Adjusted Normalized Emissivity Method for surface temperature and emissivity retrieval from optical and thermal infrared remote sensing data

Abstract

A methodology for the retrieval of surface temperatures and emissivities combining visible, near infrared and thermal infrared remote sensing data was applied to Digital Airborne Imaging Spectrometer (DAIS) data and validated with coincident ground measurements acquired in a multiyear experiment held in an agricultural site in Barrax, Spain. The Adjusted Normalized Emissivity Method (ANEM) is based on the use of visible and near infrared data to estimate the vegetation cover and model the maximum emissivity according to the Vegetation Cover Method. The pixel‐dependent maximum emissivity is used as the initial guess of the Normalized Emissivity Method to obtain the surface temperature and emissivity from the thermal infrared data. ANEM allows adjusting the initial emissivity with regard to the spatial variation of emissivity with vegetation cover, instead of using a fixed emissivity. Surface temperatures derived with ANEM agreed well with ground data, with a standard deviation of ±0.8 K and nearly zero bias for all the surface types. Retrieved emissivities were mostly within ±0.01 of the measured values, despite certain instrumental problems apparent in the thermal part of DAIS. An analysis of the emissivity spectra was performed, showing the utility in the discrimination of different agricultural surface types in the area.