Assimilating the Subic radar data in the WRF model for tropical cyclone-enhanced heavy monsoon rainfall prediction in Metro Manila

Metro Manila's rapid urbanization in recent decades has made it vulnerable to flooding during its rainy season. Disaster mitigation of such hazards requires the synergistic use of advanced weather observation and forecasting systems. This study explores the impact of radar data assimilation (DA) on the quantitative precipitation forecast skills of the Weather and Research Forecasting (WRF) model during three selected tropical cyclone-enhanced heavy monsoon rainfall events. The radar reflectivity data from the S-band Subic radar station was assimilated into the WRF model to assess the efficacy of using the three-dimensional variational (3DVAR) DA system. Results indicate that while the DA of radar reflectivity resulted in changes in the model’s hydrometeor budgeting where the presence of higher reflectivity observations increased the rainwater mixing ratio in the assimilated model, and slight improvements were observed for rainfall < 30 mm (i.e. lower bias, higher accuracy score) on 1 to 2-day lead time forecasts, overall, those changes were not significant enough to statistically improve the WRF model’s rainfall forecast skills compared to the non-DA model runs. In the conduct of this study, we found several limitations in the radar data. Hence, recommendations in the radar data characteristics and best practices for a successful DA implementation are provided, highlighting the challenges of DA forecasting for areas in the Philippines prone to weather-related hazards.