Application of a global aerosol forecast model for multi-spectral ocean color atmospheric corrections

The presence of tropospheric aerosols obscures the ocean and reduces the ability of satellites to retrieve ocean color products from water leaving radiances (Lw(λ)). NASA standard atmospheric corrections are based on the Gordon and Wang method and use 80 aerosol models with varying relative humidities and size fractions to account for the presence of aerosols; however, if the wrong aerosol model is selected, negative Lw(λ) are often returned. Our goal is to reduce the number of negative Lw(λ) by improving the model selection process through the incorporation of external data derived from the Navy Aerosol Analysis and Prediction System (NAAPS). NAAPS is a global aerosol forecast model capable of providing the location, composition and optical properties, including optical depth, of dust, smoke and sulfates in the atmosphere. We investigate application of NAAPS-derived composition and aerosol optical depths to select the appropriate aerosol model for retrievals using Moderate Resolution Imaging Spectroradiometer (MODIS) data. Spectral aerosol optical depths and water-leaving radiances from both the standard and NAAPS-based aerosol correction methods are compared to in situ NASA AERONET data for 25 MODIS images covering three regions with different atmospheric aerosol conditions: smoke over the Gulf of Mexico, pollution off the Northeast US Coast, and dust over the Arabian Gulf. Results will be discussed.