Aerosol Optical Depths from Airborne Sunphotometry in INTEX-B/MILAGRO as a Validation Tool for the Ozone Monitoring Instrument (OMI) on Aura

J. Livingston1, J. Redemann2, P. Russell3, S. Ramirez2, Q. Zhang2, O. Torres4, A. Smirnov4, B. Holben5, E. Browell6, J. Hair6, Y. Shinozuka7, C. McNaughton7, A. Clarke7

1SRI International, Menlo Park, CA, 2Bay Area Environmental Research Institute, Sonoma, CA, 3NASA Ames Research Center, Moffett Field, CA, 4UMBC/NASA Goddard Space Flight Center, Greenbelt, MD, 5NASA Goddard Space Flight Center, Greenbelt, MD, 6NASA Langley Research Center, Hampton, VA, 7University of Hawaii, Hilo, HI

Current and Future Analyses

We have identified four J31-AATS/Aura-OMI coincident or near-coincident MILAGRO/INTEX-B measurement events for which OMI AOD spectra have been retrieved using the UV retrieval algorithm and AATS AOD retrievals have been calculated. Three of these (March 3, 10, and 17) were over the Gulf of Mexico, and one (March 19) was over Mexico City.

For each of these events, we have compared OMI AOD retrievals with corresponding AATS AOD retrievals derived from measurements obtained during minimum altitude J31 flight legs. In general, OMI AOD retrievals exceed corresponding AATS AOD retrievals.

We are currently in the process of investigating the differences between the OMI and AATS AOD retrievals. The OMI UV retrieval algorithm is particularly sensitive to absorbing aerosol, and relatively insensitive to non-absorbing aerosol. Future work will likely involve the inclusion of other information, such as coincident measurements of water leaving radiance and/or observations of sea surface state, to the UV retrieval algorithm. We also plan to extend these comparisons by using an alternate OMI AOD retrieval algorithm that uses measurements at multiple wavelengths.

The March 19 event over Mexico City is quite complicated due in large part to the effect of the complex terrain, but it is a rich data set to be studied due to coincident remote and in-situ measurements available from the DC-8, in addition to measurements of surface albedo (J31 SSFR) and BRDF (J31 RSP and CAR), and retrievals of AOD and other aerosol optical properties by AERONET at three ground sites.

Abstract

Aerosol data products produced by the Ozone Monitoring Instrument (OMI) on the Aura satellite include aerosol extinction optical depth (AOD) and absorption optical depth (AAOD), in addition to the aerosol index. AAOD is the most reliable OMI aerosol parameter. These OMI aerosol products are derived using wavelengths and algorithms that differ significantly from those of other nadir-viewing satellite aerosol-measuring instruments, such as MODIS and MISR. These differences produce some advantages (such as high sensitivity to aerosol absorption and aerosol retrievals over bright surfaces and clouds) and disadvantages (such as dependence on aerosol layer height) that lead to unique needs when validating and improving OMI aerosol retrievals. An example is the desire for measurements of aerosol vertical profiles in conjunction with AOD.

In March 2006 during INTEX-B/MILAGRO (Phase B of the Intercontinental Chemical Transport Experiment/Megacity Initiative-Local And Global Research Observations) the 14-channel Ames Airborne Tracking Sunphotometer (AATS-14) flew on the Jetstream 31 (J31) aircraft. AATS measured AOD at 13 wavelengths (354-2139 nm) and water vapor columns in 13 flights based in Veracruz, Mexico, sampling clean and polluted airmasses over the Gulf of Mexico and Mexico City. Vertical differentiation of AOD and water vapor column data from J31 vertical profiles yields vertical profiles of multiwavelength aerosol extinction and water vapor density. J31 flights were coordinated with overflights of several satellites, including Aqua, Aura, Terra, and Parasol, plus other aircraft, including the NASA DC-8 and King Air and the NCAR C-130. These coordinated flights produced a very rich data set with strong potential for OMI aerosol validation studies during J31-OMI coincidences.

