Nonparametric statistical approach to wind farm siting in Poland using GIS

The present work examines the influence of environmental, topographical, and natural factors on the operational parameters of wind farms in Poland. Motivated by the ongoing transition to renewable energy and recent regulatory changes, the analysis utilises a comprehensive dataset of all wind farms in Poland, 2585 wind turbines. In statistical analysis, two grouping parameters were applied-wind farm area (0-5 km², 5-10, 10-15, 15-20 and 25-30 km²) and the power capacity of single turbines (0-1 MW, 1-2, 2-3, 3-4 MW)-to assess spatial and technical variability. A range of non-parametric statistical tests, including the Kruskal-Wallis test and Dunn's post-hoc test with false discovery rate correction, was used to evaluate differences among subgroups. The results reveal that wind farms of 15-20 km² were farthest from roads (126.7 m), while those of 25-30 km² were the closest (27.3 m). In terms of forest proximity, farms of 10-15 km² were the closest (median: 280.2 m), while those of 5-10 km² were significantly more distant (479.7 m). Wind farms with 3-4 MW turbines were more distant from surface water (744.5 m) than those with 1-2 MW turbines (615.0 m). Surprisingly, larger farms (25-30 km²) were closer to Natura 2000 areas (3494 m) than smaller ones (0-5 km², 5505 m). This indicates a complex interaction of factors influencing the location of wind farms. These findings provide valuable insights for optimising site selection and supporting sustainable wind energy development in Poland, depending on the planned area of investment and wind turbine model.