Effective Training Strategies for Deep-learning-based Precipitation Nowcasting and Estimation

TL;DR

This paper introduces a pre-training scheme and a new loss function for deep learning-based precipitation nowcasting and estimation, significantly improving prediction accuracy and handling class imbalance.

Contribution

It proposes a novel pre-training method and loss function for U-Net models applied to radar image-based precipitation tasks, enhancing performance in nowcasting and estimation.

Findings

CSI for heavy rainfall nowcasting improved by up to 95.7% at 5 hours lead time

Pre-training reduces precipitation estimation error by up to 10.7% for light rainfall

Approach is sensitive to different spatial resolutions and analyzed through case studies

Abstract

Deep learning has been successfully applied to precipitation nowcasting. In this work, we propose a pre-training scheme and a new loss function for improving deep-learning-based nowcasting. First, we adapt U-Net, a widely-used deep-learning model, for the two problems of interest here: precipitation nowcasting and precipitation estimation from radar images. We formulate the former as a classification problem with three precipitation intervals and the latter as a regression problem. For these tasks, we propose to pre-train the model to predict radar images in the near future without requiring ground-truth precipitation, and we also propose the use of a new loss function for fine-tuning to mitigate the class imbalance problem. We demonstrate the effectiveness of our approach using radar images and precipitation datasets collected from South Korea over seven years. It is highlighted that our pre-training scheme and new loss function improve the critical success index (CSI) of nowcasting of heavy rainfall (at least 10 mm/hr) by up to 95.7% and 43.6%, respectively, at a 5-hr lead time. We also demonstrate that our approach reduces the precipitation estimation error by up to 10.7%, compared to the conventional approach, for light rainfall (between 1 and 10 mm/hr). Lastly, we report the sensitivity of our approach to different resolutions and a detailed analysis of four cases of heavy rainfall.