Evaluation of Spatial Interpolation Methods for Wind Speed and Direction: A Case Study in Split-Dalmatia County

Abstract. Wind speed and direction are spatial variables that vary over both time and space. These variables are crucial for urban and spatial planning, agriculture and crop management, sports activity planning, aerial navigation, air pollution modeling and fire management. This paper investigates the effectiveness of several interpolation methods for predicting wind speed and direction at unknown locations, using measurements from a network of weather stations. Four well-established methods were considered: Natural Neighbor, Inverse Distance Weighting (IDW), Kriging, and Ordinary Kriging.
Data were collected from 28 weather stations distributed across Split and Dalmatia County. In two experiments, the four unknown stations were chosen to represent: 1) a station spatially surrounded by known measurements, and 2) stations representing typical geographical challenges, such as land, coast, canyon, and island locations. For each experiment, scenario, and interpolation method, we calculated and analyzed the Root Mean Squared Error (RMSE), Mean Absolute Error in the u-direction (MAE u), and Mean Absolute Error in the v-direction (MAE v). The analysis revealed that the highest errors occurred during Bora wind conditions. Among the methods, Ordinary Kriging demonstrated the lowest prediction error.