Simulation Of Atmospheric Correction Algorithm For Asian Absorptive Aerosols

This dissertation gives a brief introduction to the recent development of ocean color remote sensing and atmosphere aerosols. The atmospheric correction algorithm,bio-optical algorithms and optical properties of aerosols (especially absorbtive aerosols) are also described. To investigate the failure of standard atmospheric correction algorithm in the East Asian area, the simulation of atmospheric correction algorithm for absorptive aerosols are performed.Radiative transfer properties of absorbing aerosol, non-absorbing aerosol and cloud were analyzed, a simple, fast method was proposed based on the conception of the normalized difference vegetation index (NDVI). A near-infrared band (865nm) and a visible band (510nm) were used to calculate the normalized difference aerosol index (NDAI). NDAI values were used to distinguish between different aerosol types and concentration, such as dust, soot and non-absorptive maritime aerosols. Radiative transfer model Rstar5b was used for simulation analysis, to investigate the influence of contribution from water body. This method was successfully applied to several satellite images from SeaWiFS.The ACE-Asia in situ measurements are used to compare with the match-up SeaWiFS-derived products (processed by SeaDAS). The results confirm that the phenomenon of under-estimation of water-leaving reflectance in shorter blue wavelengths is very common in the East-Asian region. Based on in situ measurements, soot,dust particles are added to the SeaWiFS standard aerosol models separately, then an iterative radiative transfer simulation is carried out to select the best combination of particles. The resulting water-leaving reflectance and aerosol optical thickness agreed well with in situ measurements when the effect of soot or dust particle is considered in the simulation. For retrieving water-leaving reflectance, soot aerosol is better; but dust aerosol can improve the retrieval of aerosol optical thickness. While adding both soot and dust to the SeawiFS standard aerosol models does not show good performance. Then a set of absorbtive aerosol models with soot are selected based on simulation with and without in situ measurements. They are appropriate for the Asian aerosols covered in this study. An area polluted by dust is used to validate the soot-absorbtive aerosol models. The chlorophyll a concentrations are calculated with the soot-absorbtive aerosol models. The results are better than the standard SeaWiFS algorithm in terms of chlorophyll a retrieval.