Development of new methods for detailed soil mapping on sloping lands using geogrhaphic information systems and remote sensing tecniques

In this study, it is attempted to develop new methods for the preparation of detailed soil maps for sloping lands in order to save time and lower costs. To achieve this goal 38 000 hectares of land was selected at the north of Adana, Turkey located between Seyhan and Catalan Dams. New methods were tested in order to determine the probable land soil mapping units and their boundaries prior to field work. Random test areas were selected in different sizes from the study area. Slope maps and Digital Elevation Models (DEM) were calculated from 1:25000 scale topographic maps to be used as base maps also; geological maps and Quick Bird digital satellite imagery were used during the course of the project.The first stage of the study was to develop software to collect morphological soil field data on the identified soil series. In order to establish the most accurate, time saving and cost effective method for soil surveys Remote Sensing Techniques, GIS methods, Artificial Neural Networks and Soil Landscape Inference Model (SoLIM) were applied and tested to selected test areas within the study area. However, objectives on identifying and predicting soil mapping units and their boundaries prior to field work were not achieved using digital soil mapping techniques. The main reason for not achieving the objectives were; the lack of accurate data sources and heterogenic soil distribution in the study area. These goals could be achieved if there is accurate geological and topographic information. It is concluded that; digital soil mapping techniques could be used effectively and accurately for the prediction of soil mapping units and their boundaries if topographic and geological maps scales are larger than the soil map scales.