Quantification of soil patterns and field soil fertility using spectral reflection and digital processing of aerial photographs

Abstract

Direct spectral reflectance measurements of soils and multi-emulsion analysis of aerial photographs were carried out to document possible uses of these techniques for soil and fertility management in agricultural fields where the within field soil variability is high. A spectrometer with four broadband filters and a multi-emulsion film scanner were used to quantify the soil conditions in an agricultural field where the soil pattern was made up of three distinctly different soil types. The spectral data from soil samples and the data from the multi-emulsion analysis of the aerial imagery were then compared and related to the soil chemistry. The spatial pattern of the soils in the test field was quantified by cluster analysis using the brightness values obtained from the multi-emulsion analysis of the aerial photo. Percent organic carbon, water content, and color value were the most sensitive variables to be related to the remote sensing data. Using these techniques in conjunction with conventional soil analysis facilitated the determination of fertilizer requirements in the field. Given the field soil pattern, optimum fertility management can be achieved by using differential rates of fertilizers for two of the three soils identified via remote sensing. The results suggest that for the field used in this study the remote sensing techniques can facilitate the application of variable rates of fertilizers.