Field Scale Mapping of Surface Soil Clay Concentration

The surface soil clay concentration is a useful soil property to map soils, interpret soil properties, and guide irrigation, fertilizer, and agricultural chemical applications. The objective of this study was to determine whether surface soil clay concentrations could be predicted from remotely sensed imagery of bare surface soil or from soil electrical conductivity for a 115 ha field located in Crisp County, Georgia. The soil clay concentrations were determined for soil samples taken at 28 field locations. Three different data sources–an aerial color photograph image, two infrared bands from an ATLAS data set, and the electrical conductivity of the surface soil layer were used in the research. Principal components analysis was applied to the color photograph image, whereas the ratio of two infrared bands was applied to the ATLAS data set. Filtering was applied to both resulting images. The distribution of soil electrical conductivity was derived from the measured soil electrical conductivity data by spatial analysis. Statistical relationships between soil clay concentrations and the principal component 3, the ratio of two ATLAS infrared bands, and the soil electrical conductivity were analyzed, and three linear equations were derived with r2 values 0.83, 0.52, and 0.78, respectively. The distribution of the soil clay concentrations was derived based on these three equations. Six levels of soil clay concentrations were classified in these three methods, and the advantages and disadvantages were discussed. The predicted and measured soil clay concentrations, based on additional soil samples from 30 field locations, were compared using linear regression (r2=0.76, 0.45, and 0.77 for the three methods). The overall accuracy for these methods were 84%, 66%, and 76%, respectively. The principal components method had the highest accuracy in our research, while the result for the depressional areas is the best from the ratio method.