Zhiqiu Gao

To address the nonlinear amplification of numerical weather prediction (NWP) errors and the difficult trade-off between coverage and sharpness in short-term offshore wind power forecasting, this paper proposes PRWind, a physics-guided framework for short-term probabilistic forecasting. Built upon a Transformer encoder, the framework integrates horizon-adaptively weighted power-curve, persistence, and wind-speed cubic-law baselines; propagates wind-speed uncertainty into a power distribution through a probabilistic wind-speed distribution with Monte Carlo integration; and employs spatial graph convolution to model inter-turbine correlations, together with online bias correction and segmented post hoc interval calibration. The framework is validated on two offshore wind farms in coastal China (8350 kW and 10,000 kW units), with all data analyzed at hourly resolution and forecasts issued for 23 hourly steps ahead. For point forecasting, the NMAE reaches 8.74% and 8.24%, with MAE reductions of 19.1% (95% CI: 18.1–19.9%) at Wind Farm A and 7.8% (95% CI: 7.3–8.4%) at Wind Farm B relative to the strongest baselines, both statistically significant under the Diebold–Mariano test (p < 0.001). For probabilistic forecasting, the PICP remains stable at 0.89, balancing coverage and sharpness. The error of PRWind accumulates markedly more slowly with the forecast horizon than that of the baselines, and ablation studies confirm the synergistic effect of the modules. Trained and evaluated independently on the two wind farms, PRWind performs consistently, demonstrating cross-wind-farm stability (rather than zero-shot generalization) conferred by physical guidance, and offering significant engineering value for intra-day and day-ahead offshore wind power dispatch.

To improve solar irradiance simulation accuracy for precise photovoltaic power forecasting, we developed a hybrid framework combining WRF-Solar numerical simulation and random forest (RF) machine learning for a PV plant in Guangdong, China. Weather conditions were objectively classified into clear, intermittent cloudy, and overcast using the Daily Variability Index (DVI) and Daily Clear-sky Index (DCI). We calibrated the WRF-Solar model’s microphysics and radiative transfer schemes via sensitivity tests to optimize overcast-sky performance, then applied RF correction to the simulated irradiance. Results show that RF correction significantly reduces simulation errors for intermittent and overcast conditions, while the original WRF-Solar outperforms the corrected results under clear skies due to RF overfitting.

This paper evaluates the performances of mean diurnal variation (MDV), nonlinear regression (NR), lookup tables (LUTs), support vector regression (SVR), k-nearest neighbors (KNNs), gradient boosting (XGBoost), long short-term memory (LSTM), gated recurrent units (GRUs), and the Transformer model with a deep self-attention mechanism to interpolate the turbulent heat fluxes missing from a prairie observation on the Tibetan Plateau. Results indicated that the Transformer model outperformed the other methods that were tested. To further enhance the interpolation accuracy, a combined model of Transformer and a convolutional neural network (CNN), termed Transformer_CNN, was proposed. Herein, while Transformer focused primarily on global attention, the convolution operations in the CNN provided the model with local attention. Experimental outcomes revealed that the interpolations from Transformer_CNN surpassed the traditional single artificial intelligence model approaches. The coefficient of determination (R²) reached 0.95 in the sensible heat flux test set and 0.90 in the latent heat flux test set, thereby confirming the applicability of the Transformer_CNN model for data interpolation of turbulent heat flux on the Tibetan Plateau. Ultimately, the turbulent heat flux observational database from 2007 to 2016 at the station was imputed using the Transformer_CNN model.

Accurate real-time icing grid fields are critical for preventing ice-related disasters during winter and protecting property. These fields are essential for both mapping ice distribution and predicting icing using physical models combined with numerical weather prediction systems. However, developing precise real-time icing grids is challenging due to the uneven distribution of monitoring stations, data confidentiality restrictions, and the limitations of existing interpolation methods. In this study, we propose a new approach for constructing real-time icing grid fields using 1339 online terminal monitoring datasets provided by the China Southern Power Grid Research Institute Co., Ltd. (CSPGRI) during the winter of 2023. Our method integrates static geographic information, dynamic meteorological factors, and ice_kriging values derived from parameter-optimized Empirical Bayesian Kriging Interpolation (EBKI) to create a spatiotemporally matched, multi-source fused icing thickness grid dataset. We applied five machine learning algorithms—Random Forest, XGBoost, LightGBM, Stacking, and Convolutional Neural Network Transformers (CNNT)—and evaluated their performance using six metrics: R, RMSE, CSI, MAR, FAR, and fbias, on both validation and testing sets. The stacking model performed best, achieving an R-value of 0.634 (0.893), RMSE of 3.424 mm (2.834 mm), CSI of 0.514 (0.774), MAR of 0.309 (0.091), FAR of 0.332 (0.161), and fbias of 1.034 (1.084), respectively, when comparing predicted icing values with actual measurements on pylons. Additionally, we employed the SHAP model to provide a physical interpretation of the stacking model, confirming the independence of selected features. Meteorological factors such as relative humidity (RH), 10 m wind speed (WS₁₀), 2 m temperature (T₂), and precipitation (PRE) demonstrated a range of positive and negative contributions consistent with the observed growth of icing. Thus, our multi-source remote-sensing data-fusion approach, combined with the stacking model, offers a highly accurate and interpretable solution for generating real-time icing grid fields.

