Severe Storms Nowcasting over the New York metropolitan area

Nowcasting severe storms and catastrophic events such as extreme rainfall has a broad range of applications including: flood prediction, severe storms warning, and navigation purposes. This study evaluates the predictability of convective storms and associated extreme phenomena within storms lifecycle over the New York metropolitan area. As the most populous area in the U.S, with low-level lands, and significant air traffic, the New York metropolitan area has been continuously suffering from severe storms. The current research proposes a framework to provide short-term precipitation forecasting using extrapolation-based models such as a newly developed Pixel-Based Nowcasting algorithm (PBN), PERCAST (PERsian-foreCAST), NOAA's Warning Decision Support System-Integrated Information (WDSS-II) which all work with remote sensing observations, along with high resolution Numerical Weather Prediction (NWP) models. The ground-based radar observations have been used to verify the skill of nowcasting models. Different verifications metrics including probability of detection and false alarm ratio have revealed pros and cons of models. It is shown that extrapolation-based models have robust capabilities in providing very short-term forecasting. Even though forecasting skill is affected during storms initiation, overall, they demonstrate promising performance to provide short-term prediction. Considering storm growth and decay factor, it can improve the predictability of convective storms up to 20% in regards to verification metrics.