Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Soil nutrient map can be produced through spatial interpolation using grid sampling data. However, results from the interpolation approach are affected by the interpolation methods (inverse distance weighting or geostatistical methods such as kriging) and may not accurately represent the real field condition in between the sampling locations. Remote sensing image, especially aerial hyperspectral image can provide detailed pixel by pixel spectral information. Geostatistical techniques provide a means for utilizing both spatial information from soil sampling data and spectral information from remote sensing imagery. This paper used ground sample data as the primary information and aerial hyperspectral imagery as secondary information for soil nutrient spatial interpolation to produce nutrient map. The nutrient properties under investigation include soil PH, potassium, phosphorus, and organic matter. Sequential Gaussian cosimulation is used in this paper for soil nutrient mapping and the result is compared with another geostatistical procedure, colocated ordinary cokriging estimation. The results showed that Gaussian cosimulation gave the best correlations for all soil nutrient factors except phosphorus. The results also indicated that both hyperspectral image-based geostatistical methods worked best for soil organic matter content mapping.