This study analyzes the surface winds of Typhoon In‐fa during landfall at a coastal area using Doppler wind lidar (DWL) observations in the Plane Position Indicator (PPI) scanning mode. High‐resolution (30 m) wind field structures in the lower boundary layer at the land‐sea interface were derived using the velocity‐azimuth display technique. This approach captures the wind field structure in unprecedented detail at the scale of tens of meters, providing a fine‐scale view of the coastal wind field during typhoon landfall. It was found that stronger winds were predominantly located at the periphery of the scanning area, where the underlying surface is water, whereas weaker winds tended to occur near the radar, where the underlying surface is land. Analysis of vorticity and divergence revealed small‐scale vortex structures, approximately 100 m in size, embedded within the large‐scale circulation of the typhoon. These structures became less distinct as the typhoon center approached. Furthermore, the turbulent energy spectrum derived from the directly observed radial winds followed Kolmogorov's −5/3 law, indicating that the turbulence was generally isotropic. Additionally, the turbulent kinetic energy (TKE) calculated from the DWL data showed good agreement with previous tower‐based observations, particularly in terms of the relationship between TKE and wind speed, confirming the reliability of DWL measurements in capturing turbulence characteristics. These results suggest that DWL can serve as an effective alternative for future turbulence observations during typhoon landfall.

Accurate real-time icing grid fields are critical for preventing ice-related disasters during winter and protecting property. These fields are essential both for mapping ice distribution and for predicting icing using physical models combined with numerical weather prediction systems. However, developing precise real-time icing grids is challenging due to the uneven distribution of monitoring stations, data confidentiality restrictions, and the limitations of existing interpolation methods. In this study, we propose a new approach for constructing real-time icing grid fields using 1,339 online terminal monitoring datasets provided by the China Southern Power Grid Research Institute Co., Ltd. (CSPGRI) during the winter of 2023. Our method integrates static geographic information, dynamic meteorological factors, and ice_kriging values derived from parameter-optimized Empirical Bayesian Kriging Interpolation (EBKI) to create a spatiotemporally matched, multi-source fused icing thickness grid dataset. We applied five machine learning algorithms—Random Forest, XGBoost, LightGBM, Stacking, and Convolutional Neural Network Transformers (CNNT)—and evaluated their performance using six metrics: R, RMSE, CSI, MAR, FAR, and fbias, on both validation and testing sets. The stacking model performed best, achieving an R value of 0.634 (0.893), RMSE of 3.424 mm (2.834 mm), CSI of 0.514 (0.774), MAR of 0.309 (0.091), FAR of 0.332 (0.161), and fbias of 1.034 (1.084), respectively, when comparing predicted icing values with actual measurements on pylons. Additionally, we employed the SHAP model to provide a physical interpretation of the stacking model, confirming the independence of selected features. Meteorological factors such as relative humidity (RH), 10-meter wind speed (WS10), 2-meter temperature (T2), and precipitation (PRE) demonstrated a range of positive and negative contributions consistent with the observed growth of icing. Thus, our multi-source remote sensing data fusion approach, combined with the stacking model, offers a highly accurate and interpretable solution for generating real-time icing grid fields.

High-quality observational data play a crucial role in deepening the investigation of the Tibetan Plateau’s influence on the Asian climate. This study employs eight machine learning models (support vector regression (SVR), k-nearest neighbors (KNN), extreme gradient boosting (XGBoost), random forest (RF), long short-term memory (LSTM), gated recurrent unit (GRU), Transformer, and Transformer–convolutional neural network (Transformer-CNN)) to interpolate missing observational data on surface net radiation (Rn), soil surface temperature (Ts), soil water content (SWC), air temperature (Ta), relative humidity (RH), and wind speed (WS) from the QOMS observation site. The data covers the period from 1 January 2007 through to 31 December 2016. A comparative evaluation of these models shows that the Transformer-CNN model consistently outperforms the other models in terms of prediction accuracy. On the test dataset, the coefficients of determination for the interpolated results of Ta, RH, WS, SWC, Ts, and Rn were 0.97, 0.92, 0.97, 0.79, 0.93, and 0.98, respectively. Secondly, the Transformer-CNN model was then applied to generate a complete meteorological dataset for the full period. A time series analysis of this dataset reveals statistically significant trends over the past decade: air temperature (Ta) increased by 0.60 °C (p = 0.022) and soil temperature (Ts) by 1.85 °C (p = 1.37 × 10⁻⁵). Meanwhile, wind speed (WS), soil water content (SWC), and net radiation (Rn) declined by 0.42 m/s (p = 1.18 × 10⁻¹²), 1.24% (p < 0.001), and 9.21 W/m² (p = 8.81 × 10⁻⁶), respectively.

