ارزیابی دقت روش‌های میان‌یابی در تخمین سطح ایستابی آب زیرزمینی (مطالعه موردی: آبخوان‌های فارسان ـ جونقان و سفید دشت)

The accuracy and precision of the input data in decision making is important. Error originates from data collection, data entry, storage, retrieval and analysis of the data which consequently result in model error. One of the errors in spatial analysis is interpolation error. The main objective of this research was the suitability assessment of some interpolation methods for estimation of groundwater level in Farsan-Joneghan and Sefiddasht aquifers, located in Beheshtabad catchment, Chaharmahl-Va-Bakhtiyari province, Iran. Cross-validation technique was employed for the determination of each method's error. The RMSE and MAE indices were used for the error comparison. The results show that the modified Shapard's method with an MAE=6 and RMSE=7 was the most accurate for interpolation of groundwtaer level in the Sefiddasht aquifer. The inverse distance power method with MAE=6 and RMSE=9 was the best interpolation method for Farsan-Jonaghan aquifer. The Kriging with MAE=7 and RMSE=12 is the second best method in these aquifers. The moving average, minimum curvature and polynomial regression procedures produce the maximum error in the aquifers (17<MAE<25 and 25<RMSE<37). It was concluded that no particular method can be recommended for the interpolation of groundwater level in either aquifer as any specific recommendation depends upon (1) the distribution of the measurement points relative to aquifer centered, (2) total number of the measurement points and (3) the difference between maximum and minimum topographic levels of the aquifer. Therefore, for any aquifer the suitable interpolation methods should be identified.