Determining the most important geomorphometric parameters in classification of desert plains with using artificial neural networks and sensitivity analysis

Plains are one of the most important Geomorphological units, and different parameters for classification are presented in plain areas. One of The classification methods that are used in studies of natural resources in Iran, The plain division are in three types: bare plains, Apandazh and covered plain .The qualitative and quantitative factors are used to distinguish them from each other. In this study the geomorphometrical parameters used for plain classification by using artificial neural networks and sensitivity analysis. These parameters extracted by using mathematical equations and the corresponding relations on digital elevation models. Geomorphometric parameters that using in this study included Percent of slope, Plan Curvature, Profile Curvature, Minimum Curvature, the Maximum Curvature, Cross sectional Curvature, Longitudinal Curvature and Gaussian Curvature. These parameters calculated in area 125000 hectare and sample point are 1500; the result with ground truth map was compared and calibrated. Sampling method in this study is Latin Hyper cube that is kind of stratified random sampling. Result of this study shows the importance Geomorphometric parameters to classify desert plains include: Plan Curvature, Profile Curvature that is more sensitivity in plains types. As the topography is reduced, role and importance of these factors are reduced And highest parameters of the parameters is bare plain and lowest performance are covered plains.