M. Kargar; Mohammad Reza Javadi; S.A.A. Hashemi
Abstract
Soil erosion and sediment production are among most important problems in developing countries including Iran. In this study it has been endeavored that applicability of four (AOF, MUSLE-S, MUSLT and USLE-M) models is investigated in Srfiddasht Research Site, Semnan province, at event scale to estimate ...
Read More
Soil erosion and sediment production are among most important problems in developing countries including Iran. In this study it has been endeavored that applicability of four (AOF, MUSLE-S, MUSLT and USLE-M) models is investigated in Srfiddasht Research Site, Semnan province, at event scale to estimate the sediment. For this, all required variables and inputs of the model have been calculated in the watershed and the estimations from considering statistical models with measured sediments of 15 cloudbursts have been compared. The results for t-student correlation test showed that there is no significant difference (at 1%) between MUSLT, MUSLE-S models and measured sediment. Based on these, it can be said that in this study, the results from these two models have higher accuracies to estimate the sediment from cloudbursts than other methods. Also, the results of evaluation and efficiency of the model using Nash-Suttcliffe criterion and root relative mean squared error (RRMSE) statistic showed that MUSLE-S and MUSLT models have higher efficiencies than other models and inefficiencies of USLE-M and AOF models to estimate sediments from cloudburst have been confirmed in the studied research station in this study.
H. Kashi; H. Ghorbani; S. Emamgholizadeh; S.A.A. Hashemi
Abstract
With respect to the problem of direct measurement of soil parameters in recent year using indirect method such as artificial neural networks has been considered. In the present study, 200 soil samples were collected from Ghoshe location in Semnan province. Half of samples were collected from disturbed ...
Read More
With respect to the problem of direct measurement of soil parameters in recent year using indirect method such as artificial neural networks has been considered. In the present study, 200 soil samples were collected from Ghoshe location in Semnan province. Half of samples were collected from disturbed agricultural lands and the other half were collected from undisturbed nearby lands. Some soil chemical as well as physical properties such as electrical conductivity (EC), soil texture, lime percentage, sodium adsorption ration (SAR) and bulk density were considered as easy and fast obtainable features and soil cation exchange capacity as difficult and time consuming feature. The collected data randomly divided in two categories of training (70%) and testing (30%) and they used for train and test of two artificial neural networks, multi-layer perception using back-propagation algorithm (MLP/BP) and Radial basis functions (RBF) and nonlinear regression model. Results of this research show high efficiency of artificial neural network compared with nonlinear regression and also MLP network was better than RBF network. Sensitivity analysis was also performed for all parameters to find out the relationship between soil mentioned parameters and soil cation exchange capacity for both disturbed and undisturbed soils. At last, the correlation between soil parameters and soil cation exchange capacity was determined and most important parameters which could influence the soil cation exchange capacity were described.
