Spatial prediction of Soil physical properties using Infrared Spectroscopy and GIS

Abstract

Soil physical properties include soil consistency limits and soil aggregate stability indices. These properties have influence on the soil agronomic characteristics as well as management of watershed hydrology, civil constructions and modeling environmental quality. The current methods for their determination are expensive and time consuming. There is a need for rapid and fairly accurate methods that can guarantee speed and allow comparison of point-measurements over time and space. This study developed a new approach for predicting soil physical properties using infrared spectroscopy and GIS. The study involved measurement of the soil physical properties using near infrared spectroscopy and mechanical methods. The calibration of the spectral values with the mechanical values developed a calibration model coefficient of determination of 0.78, 0.81 and 0.88 for aggregate stability index, liquid limit and plastic limit respectively. Spectral reflectance was found to be a reliable surrogate predictor of the soil physical properties.

A prediction model was established for estimation of the soil physical properties using Kriging method and implemented in R software (R Development Core Team, 2008, www.cran.r-project.org). The findings give a visual and numerical prediction of the soil physical properties in the upper Athi river watershed, eastern Kenya. A map of the physical properties of soil was produced and is expected to provide useful insights for planning civil constructions, control of land degradation, and for environmental conservation in upper Athi river watershed. This study should be tested with high resolution data sources and other related models which can improve the accuracy of the input data as well as prediction of the soil physical properties. Further testing and worldwide application with different models in different soil types and watersheds is highly recommended.