Evaluación de sensores hiperespectrales para la estimación del carbono orgánico del suelo

Abstract

The aim of this work is to identify the most suitable sensors and regions of the spectrum for estimation of soil organic carbon in southwest Asturias (northwest Spain). Three hyperspectral sensors (Hyperion, CASI, and AHS) and one multispectral, Landsat-5 TM, are evaluated. The reflectance spectra of a set of soil samples are resampled for the different sensors and PLSR (Partial Least Squares Regression) models of TOC are calibrated for several regions of the spectrum. The results are compared and the most important regions of the spectrum for estimation of TOC are determined.