<p>Rainfall distribution in Thailand is strongly influenced by monsoonal circulation and complex terrain, creating persistent challenges for radar-based quantitative precipitation estimation (QPE). This study presents a storm-scale evaluation of GIS-based gauge correction applied to reflectivity-derived radar rainfall during Tropical Storm Son-Tinh (15–23 July 2018) in Central Thailand. C-band radar data from the Phitsanulok station were preprocessed using standard quality-control procedures (clutter and noise filtering, attenuation correction) and converted to rainfall using the Marshall–Palmer Z–R relationship as a baseline estimator. A bias-ratio-based correction framework was applied to reduce systematic magnitude bias, while Mean Field Bias (MFB) was computed as a diagnostic indicator of event-scale bias magnitude. Spatial bias adjustment was performed using eight deterministic and geostatistical interpolation methods, including Inverse Distance Weighting (IDW), Ordinary Kriging, and Universal Kriging. Performance was evaluated using five-fold cross-validation based on RMSE, MAE, correlation coefficient (r), coefficient of determination (R²), and mean bias. Prior to correction, radar rainfall exhibited moderate agreement with gauge observations and substantial systematic underestimation. After correction, radar–gauge consistency improved across all methods, with IDW producing the lowest numerical errors, while Gaussian Ordinary Kriging generated smoother spatial fields and bias values closest to unity, particularly in terrain-affected areas. These improvements primarily reflect statistical alignment with gauge observations rather than resolution of intrinsic radar measurement limitations. The findings demonstrate that GIS-based bias correction can enhance event-scale radar rainfall consistency for hydrological and post-event applications, while its effectiveness remains conditional on storm characteristics, gauge distribution, and inherent constraints of reflectivity-only radar retrieval in tropical environments.</p> Graphical Abstract <p></p> <p>The graphical abstract presents a structured overview of the data integration, processing workflow, methodological development, and evaluation framework used for event-scale radar rainfall bias correction during Tropical Storm Son-Tinh (15–23 July 2018) in central Thailand. The study integrates C-band weather radar data with observations from 89 automatic rain-gauge stations. Radar data were processed using Python and open-source libraries (e.g., Py-ART), including clutter removal, Signal-to-Noise Ratio (SNR) filtering, attenuation correction, and polar-to-Cartesian transformation to generate a 2&#xa0;km CAPPI. Rainfall estimates were derived using the Marshall–Palmer Z–R relationship. A bias-ratio-based correction framework was developed, where station-based bias ratios (Gauge/Radar) were computed and spatially interpolated using both deterministic and geostatistical methods, including Inverse Distance Weighting (IDW), Ordinary Kriging, and Universal Kriging. These interpolated bias fields were applied multiplicatively to adjust radar-derived rainfall. The framework was evaluated using five-fold cross-validation, comparing corrected radar rainfall against independent gauge observations. The graphical abstract emphasizes the integration of radar and gauge data, the use of GIS-based spatial interpolation, and the systematic comparison of interpolation techniques to improve radar–gauge consistency under complex tropical conditions.</p>