Intelligent Reconstruction of Radar Composite Reflectivity Based on Satellite Observations and Deep Learning

Jianyu Zhao,Jinkai Tan,Sheng Chen,Qiqiao Huang,Liang Gao,Yanping Li,Chunxia Wei

DOI: https://doi.org/10.3390/rs16020275

IF: 5

2024-01-11

Remote Sensing

Abstract:Weather radar is a useful tool for monitoring and forecasting severe weather but has limited coverage due to beam blockage from mountainous terrain or other factors. To overcome this issue, an intelligent technology called "Echo Reconstruction UNet (ER-UNet)" is proposed in this study. It reconstructs radar composite reflectivity (CREF) using observations from Fengyun-4A geostationary satellites with broad coverage. In general, ER-UNet outperforms UNet in terms of root mean square error (RMSE), mean absolute error (MAE), structural similarity index (SSIM), probability of detection (POD), false alarm rate (FAR), critical success index (CSI), and Heidke skill score (HSS). Additionally, ER-UNet provides the better reconstruction of CREF compared to the UNet model in terms of the intensity, location, and details of radar echoes (particularly, strong echoes). ER-UNet can effectively reconstruct strong echoes and provide crucial decision-making information for early warning of severe weather.

environmental sciences,imaging science & photographic technology,remote sensing,geosciences, multidisciplinary

What problem does this paper attempt to address?

The paper addresses the challenge of limited radar coverage, especially in complex terrains, which hinders accurate monitoring of severe weather conditions. To overcome this limitation, the authors propose an intelligent technology called "Echo Reconstruction UNet" (ER-UNet). This technology leverages observations from Fengyun-4A geostationary satellites, which have broader coverage, to reconstruct radar composite reflectivity (CREF).

### Problem Addressed:

1. **Limited Radar Coverage:** Weather radars, despite being valuable tools for monitoring severe weather, have limited coverage due to obstacles like mountains. This makes it challenging to monitor severe weather over large regions like Asia.

2. **Need for Accurate Monitoring:** Severe weather events, such as strong convection, rainstorms, and typhoons, cause significant economic losses and pose threats to human life. Accurate monitoring is crucial for disaster prevention and mitigation.

3. **Challenges with Traditional Approaches:** Deploying and operating weather radars in complex terrains is difficult. Geostationary satellites provide global coverage but lack the ability to observe the three-dimensional structure of convective systems. Traditional physical retrieval algorithms have limitations in converting satellite data into radar reflectivity.

### Proposed Solution:

1. **ER-UNet Model:** The proposed ER-UNet model is a deep learning-based approach that uses geostationary satellite observations to reconstruct radar composite reflectivity (CREF).