Estimation of a regional land surface albedo using narrow band satellite radiance data

The primary objective of the study is to derive an algorithm for regional albedo estimation using narrow band and moderate spatial resolution satellite data. A “uniform spectral region reflectance” model for converting satellite derived reflectance (emphasis on NOAA AVHRR) to broader spectral regions in the visible (0.38-0.72 μm), near-IR (0.72-1.30 μm) and shortwave mid-IR (1.30-4.0 μm) was further developed. Analysis of spectra from surface grasses indicated the usage of a satellite waveband reflectance to directly estimate a wideband visible and wide-band near-IR may produce errors to over 30%, when not incorporating a linear correction. For sensors such as NOAA AVHRR without a shortwave mid-IR band, visible and near-IR reflectivities may be used to estimate a mid-IR reflectance to reduce albedo errors to 20%. Examination of atmospheric data indicates that for many different scattering atmospheres there is a compensating effect between diffuse and direct radiation providing a small variation between spectral regions for global radiation, improving the algorithm capability to provide regional albedo estimation