OPTIMIZED PARAMETERIZATION FOR SPATIAL INTERPOLATION OF DAILY TEMPERATURE DATA AT BLACK SEA REGION, TURKEY

Spatial interpolation methods are frequently used to estimate values of meteorological data in locations where they are not measured. The correct determination of the spatial distribution of meteorological variables is as important as their measurement. Although there are several methods to perform this, it can be difficult to determine which one best reproduces actual condition. Many factors may affect the performance of the spatial interpolation methods. In literature, few research work have been performed to compare the effectiveness of different spatial interpolations in meteorological variables. Especially interpolating regionalized variables such as temperature which is strongly correlated with elevation is worth to investigate. This paper focuses on the performance of three commonly applied spatial interpolation methods, which are Inverse Distance Weighted (IDW), Temperature Lapse Rate method (TLR) combined version of IDW with TLR for daily mean temperature data in the Black Sea Region of Turkey w.r.t. spatial and temporal properties of data points. Up to now, very little research has been investigated for the performance of different interpolation methods in meteorological data of this region. Furthermore, this study targets to determine the optimum parameterization of three commonly used methods based on RMSE (root mean square error), MAE (mean absolute error), MAXE (maximum of error) and MINE (minimum of error) statistics of interpolation errors for daily mean temperature from 1981 to 2012. In order to determine optimized parameterization options of both IDW and combined version of IDW, the crossvalidation method was applied over long periods of 32 years on 52 different sites. The algorithms of three interpolation methods were programmed in MATLAB programming language by authors of paper.