Deep Learning-Based Radar Composite Reflectivity Factor Estimations from Fengyun-4A Geostationary Satellite Observations

Abstract

Ground-based weather radar data plays an essential role in monitoring severe convective weather. The detection of such weather systems in time is critical for saving people’s lives and property. However, the limited spatial coverage of radars over the ocean and mountainous regions greatly limits their effective application. In this study, we propose a novel framework of a deep learning-based model to retrieve the radar composite reflectivity factor (RCRF) maps from the Fengyun-4A new-generation geostationary satellite data. The suggested framework consists of three main processes, i.e., satellite and radar data preprocessing, the deep learning-based regression model for retrieving the RCRF maps, as well as the testing and validation of the model. In addition, three typical cases are also analyzed and studied, including a cluster of rapidly developing convective cells, a Northeast China cold vortex, and the Super Typhoon Haishen. Compared with the high-quality precipitation rate products from the integrated Multi-satellite Retrievals for Global Precipitation Measurement, it is found that the retrieved RCRF maps are in good agreement with the precipitation pattern. The statistical results show that retrieved RCRF maps have an R-square of 0.88-0.96, a mean absolute error of 0.3-0.6 dBZ, and a root-mean-square error of 1.2-2.4 dBZ.

Radar Composite Reflectivity Reconstruction Based on FY-4A Using Deep Learning

“…The RMSE, MAE, and R2 of the U_VIS_IR model are 2.5024 dBZ, 0.8457 dBZ, and 0.8954, respectively, which are superior to those of the U_IR model. This is probably because the VIS/NIR bands include additional information on cloud optical depth, including VIS/IR bands, which could better reflect the evolution of cloud and convective systems [ 28 ]. Overall, the AU_VIS_IR model shows the best performance among all models in terms of all statistics, which is consistent with the results mentioned above.…”

“…The input imager spectral bands in the model are selected according to the physical interpretation of radar reflectivity–convection–satellite observation relationships, mainly focusing on bands that are most sensitive to clouds and hydrometeors, and the study of Sun et al [ 28 ] was also referred to. For example, the VIS band at 0.65 μm is a weak absorption band sensitive to COT and cloud phase, especially for strong convective clouds.…”

“…This region was selected because the radar coverage is larger than most other areas in western China, thus providing sufficient radar observations for model training. CREF is the maximum base reflectivity factor that can be found in a given vertical column in the radar umbrella [ 28 ]. CREF can visually reflect the intensity structure and variation of a strong convective system.…”

“…Recently, several studies have been presented using DL techniques to reconstruct radar composite reflectivity (CREF) [ 19 , 27 , 28 ]. This study is a further exploration based on the previous works, a DL technology is developed to reconstruct radar CREF from observations of China’s new-generation geostationary meteorological satellite FY-4A.…”

Toward a Deep-Learning-Network-Based Convective Weather Initiation Algorithm From the Joint Observations of Fengyun-4A Geostationary Satellite and Radar for 0–1h Nowcasting

“…The specific reason for choosing these interest fields is that they are sensitive to the cloud-top microphysical composition [27,54,55]. The study of retrieving the radar composite reflectivity factor (RCRF) from Fengyun-4A /AGRI interest fields has explored the spectral response of cloud top attributes and the sensitivity of these interest fields [56]. And the statistical results show that visible models exhibit a better performance of retrieving RCRF.…”