S426 : The preparation of soil fertility map of Tavoni Tangali lands with geographic information system (GIS) and artificial neural network (ANN)
Thesis > Central Library of Shahrood University > Agricultural Engineering > MSc > 2016
Authors:
Abstarct: One of the main goals of Soil Science, proper use of lands in a way that benefits the most is the preservation of natural resources. to achieve this goal, understanding the spatial variability of soil characteristics for optimum use of resources and proper management of soil fertility is very important. The aim of this study is to map the spatial distribution of the variables affecting the fertility of the soil on the surface of 100 hectares of Tangali Tavoni Incheboron area in Golestan province was to use olive groves. For this purpose 20 profiles were prepared and soil collected from 0-30, 30-60 and 60-90 cm depth and GPS used for the exact position of profiles. Soil physical and chemical properties such as Soil pH, EC, texture, organic carbon, available potassium, available Magnesium, available zinc and total sulphur estimated. Soil physico-chemical results were analysis with ArcGIS software. Different methods were applied for median including distance reverse weight, spline, normal kriging with circle variogram, expontional, gaussian and spherical variants in soil depth of 0-30 cm. With the help of these software, the best technique were selected to prepare the soil fertility distribution. Also with Qnet software for evaluation of dependence transfer of sigmoid, tangent, hyperbolic and secant hyperbolic, the best for model of ANN were selected and then with the use of statically parameters MAE, RMSE and R2 for competence comparison for GIS and ANN. Statistical analysis of GIS and ANN methods showed that ANN model has a best performance than the GIS to estimate the percentage of organic carbon with hyperbolic tangent function, available potassium with Gaussian function, sigmoid function for magnesium, available zinc and total sulfur with Hyperbolic Secant function in 0-30 cm soil depth. also the highest coefficient of determination and the lowest error is related to artificial neural network model.
Keywords:
#Site distribution #Olive #GIS #ANN #Soil fertility map
Keeping place: Central Library of Shahrood University
Visitor:
Keeping place: Central Library of Shahrood University
Visitor: