HR-Extreme: A High-Resolution Dataset for Extreme Weather Forecasting
This addresses the need for better extreme weather forecasting datasets and models, which is crucial for practical applications in meteorology and disaster management, though it is incremental as it builds on existing deep learning and NWP systems.
The study tackled the problem of limited datasets for extreme weather forecasting by introducing HR-Extreme, a high-resolution dataset derived from NOAA's HRRR data, and found that errors in extreme weather cases are significantly larger than overall forecast errors, with their improved baseline model HR-Heim showing superior performance.
The application of large deep learning models in weather forecasting has led to significant advancements in the field, including higher-resolution forecasting and extended prediction periods exemplified by models such as Pangu and Fuxi. Despite these successes, previous research has largely been characterized by the neglect of extreme weather events, and the availability of datasets specifically curated for such events remains limited. Given the critical importance of accurately forecasting extreme weather, this study introduces a comprehensive dataset that incorporates high-resolution extreme weather cases derived from the High-Resolution Rapid Refresh (HRRR) data, a 3-km real-time dataset provided by NOAA. We also evaluate the current state-of-the-art deep learning models and Numerical Weather Prediction (NWP) systems on HR-Extreme, and provide a improved baseline deep learning model called HR-Heim which has superior performance on both general loss and HR-Extreme compared to others. Our results reveal that the errors of extreme weather cases are significantly larger than overall forecast error, highlighting them as an crucial source of loss in weather prediction. These findings underscore the necessity for future research to focus on improving the accuracy of extreme weather forecasts to enhance their practical utility.