Estimation of soil hydraulic properties using proximal spectral reflectance in visible, near-infrared, and shortwave-infrared (VIS-NIR-SWIR) region

Abstract

Characterization of soil hydraulic properties is an important step for assessing soil water regime in agricultural fields. Because direct measurement of soil hydraulic properties at multiple locations is costly and time-consuming, pedotransfer functions (PTF) are conveniently used to estimate these properties from easily measurable basic soil properties. Over the last two decades, several studies have demonstrated that basic soil properties of surface soils may be rapidly estimated by measuring soil spectral reflectance. In this study, we evaluated a PTF approach to use proximal spectral reflectance over the visible, near-infrared, and shortwaveinfrared (VIS–NIR–SWIR) region (350–2500 nm) as predictor variable in place of basic soil properties. To develop these transfer functions, spectral reflectance of air-dried and sieved soil samples was measured using a handheld spectroradiometer equipped with a contact probe. Transfer functions in the form of multiple linear regression relationships between soil hydraulic properties and different attributes of measured spectral reflectance were developed. These new transfer functions are called spectrotransfer functions (STF). Both the parametric PTFs and STFs for the parameters of van Genuchtenwater retention model (α and n) and point PTF for saturated hydraulic conductivity (Ks) were evaluated using the root-mean-squared error (RMSE). Results show that STFs have the similar accuracy as PTFs for estimating hydraulic properties. Specifically, STFs developed with the absorption features of proximal spectral reflectance performed better than the PTFs for estimating α. Among three hydraulic parameters for which the STFs were developed, van Genuchten parameter n is well predicted with comparatively lower values of RMSE. Thus, this study shows that proximal spectral reflectance of soil may be used for rapid estimation of soil hydraulic properties in a large area with accuracy comparable to PTFs. A rigorous testing in different geographical regions is warranted to establish the utility of STFs as a method for estimating soil hydraulic properties.