*3.3. The 10-Year Return Levels of Maximum Wind Speed for Frozen vs. Unfrozen Soil*

Generally differences in maximum 10-year return level of wind speed between seasons of frozen and unfrozen soil are rather small (difference +/−1 m/s) in large part of Finland. Small differences were observed especially for CSS (Figure 4a maps) and SP (Figure 4b maps), of which results as a whole resemble each other closely with similar spatial patterns and direction of differences. On large scale, larger differences were observed for CSS and SP only in parts of northernmost Finland and in the coastal are of southwestern Finland, i.e., maximum 10-year return level of wind speed was about 1–2 m/s larger in soil frost season. Differences larger than +/−1 m/s were a bit more common for PP (Figure 4c maps). Areas with stronger winds during the unfrozen season were found across coastal areas, parts of eastern Finland, and northernmost Finland, respectively. Noteworthy, compared to CSS and SP, sign of the difference is opposite in the coastal areas and in the most northwestern part of Finland. Notable positive differences in PP restricts to western and central parts of Lapland.

**Figure 4.** Aggregated maps presenting the maximum of 10-year return level of maximum wind speed [m/s] from cardinal and intercardinal wind directions during season of frozen (left) and unfrozen (middle) soil on: (**a**) soil type CSS (spruce on clay/silt), (**b**) SP (pine on sand), and (**c**) PP (pine on peat). Map on right presents the difference between two seasons (m/s). Distributions present the differences in 10-year return levels between frozen and unfrozen soil seasons wind direction wise, divided into northern (NF, top), central (CF, middle), and southern (SF, bottom) Finland.

There were also some quite notable differences direction wise whether the bulk/peak of the distribution of differences was over or below zero (Figure 4a–c distributions). The effect of wind direction on distribution of differences was smallest in northern Finland, regardless of combination of forest and soil type. In southern Finland, for S and SE wind directions return level of wind speeds were stronger during soil frost season. Conversely, winds from W, SW, and E were characterized by stronger winds during unfrozen soil season, whereas for the rest of the directions, differences were more or less evenly distributed around zero. In central Finland, dependence on wind direction was similar for CSS and SP, with N, NE, SE, S, and NW directions having dominantly stronger winds during frozen soil season and W, SW, and E directions were characterized by stronger winds during unfrozen soil season or the differences were distributed rather evenly. For PP, more directions were characterized by stronger winds during unfrozen soil season, namely N, E, SW, W, and NW. Only NE, SE, and S directions had stronger winds mainly during frozen soil season.

All in all, large-scale differences were in general quite subtle and/or restricted to few areas. On the other hand, small-scale features were visible in the maps over the whole Finland (Figure 4). In detail, these local scale nuances in the behavior of wind multiplier downscaled return levels of 10-year maximum wind speeds are demonstrated, only for PP in this study, on 30 × 30 km area from northern Finland with more complex topography (Figure 1b) including multiple hills/fells with elevation changing between 40 and 270 m above sea level.

In the example area (Figure 1) used for more detailed analysis about the effects of wind multiplier downscaling, the strongest winds were from the south (Figure 5). This dictates the general large-scale characteristics of aggregated differences (Figure 6, middle), i.e., difference was positive on the majority of the study area. However, generally weaker winds from NW (Figure 5) also had a significant role when the effect of topography was taken into account via wind multipliers. As winds from NW were conversely stronger during unfrozen soil season, this together with relatively strong topographical forcing created isolated areas on hillsides where wind speed return level characteristics are deviating quite a lot from the general conditions of the area (Figure 6, middle). In this example case, stronger winds occurred during soil frost season, but there were also areas, mainly northwestern hillsides, where the situation was opposite.

**Figure 5.** Ten-year return levels of wind speed [m/s] for PP (pine on peat) (**a**) frozen soil season and (**b**) unfrozen soil season in the detailed study area of northern Finland (location and topography, see Figure 1). The aggregated maximum values are in the middle, surrounded by the return levels of wind speeds from each of the cardinal and intercardinal directions.

**Figure 6.** Differences of values presented in the Figure 5a,b. Positive values correspond to stronger winds during frozen soil season.
