Gholamreza Zehtabian; Hassan Ahmadi; Aliakbar Nazari Samani; amir houshang ehsani; Mahdi Tazeh
Abstract
Plains are one of the most important geomorphological units and different parameters have been considered for classification of plain areas. One of most common classifications in natural resources studies in Iran entailing different qualitative and quantitative factors is: bare plains, apandazh plain ...
Read More
Plains are one of the most important geomorphological units and different parameters have been considered for classification of plain areas. One of most common classifications in natural resources studies in Iran entailing different qualitative and quantitative factors is: bare plains, apandazh plain and covered plain. Such classifications are used to make plains distinguishable from one another. In this study, the geomorphometrical parameters were considered for plain classification by using artificial neural networks and sensitivity analysis. These parameters were extracted by using mathematical equations and applying the corresponding relations on digital elevation models and they are not widely used in Iran. Geomorphometric parameters that were used 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 were calculated in an area of 125000 hectare and at 1500 points, and the result was compared and calibrated with ground truth map. Sampling method in this study was Latin Hyper cube that is a kind of stratified random sampling. Results of this study show that the most important geomorphometric parameters to classify desert plains include Plan Curvature and Profile Curvature that have the highest sensitivity among different plain types. The more the topography of the area reduced the more the contribution and importance of these factors for separating plain types decreased so that these parameters were most prominent in bare plains but had the lowest efficiency in covered plains.
amir houshang ehsani; Marzieh Foroutan
Abstract
In this study the land surface in western half of hyper-arid Lut desert, in south east of Iran, which is covered by Yardangs, a worldwide typical landform for Aeolian erosion, were classified by Self Organizing Maps (SOM) method. In the first step by using Digital Elevation Model with 10 m resolution ...
Read More
In this study the land surface in western half of hyper-arid Lut desert, in south east of Iran, which is covered by Yardangs, a worldwide typical landform for Aeolian erosion, were classified by Self Organizing Maps (SOM) method. In the first step by using Digital Elevation Model with 10 m resolution and Matlab software, 22 morphometric parameters were calculated based on derivative of the surface elevation with first, second and third orders. In the second step most affective parameters for classification and the optimum number of classes were found through utilizing Optimum Index Factor and Davies Bouldin Index. Finally SOM classification was performed on seven morphometric parameters to result in seven classes. The results showed that most appropriate parameters in classification of area are plan curvature, rotor, hypsometric Integral, total accumulation curvature, slope steepness, extreme curvature and mean curvature. The study area were divided to seven classes including saddle valley, Concave ellipsoid, Gentle slope corridor, shoulder with concave slope, shoulder with convex slope, ridge, corridor channels. Sensitivity analysis results revealed that the most sensitive parameters are rotor, mean curvature and hypsometric Integral. Also the results of Jeffreys-Matusita Distance illustrated that parameter pair hypsometric integral / extreme curvature has the most ability in separation of classes in this area. Comparison of the separated classes with the landforms on aerial photographs confirms our classification results.
Arash Malekian; Mahrou Dehbozorgi; Amir Houshang Ehsani; Amir Reza Keshtkar
Abstract
Consecutive droughts in Sistan and Baloochestan province cause water resources restriction and this isa very significant problem for this region. In this study, in order to forecast the drought cycle in 9climatological stations in the province, we used Artificial Neural Networks. The input data wereaverage ...
Read More
Consecutive droughts in Sistan and Baloochestan province cause water resources restriction and this isa very significant problem for this region. In this study, in order to forecast the drought cycle in 9climatological stations in the province, we used Artificial Neural Networks. The input data wereaverage of annual rainfall data in all stations and also deciles precipitation index, which the first 30years from 1971 to 2000 used for training the network and the last 8 years from 2001 to 2008 forsimulating it. The network consists of Multilayer Perceptron (MLP) and Back Propagation Algorithm(BP) and also sigmoid transfer function. Number of Neurons in hidden layer was 10 with 1-10-1structure and was calculated based on the lowest RMSE. Then drought prediction was done in neuralnetwork with the trained algorithm and without using actual and observed data in 2009 to 2012.Results showed that, the network was able to simulate and forecast DPI index with 97% regressionand average RMSE error less than 5%. According to drought indices, results showed that the droughtwill have an increasing trend in all stations in this region in 2009 to 2011. Therefore, by using thismethod, drought can be predicted in later years without any need to have actual meteorological dataand also can be used in water resources management, drought management and climate changes.
