<p>Spatial and temporal variations in wind speed critically influence wind energy development and planning. This study investigates the spatial–temporal characteristics of mean wind power density (WPD) in Yunnan Province, a mountainous region in southwest China with abundant wind resources. We use gridded observational data and outputs from 16 CMIP6 global climate models (GCMs) to analyze historical WPD changes (1970–2020) and project future trends (2025–2060). Two Shared Socioeconomic Pathways are considered: SSP2-4.5 (medium emissions) and SSP5-8.5 (high emissions). The spatial distribution of WPD reveals that eastern and northern Yunnan consistently exhibit high WPD values (&gt; 60 W·m⁻²), while southwestern regions show much lower values (&lt; 20 W·m⁻²). Historically, WPD declined significantly, particularly in Kunming and Qujing, rates exceeding -20 W·m⁻²·decade⁻¹ (p &lt; 0.05). Spring and winter had the highest WPD, highlighting strong seasonal variability. To further quantify WPD variability and resource stability, several metrics were computed. The STR, M_V, C_V, and RCoV for the region were 0.18, 4.74, 0.77, and 0.51, respectively, indicating moderate resource richness and variability. Future projections suggest that under both scenarios, WPD will generally decline in the near-term and mid-term, particularly in northern Yunnan. However, under SSP2-4.5, localized increases (up to + 4 W·m⁻²·decade⁻¹) are projected in central areas, such as Chuxiong and Dali. In contrast, under SSP5-8.5, long-term reductions in WPD exceed -4 W·m⁻²·decade⁻¹. Despite projected declines in WPD, especially under high-emission scenarios, the overall spatial distribution and seasonal peaks suggest that wind energy utilization efficiency may remain relatively stable in key regions.</p>