Impact of Atmospheric Correction Methods Parametrization on Soil Organic Carbon Estimation Based on Hyperion Hyperspectral Data

Visible Near infrared and Shortwave Infrared (VNIR/SWIR, 400–2500 nm) remote sensing data is becoming a tool for topsoil properties mapping, bringing spatial information environmental modeling land use management. These estimates are based on regression models, linking key property to VNIR/SWIR reflectance data. Therefore, the model’s performances depend quality of both analysis (measured laboratory over-ground soil samples) Bottom-of-Atmosphere (BOA) which retrieved from Top-Of-Atmosphere radiance using atmospheric correction (AC) methods. This paper examines sensitivity organic carbon (SOC) estimation BOA images depending two parameters used in AC methods: aerosol optical depth (AOD) FLAASH (Fast Line-of-Sight Atmospheric Analysis Spectral Hypercubes) method water vapor (WV) ATCOR (ATmospheric CORrection) method. work was Earth Observing-1 Hyperion Hyperspectral acquired over cultivated area Australia 2006. were converted seven values AOD (from 0.2 1.4) six WV 0.4 5 cm), ATCOR, respectively. Then Partial Least Squares (PLSR) model built each estimate SOC bare pixels. study demonstrated that PLSR models insensitive variation method, with R2cv RMSEcv 0.79 0.4%, The slightly sensitive ranging 0.72 0.41 0.47. Regardless values, best parametrization provided similar prediction accuracy Variation did not impact identification pixels coverage corresponded 82.35% area, while 75.04 84.04%. this recommends (1) provide before mapping or (2) careful selection parameter when ATCOR.