We have identified four Aura overpasses during MILAGRO/INTEX-B when both AATS and OMI AOD spectra have been retrieved at coincident or near-coincident times and locations. Three of these (March 3, 10 and 17) were over the Gulf of Mexico, and one (March 19) was over Mexico City. In this poster we present results from preliminary comparisons of AATS AOD retrievals and OMI AOD retrievals (using the UV retrieval algorithm) for each of these events. During the 19 March flight, AATS measured vertical profiles and horizontal transects of AOD at the T2 site NNE of Mexico City and at the T0 site in the heart of the City, both of which were in OMI grid cells where aerosol retrievals were performed. The suborbital data set is particularly rich, including AERONET retrievals of aerosol properties from T2 and T0, additional aerosol retrievals from radiometers on the J31, and lidar and in situ measurements from the DC-8. This data set provides not only AOD values for comparison to OMI results, but information on aerosol height, size, and composition for constraining the OMI aerosol retrieval model.

19 March 2006

Photo from DC-8 over Mexico City

Aeronet sites

T2 N of T1

“Visibility poor. Eyes burning in cockpit.”

J31/AATS near T0 ~20.14-20.48 UT

“Much clearer as we approach T0.”~480 m above T0,

~510 m AGL

Comparison of AATS, OMI, and Aeronet AOD

Coordination of J31 and DC-8 flights

AATS measurements of AOD and CWV; J31 altitude

J31 flight track and OMI AOD grid cells

DC-8 over J31

19:25 UT 20:15 UT DC-8 over J31

between T1 & T0DC-8 over J31

NW of T0

20:28 UT

at T2DC-8 over J31 Large depolarization suggests dust: consistent

with AATS-14 AOD wavelength dependenceIn-situ aerosol scattering measurementsalso suggest dust in the boundary layer

Aura overpass: 20.11 UT

Results shown in the middle frame above use the same version of the retrieval algorithm as applied to the March 3 and the March 17 measurements. There are no acceptable OMI retrievals in grid cells coinciding with the J31/AATS near-surface transect 19.68-20.02 UT. However, OMI AOD retrievals for the four grid cells closest to the J31 minimum altitude leg yield the best agreement with the corresponding AATS measurements. The bottom frame shows results using a modified version of the algorithm that yields a larger number of acceptable AOD retrievals. One of the changes assumes the aerosols are non-absorbing for small AOD. The OMI grid cells with the highest retrieval quality (qual=0, see cells within oval) actually yield AODs less than the corresponding AATS AODs.

OMI retrievals of AOD using the UV technique are particularly difficult for background aerosol conditions such as those that prevailed during the three over-water MILAGRO/INTEX-B J31-AATS/Aura-OMI comparisons on March 3, 10 and 17. For these cases, the algorithm yielded a limited number of OMI grid cells with acceptable AOD retrievals due to the presence of clouds and/or other algorithmic conditions that deemed the retrievals not fully reliable. Here we compare OMI and AATS AODs for J31 near-surface horizontal transects during the time periods 18.84-19.02 UT and 19.39-19.48 UT. The latter is much closer in time to the Aura overpass at ~20.12 UT.

OMI AOD over-water retrievals from measurements obtained during the March 17 Aura overpass at ~20.28 UT are compared with AATS AOD retrievals from measurements acquired during the J31 near-surface transect during the period 19.42-19.53 UT. All OMI retrievals within 120 km of the mean J31 location during the low level leg are shown. The version of the OMI retrieval algorithm used here is the same as that used to produce the results shown for March 3 and for the middle frame of the March 10 presentation.

T2-----------------T1--------------------T0--------------------------

Aura/OMI overpass ~15 minutes later at 20:07 UT

3 March 2006

Aura/OMI overpass ~15 minutes later at 20:07 UT

2 grid cells intersect J31 near surface transect

18.84-19.02 UT

J31/AATS measurements

J31 near surface transect 19.39-19.48 UT

J31/AATS measurements

10 March 2006

Aura/OMI overpass ~15 minutes later at 20:13 UT

19.68-20.02 UT

19.68-20.02 UT

J31/AATS measurements

17 March 2006

19.42-19.53 UT

Aura/OMI overpass ~15 minutes later at 20:19 UT

J31/AATS measurements

AATS ~460 m AGL: above much of AOD

Surface albedo effect?