Modelling the risk of snow damage to forests under short-term snow loading

Abstract

Regression models are developed to assess the risk of snow damage to Scots pine ( Pinus sylvestris L.), Norway spruce ( Picea abies (L.) Karst) and birch ( Betula spp.) stands based on simulated data, employing a mechanistic wind and snow damage model developed by Peltola et al., 1998a. The risk is predicted in terms of the critical windspeed needed to cause stem breakage and uprooting of trees at forest edges under short-term snow loading. Separate regression models are developed for each tree species using stem taper (breast height diameter of stem relative to tree height, d 1.3/ h), stand density, snow loading and distance from the stand edge as variables, and a general model for stem breakage and uprooting is also proposed having tree species as an additional dummy variable. The overall risk of stem breakage and uprooting is shown to increase with snow loading and decrease with increasing stem taper and stand density for all three tree species, although Scots pines and Norway spruces are predicted to be much more susceptible to snow damage than birches, which, being leafless, had much less crown area for snow attachment and wind loading. The greatest susceptibility to stem breakage and uprooting is seen at the stand edge, where the risk due to wind loading is much greater than inside the stand. Under these circumstances, slightly tapering Scots pines and Norway spruces are found to be the most vulnerable under a snow load of 60 kg m −2, suffering damage at windspeeds of <9 m s −1 at a constant height of 10 m above the ground, i.e. these windspeeds enhance the risk, whereas higher speeds can be expected to dislodge the snow from the crowns. