Evaluation of Fuzzy Clustering and Artificial Neural Network Methods in Spatial Zoning of Annual Precipitation in Iran

Precipitation is one of the main elements of the Earth's hydro-climatic cycle and its variability depends on the complex and non-linear relationships between the climate system and environmental factors. Understanding these relationships and doing environmental planning based on them is difficult. Therefore, classifying data and dividing information into homogeneous and small categories can be helpful in this regard. In the present study, an attempt was made to prepare precipitation, altitude, slope, slope direction, and station density data for 3423 synoptic, climatological, and gauge stations in Iran in the 1961-2015 years’ period. These data were entered into fuzzy (FCM), self-organizing map neural network (SOM-ANN) models and precipitation-spatial zoning. The outputs of the two models were compared in terms of accuracy and efficiency. The results obtained from the output of the models have divided the rainfall conditions of Iran into four zones concerning environmental factors. Evaluations also showed that both models had high accuracy in classifying precipitation parameters; However, the fuzzy model has a relative advantage over the neural network model in the accuracy of results.