Evaluating the Efficiency of Geostatistical Methods for Mapping Spatial Distributions of Springs TDS and PH (Case Study: Mirdeh Watershed, Kurdistan)



The aim of this study was to determine the best interpolation method to achieve spatial and quality distribution of springs from point view (TDS) and PH in Mirdeh watershed NW Kurdistan province. Field sampling collections was done and 268 springs of total 904 springs of whole watershed selected randomly. Electrical conductivity and PH of springs was determined and 168 springs was used as training springs and remain 100 springs was used to testing model and validation purpose. The comparative analysis between geostatistical methods such as: Ordinary Kriging, Simple Kriging, Universal Kriging, IDW, RBF, LPI and GPI were done. The result of accuracy assessment on test springs showed that Global Polynomial with order 3 have the lowest RMSE (66.69) and also to prediction PH, Simple Kriging interpolating of test spring were obtained lowest RMSE and therefore prediction of PH and TDS of other springs were obtained in the study area. Investigating of results revealed that springs have good conformity with geomorphological and topographical variations of study area and in the Northeast and southeast of study area that agricultural and human activity is more than other area, salinization and Total Dissolved Solid (TDS) is more increased.