Figure 3 shows the retrieved total CO column using the ground-based grating spectrometer (Yurganov et al. 1997) from UT during Pre-MOVE on March 3, 1998. As a comparison, the total CO column amounts retrieved from SORTI measurements on March 3, 1998 are also included in this figure. The total CO column retrieved from the grating spectrometer measurement using the nonlinear least square (NLLS) technique agrees fairly well with that from the SORTI measurement. Further analysis is in progress to understand the differences between the NLLS technique and the equivalent width (EQW) technique.
Figure 3. Retrieved total CO column using the ground-based grating spectrometer from the University of Toronto during Pre-MOVE on March 3, 1998 (triangles). Open triangles show the retrieved CO total column using the nonlinear least square (NLLS) technique. Filled triangles show the retrieved CO total column using the equivalent width (EQW) technique. As a comparison, the retrieved total CO column from the SORTI measurement on March 3, 1998 is shown (circles) (Leonid Yurganov, Eamonn Mckernan, and Boyd Tolton. SORTI spectra were processed by N. Pougatchev, C. Rinsland, B. Connor, and N. Jones using the NLLS technique).
The top panel of Figure 4 summarizes the total CO column amounts retrieved from AERI spectra by Wallace McMillan and Hui He of UMBC during Pre-MOVE. AERI is a CART site facility instrument in autonomous operation acquiring an up-looking atmospheric emission spectrum roughly every 10 minutes. Gaps in retrieved CO column amounts occur where cloudy spectra have been removed (dot-dash lines in Figure 4). CO retrievals under cloudy conditions would underestimate the total column. Likewise, retrieved CO columns close to cloudy periods could be cloud contaminated, and thus not representative of the CO column amounts retrieved during continuous cloud-free conditions (solid lines in Figure 4). Column CO retrievals were accomplished using a modified version of the prototype CO retrieval algorithm developed for the Atmospheric Infrared Sounder (AIRS) (McMillan et al. 1997). Although the retrieved quantity is total CO column, AERI measurements are most sensitive to CO in the boundary layer.
Figure 4. Total CO column densities retrieved from AERI spectra during Pre-MOVE, March 2-5, 1998, are presented in the upper panel. Dot-dash lines connect retrievals interrupted by cloudy sky scenes while solid lines connect continuous data points. AERI spectra were supplied by the ARM program. Temperature and water vapor profiles from AERI spectra were supplied by Bob Knuteson and Wayne Feltz, U. of Wisconsin. CO retrievals from the AERI spectra were performed by Wallace McMillan and Hui He, UMBC. The bottom panel shows an expanded view of the time period on March 3, 1998 (GMT) and compares the CO columns retrieved from AERI spectra to those retrieved from coincident ground-based SORTI and UT spectrometer measurements.
The bottom panel of Figure 4 zooms in on a particular time period of March 3 (GMT) when there were coincident measurements by the ground-based SORTI and UT grating spectrometer. The CO column amounts retrieved from measurements by the 3 different instruments are compared. We are encouraged by the initial general agreement of all these retrievals to +/-10%. Further improvements in algorithms and intercomparison approaches may lead to even better agreements. We note here that MOPITT is designed to provide tropospheric CO measurement with an accuracy of 10%.
A Pre-launch MOPITT validation exercise was successfully carried out at the DOE ARM site in Lamont, Oklahoma, March 2-6, 1998. Preliminary results are very encouraging. CO retrievals from ground-based interferometers and grating spectrometer measurements compare fairly well with measurements by the NOAA/CMDL automated sampler.
We have also started the planning of the second Pre-MOVE, Pre-MOVE II, to be conducted in the Boulder-Denver area in May 1999. The main objective is to further improve correlative measurement data quality and the data processing algorithm. We intend to conduct an end-to-end simulation of the validation process. By conducting the Pre-MOVE II in the Boulder-Denver area where the NOAA/CMDL is located, we expect to get more in situ CO profiles for comparison with remote-sensing measurements. We believe these two pre-launch validation exercises will prove to be very useful in the understanding and comparison of correlative measurements for post-launch MOPITT data validation.
There will be other MOPITT validation activities including development of an airborne simulator (MOPITT-A), which is being constructed in Canada at the Universities of Saskatchewan and Toronto. Initial test flights of that instrument are expected in the fall of 1999 on the ER-2 aircraft. These will be followed by validation flights for MOPITT and participation in field campaigns with other instrumentation.
We want to thank the DOE ARM program, especially Ted Cress, Doug Sisterson, and Donald Slater, for their help in the planning and execution of Pre-MOVE at the CART site in Lamont, Oklahoma. Their assistance and excellent logistics support were essential to the success of Pre-MOVE. MOPITT validation activities are funded by the Project Science Office at Goddard Space Flight Center (GSFC), the Canadian Space Agency (CSA), and the National Science and Engineering Research Council (NSERC) of Canada.
Drummond, J. R., 1992; Measurements of Pollution in the Troposphere (MOPITT). The use of EOS for Studies of Atmospheric Physics, J.C. Gille and G. Visconti, Eds., North Holland, Amsterdam, 77-101.
McMillan, W.W., L.L. Strow, W.L. Smith, H.E. Revercomb, H.L. Huang, A.M. Thompson, D.P. McNamara, and W.F. Ryan, 1997: Remote sensing of carbon monoxide over the continental United States on September 12-13, 1997. J. Geophys. Res., 102, 10695-10709.
Novelli, P. C., K. A. Masarie, P. P. Tans, and P. M. Lang, 1994: Recent changes in atmospheric carbon monoxide. Science, 263, 1587-1590.
Pougatchev, N.S., and C.P. Rinsland, 1995: Spectroscopic study of the seasonal variation of carbon monoxide vertical distribution above Kitt Peak. J. Geophys. Res., 100, 1409-1416.
Revercomb, H.E., R.O. Knuteson, W.L. Smith, F.A. Best, and R.G. Dedecker, 1995: Atmospheric Emitted Radiance Interferometer: Status and Water Vapor Continuum Results. Proceedings of the Fourth Atmospheric Radiation Measurement (ARM) Science Team Meeting, 259-264.
Smith, M., and S. Shertz, 1998: The Measurement of Pollution In The Troposphere (MOPITT) Airborne Test Radiometer (MATR). The Earth Observer, 10, No. 4, July/August 1998, 21-23.
Stokes, G.M., and S.E. Schwartz, 1994: The Atmospheric Radiation Measurement (ARM) Program: Programmatic background and design of the cloud and Radiation Testbed. Bull. Amer. Meteor. Soc., 75, 1201-1221.
Wang, J., J. Gille, P. Bailey, M. Smith, D. Edwards, J. Drummond, G. Davis, and H. Reichle, 1998: MOPITT Data Validation Plan, Version 4.0. Available from http://eospso.gsfc.nasa.gov/validation/valplans.html.
Yurganov, L., E.I. Grechko, and A.V. Dzhola, 1997: Variations of carbon monoxide density in the total atmospheric column over Russia between 1970 and 1995: Upward trend and disturbances, attributed to the influence of volcanic aerosols and forest fires. Geophys. Res. Lett., 24, 1231-1234.
Zhao, Y., Y. Kondo, F.J. Murcray, X. Liu, M. Koike, K. Kita, H. Nakajima, I. Murata, and K. Suzuki, 1997: Carbon monoxide column abundance and tropospheric concentrations retrieved from high resolution ground-based infrared solar spectra at 43.5 ° N over Japan. J. Geophys. Res., 102, 23403-23411.