The Use of Remote Sensing and Geographic Information System Techniques to Determine Relationships between Land Use and Landform

Tülay TUNÇAY, İlhami BAYRAMİN, Hasan Sabri ÖZTÜRK, Mümtaz KİBAR, Oğuz BAŞKAN


This paper reports on an exercise in obtaining landform information from a Digital Elevation Model (DEM) using Geographic Information Systems (GIS), and relating land cover information from satellite data to the landform classes and soil formation. The Beypazari area (northern Turkey) was chosen because of its varied landforms, land use and land cover. Soil samples, collected from forty-four soil pits were analysed and classified. Most of the soils were classified as Entisol and Inceptisol. In general, the effect of parent material on soil formation was dominated by landform. Landsat TM data were classified to determine land cover categories, and DEM data were analysed to determine landform classes. A strong correlation was found between landform and land cover. The expected results were observed between the soil formation and landform. Inceptisols were generally distributed over the hills and mountains where higher elevation and steep slopes are. As anticipated, forest, rangelands and barren lands were strongly associated with areas of high elevation and steep slopes. However, contrary to expectations, much of the agricultural land was also associated with hill and mountain areas.
The Landform classification with DEM analyses was very successful except for the narrow valleys located in hilly areas. To separate or identify narrow colluvial valleys in these hilly areas, the different resolution and window size for neighbouring have to be tested for the landform classification. The Landsat TM data were very useful for the 1st level land cover study. On the other hand, higher spatial resolution (< 30 m) and multi-temporal data were needed especially in narrow valleys where irrigated areas and trees were not separated successfully.

Tam Metin: PDF