Abstract:Accurate nowcasting (short-term prediction, 0–6 h) of high-impact weather, such as landfalling hurricanes and extreme convective precipitation, plays a critical role in natural disaster monitoring and mitigation. A number of nowcasting approaches have been developed in the past few decades, such as optical flow and the tracking radar echoes by correlation system. Most of these mainstream operational techniques are based on radar echo map extrapolation, which determines the velocity and direction of precipitation systems using historical and current radar observations. However, the skill of the traditional extrapolation method decreases rapidly within the first hour. In order to improve nowcasting skill, recent studies have proposed using deep learning methods, such as convolutional recurrent neural network and trajectory gate recurrent unit. But none of these methods focuses on high-impact weather events, and the deep learning models trained based on general precipitation events cannot meet the demand of accurate warnings and decision-making at the scales required for high-impact weather events, such as hurricanes. Using multiradar observations, this article introduces the idea of self-attention and develops a self-attention-based gate recurrent unit (SaGRU) to enhance its generalization capability and scalability in predicting high-impact weather events. In particular, two types of high-impact weather systems, namely, landfalling hurricanes and extreme convective precipitation events, are investigated. Three models are trained based on hurricane events, heavy rainfall (i.e., nonhurricane) events, and all events combined in the southeast United States during 2015 and 2020. The impacts of different data sources on the nowcasting performance are quantified. The evaluation results of nowcasting products show that our SaGRU performs very well in predicting hurricane-induced rainfall. In the new methodology, the data from nonhurricane events are shown- to provide useful information in enhancing the nowcasting performance during hurricane events as the model trained by combining all the hurricane and nonhurricane events has the best performance. In addition, this article quantifies the impact of the sequence length of input radar observations on the nowcasting performance, which shows that five consecutive observations are sufficient to obtain a stable model, and even two consecutive observations can produce reasonable results.

A Deep Learning-Based Wind Field Nowcasting Method with Extra Attention on Highly Variable Events

A Spatio-Temporal Neural Network for Fine-Scale Wind Field Nowcasting Based on Lidar Observation

DeepWind: A heterogeneous spatio-temporal model for wind forecasting

A Self-Attention Based Method for Wind Speed Forecasting in High-Speed Railway System

A Spatiotemporal Deep Neural Network for Fine-Grained Multi-Horizon Wind Prediction

A Short-Term Wind Speed Prediction Method Utilizing Novel Hybrid Deep Learning Algorithms to Correct Numerical Weather Forecasting

Extreme and severe convective weather disasters: A dual-polarization radar nowcasting method based on physical constraints and a deep neural network model

A Deep Learning Method for Real-time Bias Correction of Wind Field Forecasts in the Western North Pacific

DeepNWP4Wind: A Wind Forcasting Correction Solution in View of Multi-Criteria Evaluation.

Temporally Correlated Deep Learning-Based Horizontal Wind-Speed Prediction

Spatiotemporal wind speed forecasting using conditional local convolution and multidimensional meteorology features

A Novel Hybrid Forecasting System Based on Data Augmentation and Deep Learning Neural Network for Short-Term Wind Speed Forecasting

An Improved Deep Learning Model for High-Impact Weather Nowcasting

Nowcasting extreme rain and extreme wind speed with machine learning techniques applied to different input datasets

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

Series-wise Attention Network for Wind Power Forecasting Considering Temporal Lag of Numerical Weather Prediction

A Deep Learning‐Based Methodology for Precipitation Nowcasting with Radar

Hybrid wind speed forecasting model based on multivariate data secondary decomposition approach and deep learning algorithm with attention mechanism

Windformer: A Novel 4D High-Resolution System for Multi-Step Wind Speed Vector Forecasting Based on Temporal Shifted Window Multi-Head Self-Attention

Multi-head Attention-Based Probabilistic CNN-BiLSTM for Day-Ahead Wind Speed Forecasting