RN-Net: A Deep Learning Approach to 0–2 Hour Rainfall Nowcasting Based on Radar and Automatic Weather Station Data

Abstract

Precipitation has an important impact on people’s daily life and disaster prevention and mitigation. However, it is difficult to provide more accurate results for rainfall nowcasting due to spin-up problems in numerical weather prediction models. Furthermore, existing rainfall nowcasting methods based on machine learning and deep learning cannot provide large-area rainfall nowcasting with high spatiotemporal resolution. This paper proposes a dual-input dual-encoder recurrent neural network, namely Rainfall Nowcasting Network (RN-Net), to solve this problem. It takes the past grid rainfall data interpolated by automatic weather stations and doppler radar mosaic data as input data, and then forecasts the grid rainfall data for the next 2 h. We conduct experiments on the Southeastern China dataset. With a threshold of 0.25 mm, the RN-Net’s rainfall nowcasting threat scores have reached 0.523, 0.503, and 0.435 within 0.5 h, 1 h, and 2 h. Compared with the Weather Research and Forecasting model rainfall nowcasting, the threat scores have been increased by nearly four times, three times, and three times, respectively.

A Novel Multi-Input Multi-Output Recurrent Neural Network Based on Multimodal Fusion and Spatiotemporal Prediction for 0–4 Hour Precipitation Nowcasting

“…The dataset used in the experiment includes 20-month precipitation amount grid data, radar echo data, and reanalysis data, and two months of precipitation amount forecast data of the WRF model for comparison. The experimental results show that MFSP-Net is better than RN-Net [9] for precipitation amount nowcasting. Compared with other precipitation intensity nowcasting models, MFSP-Net has apparent advantages in heavy precipitation nowcasting.…”

“…In order to improve accuracy, deep learning has become an important development direction of precipitation nowcasting. Many scholars have conducted research [7][8][9][10][11][12][13][14][15][16][17] in this direction. Precipitation nowcasting based on deep learning includes precipitation intensity nowcasting and precipitation amount nowcasting, which predict the instantaneous value and the cumulative value of precipitation.…”

“…Zhang et al [28] used a multi-layer perceptron to forecast the precipitation amount data of 56 weather stations in China for the next 3 h. The forecast is derived from 13 physical factors related to precipitation in the surrounding area. RN-Net [9] regards precipitation amount nowcasting as a spatiotemporal prediction problem. It takes past precipitation amount grid data and radar echo data as input data, and then forecasts the precipitation amount grid data for the next 2 h.…”

“…Among the few weather nowcasting methods that use multi-source data, the multimodal fusion method is relatively simple. LightNet [30] uses WRF simulation data and observation data as the basis for 0-6 h lightning nowcasting, and RN-Net [9] uses radar echo data and observation data as the basis for 0-2 h precipitation amount nowcasting. Both use the late fusion method to fuse a variety of data.…”

Two-Stage Spatiotemporal Context Refinement Network for Precipitation Nowcasting

“…A Unet-based model on the fusion of rainfall radar images and wind velocity produced by a weather forecast model is proposed in [7] and improves the prediction for high precipitation rainfalls. Moreover, the dual-input dual-encoder network structures are also proposed to extract simulation-based and observation-based features for prediction [38,39]. The limitation of the existing deep learning models lies in the defect of the extracting ability of spatiotemporal characteristics.…”