Atmospheric Correction for Long Slant Path Imagery

Earth-observing Hyper- and Multispectral Imaging (HSI/MSI) in the Vis/SWIR provides detailed spectral signatures of natural terrain and man-made objects. However, the present inability to accurately correct for atmospheric effects in long slant paths seriously degrades target detection and discrimination, thus limiting surveillance and reconnaissance opportunities. Spectral Sciences, Inc. (SSI) proposes to develop innovative, automated, fast and accurate atmospheric correction software for the highly off-nadir, long slant path scenario. In Phase I, the main building blocks for the correction algorithm will be generated and tested. A combination of real and Monte Carlo-simulated imagery from the SSI MCScene model will be assembled and used to evaluate the AFRL/SSI FLAASH atmospheric correction code in the long slant path scenario and identify the key sources of reflectance error. In addition, one or more new, advanced algorithms for retrieving detailed aerosol properties will be developed and demonstrated. Based on the results, plans will be made for development in Phase II of an enhanced Long Slant-path version of FLAASH that includes the new aerosol retrieval algorithm(s) and an aerosol and view/solar angle-dependent adjacency correction, and for a field experiment designed to validate the enhanced correction capability.