Abstract Reanalysis products and global climate models are the foundation of weather and climate prediction; however, they encounter challenges in capturing the spatiotemporal changes of near‐surface wind speed (NSWS) in a topographically complex area. Based on deep learning with ERA5 variables, the changes in NSWS were well captured across the Tibetan Plateau (TP), and the simulated NSWS bias can be reduced by 50.0% when the factor and location self‐attentions were considered. The reanalysis products overestimated the probabilities of light air, exceeding 40.0%, and underestimated the probabilities of gentle breeze, reaching 30.0%. Compared to some mainstream reanalysis products, the prediction success rate and threat score of the deep learning model for different wind strength events were improved, and its false alarm ratio and missing alarm ratio were decreased. Meanwhile, the predictive indices showed superior spatial homogeneity. This study offers a valuable reference for improving the NSWS simulation over a complex topography region.
Observations from the Little Bunker
Last updated: August 18, 2025 04:13 PM
Atmospheric CO₂
425.4 ppm
+18.4
ppm/year
Atmospheric CH₄
1934.91 ppb
+9.11
ppb/year
Global Temperature
+1.48 °C
Peak: +1.73°C
World Population
8.1 billion
+67
million/year