In the past few decades soil mapping techniques have seen a remarkable changes in the methodology Different approaches are being used to delineate soil units to map the soils to know their extent and kind at different scales is in practice in India and other countries to meet user needs. Yet, soil units delineation based on soil properties and classification remains incomplete without extrapolating the soil information from one point to the neighborhood polygon. To overcome this limitation remote sensing data has been brought into use to derive some useful information from image processing tools. However, remote sensing data derived products are helpful in estimating a single soil property instead of delineating the soil units on the basis of soil taxonomy. Digital soil mapping procedures are also being used to map the soils at large scale but at regional scale mapping it holds severe limitations. So, an integrated approach was applied to generate a soil map by combining geostatistical methods, remote sensing and soil survey data. In the present study all the typifying pedons to represent different soil series identified in the study area by conducting the soil survey at 1: 50000 were analyzed for their physico-chemical properties and classified as per USDA soil taxonomy. Statistical analysis was performed for all the soil properties of each soil series to find out the standard deviation and each soil series selected for mapping were given a Z value. Data base was generated in Arc GIS 9.1 and all the point observations representing each soil series were interpolated using the Z value. Inverse Distance Weightage, Spline and Krigging interpolation methods were employed to generate the raster maps. The maps were compared to find the most suitable interpolation method. The final map was reclassified to generate a vector map. The vector soil map was integrated with the physiography map generated from the remote sensing image analysis. The soil map generated by the integrated approach provided soil units consisting of single soil series. It was compared with the soil map prepared by the conventional method using 3 tier approach, where each mapping unit contained two soil series in association and the spatial extent of individual series with in the unit was not known. The new soil map provided a better spatial resolution of different soils, and their delineation over the conventional soil map, which has a serious limitation to link the spatial and point data observations to predict and simulate the soil properties and pro,cess over the period of time.
