Spatial Rainfall Interpolation and Evaluation for Seasonal Precipitation in Peninsular Malaysia

Abstract

Rainfall, a major driving force in hydrology and water resources planning, poses challenges due to potential malfunctions in conventional rain gauges, necessitating effective methods for filling the data gap. This study focuses on assessing the efficacy of two spatial rainfall interpolation techniques, i.e., Local Polynomial Interpolation (LPI) and Inverse Distance Weighting (IDW) for seasonal rainfall estimation in Peninsular Malaysia. Interpolated values from both methods are compared to ground observations, and their performance is evaluated through cross-validation using statistical measures such as RMSE, MAE, and R². Additionally, Geographically Weighted Regression (GWR) is employed to analyze the relationship between interpolated rainfall and ground elevation. The findings reveal that LPI outperforms IDW, demonstrating higher R² values and lower MAE and RMSE, highlighting its superior accuracy in rainfall estimation and underscoring the importance of method selection in handling missing rainfall data.