Ground validation of high-resolution WRF model precipitation estimates over Colombia

In this study, we perform a thorough validation of precipitation estimates from the high-resolution WRF model, run with a 4-km horizontal grid spacing over Colombia during 2000-2022, using in-situ data from the Colombia weather, climate and hydrology service (IDEAM) at annual, monthly, and diurnal scales. Model outputs were validated against IDEAM rain gauge data using multiple statistical metrics, including Spearman correlation, p-value, RMSE, ME, MAE, and BIAS. Results show that the model is able to capture the main precipitation regimes, with notable contrasts between coastal (Caribbean and Pacific) and low-lying and plain regions (Orinoco and Amazon), and over the Andes cordillera. While the model generally tends to overestimate rainfall throughout most of the country, the error metrics are smaller over the Andean regions, where the spatial and seasonal variability are better represented. Comparisons across regions at monthly, interannual, and diurnal scales highlight significant differences between model estimates over the Pacific region and those over the Andes. The analysis includes the incidence of both phases of ENSO (El Niño and La Niña), showing positive and negative precipitation anomalies ranging between -300 mm and 350 mm per month, with higher anomalies during El Niño. Results of the validation at monthly and diurnal timescales highlight characteristic nighttime precipitation peaks consistent with the literature. These results confirm that, although the model effectively reproduces high-rainfall regions and their seasonal and diurnal variability, systematic biases remain, especially in the wettest periods (MAM and SON) underscoring the need for further calibration to improve its accuracy and practical applicability.