Wind speed forecasting is significant in practical applications such as energy dispatch and meteorological early warning systems. However, spatio-temporal correlations of wind speed are dynamically influenced by weather conditions, seasonal variations, and diurnal fluctuations, resulting in constantly changing spatio-temporal patterns. Meanwhile, wind speed data inherently include short-term high-frequency fluctuations and long-term low-frequency trends, which traditional methods struggle to adapt to and accurately capture multiscale features. This paper proposes a dynamic spatio-temporal graph convolutional network model (DSTGFP) for multi-station wind speed prediction to address this challenge. First, the model constructs an adaptive dynamic adjacency matrix (ADAM) by integrating geographical locations among stations, dynamic time warping (DTW), and mutual information (MI), facilitating dynamic graph modeling. Next, we introduce a novel spatio-temporal feature extraction framework, which employs residual graph convolutional networks combined with a multi-head attention mechanism to extract spatial features. We simultaneously integrate temporal-domain convolution and frequency-domain convolution to capture multiscale temporal-frequency features. Finally, the particle swarm optimization (PSO) algorithm is used for hyperparameter optimization to improve the prediction accuracy. Experimental results demonstrate that the DSTGFP model achieves reductions of 24.66% in mean absolute error (MAE) and 25.47% in root mean square error (RMSE) compared to existing deep learning methods, highlighting its superior predictive performance.

Accurate correction of gridded numerical weather prediction precipitation forecasts remains challenging because many end-to-end deep learning correction models treat meteorological variables as undifferentiated data channels and therefore provide limited physical interpretability. Inspired by general principles of biological environmental sensing, selective information processing, and regulatory constraint learning, this study proposes PCPNet, a bio-inspired and physically constrained precipitation correction framework. The framework does not imitate a specific biological organ or species; instead, it abstracts three information-processing principles into a meteorological correction task. First, key weather-system cues, including low-level shear lines, trough-ridge effects, upper-level jet-stream forcing, vorticity-divergence-related vertical motion, and water-vapor flux convergence, are quantified as structured diagnostic fields. This transforms the subjective synoptic diagnosis of forecasters into automated grid-based sensing features. Second, these diagnostic cues are fused with numerical weather prediction variables and terrain descriptors in an encoder–attention–decoder network, allowing the model to emphasize dynamically important precipitation-triggering regions. Third, water-vapor conservation and terrain-forcing relationships are embedded as differentiable constraint losses, providing training-time constraint-based regulation that guides the corrected precipitation field toward physically consistent solutions. The method is evaluated from 2021 to 2023 in Hunan Province, China, using hourly numerical weather prediction model outputs as input features, China Meteorological Administration Land Data Assimilation System gridded analysis data as the training target, and independent meteorological station observations for strict cross-validation. PCPNet reduces the mean absolute error by 22.1% compared with the uncorrected China Meteorological Administration Land Data Assimilation System gridded precipitation products and outperforms Linear Regression, Bagging, Boosting, Multi-Layer Perceptron, TabNet, and Tree-based Progressive Regression Models by 12.9%, 13.5%, 16.9%, 10.8%, 14.9%, and 15.9%, respectively. The single-day event analysis provides an initial demonstration of heavy precipitation recovery capability, while comprehensive validation across long-term continuous weather events is planned for future operational deployment to further verify model stability. These results indicate that bio-inspired sensing and regulatory constraint learning can improve both the accuracy and interpretability of precipitation nowcasting correction.