Accurate and timely precipitation forecasts are critical in modern society, influencing both economic activity and daily life. While deep learning methods leveraging remotely sensed radar data have become prevalent for precipitation nowcasting, longer-term forecasting remains challenging. This is due to accumulated errors in deep learning models and insufficient information about precipitation systems over longer time horizons. To address these challenges, we introduce the Short-Term Precipitation Forecast Network (STPF-Net), a recurrent neural network designed for longer-term precipitation prediction. STPF-Net uses a multi-tier structure with varying temporal resolutions to mitigate the accumulated errors during longer forecasts. Additionally, its transformer-based module incorporates larger spatial contexts, providing more complete information about precipitation systems. We evaluated STPF-Net on radar data from southeastern China, training separate models for 6 and 12 h forecasts. Quantitative results demonstrate STPF-Net achieved superior accuracy and lower errors compared to benchmark deep learning and numerical weather prediction models. Visualized case studies indicate reasonably coherent 6 h predictions from STPF-Net versus other methods. For 12 h forecasts, while STPF-Net outperformed other models, it still struggled with storm initiation over longer forecasting time.

Accurate state estimation of the high‐dimensional, chaotic Earth's atmosphere marks a Sisyphean task, yet is indispensable for initiating weather forecasts and gauging climate variability. While much effort is devoted to assimilating observations and forecasts to infer weather states, the inherent low‐dimensional statistical structure in atmospheric circulation, shaped by geophysical laws and geographic boundaries, is underutilized as an informative prior for state inference, or as reference for assessing representative of existing observations and planning new ones. We realize these potential by learning climatological distribution from climate reanalysis/simulation, using a deep generative model. For a case study of estimating 2 m temperature spatial patterns, the learned distribution faithfully reproduces climatology statistics. A combination of the learned climatological prior with few station observations yields strong posterior of spatial pattern estimates, which are spatially coherent, faithful and adaptive to observational constraints, and uncertainty‐aware. This allows us to evaluate each observation's value in reducing pattern estimation uncertainty, and guide optimal observation network design by pinpointing the most informative sites. Our study showcases how generative models can extract and utilize information produced in the chaotic evolution of climate system.

Deep learning models based on radar echo extrapolation have been widely used in precipitation nowcasting. However, they face the challenge of insufficient input information when extending the forecast lead time, requiring the incorporation of physics‐based numerical weather prediction (NWP). Given that the strengths of radar and NWP data vary depending on the forecast time, effectively fusing these two data sources in a unified deep learning model remains an open research problem. In this study, we propose a Time‐aware Adaptive Feature Fusion Network (TAFFNet) for very short‐term precipitation forecasts up to 12 hr. TAFFNet fuses features adaptively according to their relative contributions to forecast skill at different times. Experimental results demonstrate that TAFFNet performs the best for very short‐term precipitation forecasts. The case studies show that adaptively fusing NWP with radar can improve the accuracy of precipitation forecasts, especially for predicting the initiation and dissipation of storms at longer lead times.

Satellite and radar observations represent two fundamentally different remote sensing observation types, providing independent information for numerical weather prediction (NWP). Because the individual impact on improving forecast has previously been examined, combining these two resources of data potentially enhances the performance of weather forecast. In this study, satellite radiance, radar radial velocity and reflectivity are simultaneously assimilated with the Proper Orthogonal Decomposition (POD)-based ensemble four-dimensional variational (4DVar) assimilation method (referred to as POD-4DEnVar). The impact is evaluated on continuous severe rainfall processes occurred from June to July in 2016 and 2017. Results show that combined assimilation of satellite and radar data with POD-4DEnVar has the potential to improve weather forecast. Averaged over 22 forecasts, RMSEs indicate that though the forecast results are sensitive to different variables, generally the improvement is found in different pressure levels with assimilation. The precipitation skill scores are generally increased when assimilation is carried out. A case study is also examined to figure out the contributions to forecast improvement. Better intensity and distribution of precipitation forecast is found in the accumulated rainfall evolution with POD-4DEnVar assimilation. These improvements are attributed to the local changes in moisture, temperature and wind field. In addition, with radar data assimilation, the initial rainwater and cloud water conditions are changed directly. Both experiments can simulate the strong hydrometeor in the precipitation area, but assimilation spins up faster, strengthening the initial intensity of the heavy rainfall. Generally, the combined assimilation of satellite and radar data results in better rainfall forecast than without data assimilation.

Within Open Relation Extraction (ORE) tasks, the Zero-shot ORE method is to generalize undefined relations from predefined relations, while the Unsupervised ORE method is to extract undefined relations without the need for annotations. However, despite the possibility of overlap between predefined and undefined relations in the training data, a unified framework for both Zero-shot and Unsupervised ORE has yet to be established. To address this gap, we propose U-CORE: A Unified Deep Cluster-wise Contrastive Framework for both Zero-shot and Unsupervised ORE, by leveraging techniques from Contrastive Learning (CL) and Clustering. ¹ U-CORE overcomes the limitations of CL-based Zero-shot ORE methods by employing Cluster-wise CL that preserves both local smoothness as well as global semantics. Additionally, we employ a deep-cluster-based updater that optimizes the cluster center, thus enhancing the accuracy and efficiency of the model. To increase the stability of the model, we adopt Adaptive Self-paced Learning that effectively addresses the data-shifting problems. Experimental results on three well-known datasets demonstrate that U-CORE significantly improves upon existing methods by showing an average improvement of 7.35% ARI on Zero-shot ORE tasks and 15.24% ARI on Unsupervised ORE tasks.