Lyapustin, Alexei (Alex): NASA GSFC (Project Lead)
Project Funding:
2011 - 2015
NRA: 2009 NASA: The Science of Terra and Aqua
Funded by NASA
Abstract:
This proposal addresses the ROSES-2009 call A.41 "The Science of TERRA and AQUA", sec. 2.3 (Algorithms - New Data Products) for the new algorithms which successfully passed ATBD review. The PI is a current member of NASA MODIS, NPP and NOAA GOES-R science teams, of the VIIRS Operational Algorithm Team (VOAT) and of NASA's GeoCape Mission Formulation Team.
This proposal capitalizes on our progress on the MODIS data processing algorithms developed at NASA under funding of Dr. D. Wickland. We have developed a new Multi-Angle Implementation of Atmospheric Correction (MAIAC) algorithm for MODIS. It uses a time series approach and an image-based rather than pixel-based processing to perform simultaneous retrievals of atmospheric aerosols and surface spectral bi-directional reflectance/albedo without empirical assumptions typical of current operational algorithms. The new algorithm is generic and works over vegetated regions of the Earth as well as over bright deserts. The aerosol retrievals are performed at high 1 km resolution which is highly requested in different science and application disciplines, such as Air Quality/Urban Pollution. In snow-covered regions, MAIAC provides an estimate of the snow grain size and sub-pixel snow fraction. MAIAC has an advanced cloud mask (CM) and an internal dynamic land-water-snow classification which helps algorithm to flexibly choose processing path in changing conditions. MAIAC has been extensively tested on the local and regional scales.
A recently held NASA (Science Mission Directorate's Earth Science Division) ATBD Review Panel rated MAIAC ATBD as "A" suggesting that "the theoretical basis for the algorithm is sound, and the algorithm is ready for coding and production." The panel has recommended a "demonstration of the global product of some duration" with the following "systematic and transparent" comparison study between MAIAC and the current operational MODIS atmospheric correction algorithm (MOD09) involving the broad land community. Following these recommendations, we propose the following activities:
1) Resolve remaining algorithm issues, namely: implement retrievals over inland water bodies and coastal regions; add full polarization correction; finalize global aerosol climatology for operational processing; remove relatively small but systematic aerosol retrieval biases over bright surfaces.
2) Develop an operational version of MAIAC compatible with MODIS Adaptive Processing System (MODAPS) and demonstrate operational feasibility of algorithm.
3) Perform validation of MAIAC products. Aerosol validation will continue based on AERONET data. Surface reflectance/albedo will be validated with a) AERONET-based Surface Reflectance Validation Network (ASRVN) data record; b) available ground measurements (through established collaboration with groups of N. Coops and T. Hilker (Canada), A. Gitelson (Nebraska), E. Dutton (SurfRad) and others).
4) Perform operational evaluation of MAIAC in MODAPS and conduct a systematic comparison study of MAIAC and MOD09/MCD43 surface reflectance products. The near-term algorithm testing on MODAPS will start in Spring 2011 and will include both regular cases, such as East Coast of USA, and "hard" cases stressing the algorithm. The "hard" tests will include areas of India with perpetual haze, South Africa during biomass burning season with high aerosol levels and rapid changes in landscape, agricultural region of the US mid-West with fast and high-amplitude vegetation cycle (green-up to harvest). The testing will expand in the Collection 6 timeframe. MAIAC products for the selected tiles will be made available to the Land and other communities for evaluation against the standard MODIS products through MODAPS.
More details may be found in the following project profile(s):
