The Use of geoinformatics for estimating soil organic

Abstract

LANDSAT TM 5; band 1-3-4 was used to predict soil organic matter contents in central plain of Thailand: Nakhon Pathom province as a representative area. There were 135 plots from several agricultural land use; paddy fields, sugarcane and fruit orchard. Y=3.926+0.0176 X1-0.0117 X3+0.0476 X4 (R2=0.357) was the result from the interpolation method at RMSE=0.95. The soil organic matter map was build up via Geoinformatics techniques with additional data sources. These techniques include organic matter models and qualitative methods. Finally, validation methods used to assess the accuracy of maps produced with image data are discussed. It is concluded that a general lack of validation data is a main concern. Validation is of utmost importance to achieve regional operational monitoring systems, and close collaboration between the image data and field-based soil scientists is therefore required.