Estimation of Groundwater Level Fluctuations Using Neuro-Fuzzy and Support Vector Regression Models

Estimation of Ground Water Level (GWL) is important in the determination of the sustainable use of water resources and Ground Water resources. Groundwater level fluctuations were investigated using the variable of groundwater level, precipitation, temperature. In the present study, GWL estimation studies were conducted via Neuro-Fuzzy (NF), Support Vector Regression with radial basis functions (SVR-RBF) and Support Vector Regression with poly kernel (SVR-PK) models. The daily data of the precipitation, temperature and groundwater level are used which is taken from Minnesota, United States of America. The results were compared with NF and SVR methods. According to this comparison, it was observed that the NF and SVR models gave similar results for observation.