Soil Moisture Estimation From Remotely Sensed Hyperspectral Data

Abstract

A methodology for mapping surface soil moisture content across an agricultural field fromoptical remote sensing and ground sampling is developed. This study uses both ground-based andremotely sensed spectral measurements of soil reflectance in visible and near-infrared wavelengthsand concurrent measurements of volumetric soil moisture within the top 6 cm. After determiningappropriate wavelengths for soil moisture estimation from spectral reflectance, a cokriging schemewas used to generate soil moisture maps. Results indicate that combining reflectance and groundmeasurements can yield more detailed maps of soil moisture than ground measurement alone.