The limited understanding of the structure of the urban surface atmospheric boundary layer can be attributed to its inherent complexity, as well as a deficiency in comprehensive measurements. We analyzed one year of meteorological data and Doppler wind lidar measurements in Beijing to explore how atmospheric stability is influenced by wind speed, radiation, turbulence, and pollution levels. Results indicate that the predominant state of the urban boundary layers in Beijing is an active condition (characterized by strong unstable and unstable stability regimes) throughout the day, attributed to the significant heat storage capacity of the urban canopy. Strong stable regimes are more frequently observed during winter and autumn, peaking during transitions from night to day. Furthermore, both strong unstable and strong stable regimes occur under very weak wind conditions (indicating weak dynamic instability), with strong instability associated with high net radiation levels while strong stability correlates with low net radiation conditions (indicative of robust thermal stability). The unstable regime manifests under strong winds (reflecting strong dynamic instability) alongside moderate net radiation environments, characterized by elevated values of turbulence kinetic energy and urban boundary height, highlighting the critical role of mechanical turbulence generation during periods of high wind activity. Additionally, six instances of pronounced stable conditions observed during daytime can be partially attributed to low radiation coupled with high pollutant concentrations near the surface, resulting from prolonged temperature inversions due to intense radiative cooling effects and weak dynamic forcing. Our findings presented herein are expected to have urban boundary layer climate and environment implications for other cities with high pollution and dense urban infrastructure all over the world.

In this study, we propose a model called CNN-BiLSTM-AM that utilizes deep learning techniques to forecast severe convective weather based on ERA5 hourly data and observations. The model integrates a CNN with a Bidirectional Long Short-Term Memory (BiLSTM) system and an Attention Mechanism (AM). The CNN is tasked with extracting features from the input data, while the BiLSTM effectively captures temporal dependencies. The AM enhances the results by considering the impact of past feature states on severe weather phenomena. Additionally, we assess the performance of our model in comparison to traditional network architectures, including ConvLSTM, Predrnn++, CNN, FC-LSTM, and LSTM. Our results indicate that the CNN-BiLSTM-AM model exhibits superior accuracy in precipitation forecasting. Especially with the extension of the forecast time, the model performs well across multiple evaluation metrics. Furthermore, an interpretability analysis of the convective weather mechanisms utilizing machine learning highlights the critical role of total precipitable water (PWAT) in short-term heavy precipitation forecasts. It also emphasizes the significant impact of regional variables on convective weather patterns and the role of convective available potential energy (CAPE) in fostering conditions conducive to convection. These findings not only confirm the effectiveness of deep learning in the automatic identification of severe weather features but also validate the suitability of the sample dataset employed. Given its remarkable performance and robustness, we advocate for the adoption of this model to enhance the forecast of severe convective weather across various business applications.

Numerical experiments using the WRF model were conducted to analyze the sensitivity of Typhoon Mangkhut intensification simulations to seven widely used planetary boundary layer (PBL) parameterization schemes, including YSU, MYJ, QNSE, MYNN2, MYNN3, ACM2, and BouLac. The results showed that all simulations generally reproduced the tropical cyclone (TC) track and intensity, with YSU, QNSE, and BouLac schemes better capturing intensification processes and closely matching observed TC intensity. In terms of surface layer parameterization, the QNSE scheme produced the highest C_k/C_d ratio, resulting in stronger TC intensity based on Emanuel’s potential intensity theory. In terms of PBL parameterization, the YSU and BouLac schemes, with the same revised MM5 surface layer scheme, simulated weaker turbulent diffusivity K_m and shallower mixing height, leading to stronger TC intensity. During the intensification period, the BouLac, YSU, and QNSE PBL schemes exhibited stronger tangential wind, radial inflow within the boundary layer, and updraft around the eye wall, consistent with TC intensity results. Both PBL and surface layer parameterization significantly influenced simulated TC intensity. The QNSE scheme, with the largest C_k/C_d ratio, and the YSU and BouLac schemes, with weaker turbulent diffusivity, generated stronger radial inflow, updraft, and warm core structures, contributing to higher storm intensity.