Advancing Extreme Precipitation Forecasting for Coastal-Urban Regions: A New York City Case Study

Urban extreme precipitation poses a significant challenge due to its increasing societal impact and the limitations of timely forecasting. Historically, New York State has faced substantial social, infrastructural, and economic consequences from such events, particularly in major urban areas like New York City. This study evaluates the performance of mesoscale models in simulating two extreme precipitation events— (1) August 12, 2020, and (2) August 6, 2024—within New York City. The analysis employs the fully urbanized Weather Research and Forecasting (uWRF) model version 3.9, configured with the Building Effect Parameterization and Building Energy Model (uWRF BEP+BEM) coupled to the Mellor-Yamada-Janjic (MYJ) planetary boundary layer (PBL) scheme, hereafter referred to as uWRF. Subsequently, the model was upgraded to version 4.2, which integrates BEP+BEM with the Mellor–Yamada–Nakanishi–Niino (MYNN-3) PBL scheme, including enhancements such as a variable building drag coefficient and modifications to cooling tower effects.

Results reveal that both extreme precipitation events were preceded by elevated urban temperatures (~31°C), which declined following the rainfall and led to widespread flooding across the city. The upgraded model demonstrated improved spatial prediction of rainfall extremes but underestimated their intensity. The root means square error (RMSE) in precipitation prediction decreased from 7.5 mm to 6.1 mm, highlighting the benefit of model advancements. Future research will focus on investigating the influence of the modified PBL scheme within the BEP+BEM framework to enhance urban precipitation forecasts and further addressing spatial biases in the model using Multi-Radar/Multi-Sensor System (MRMS) data.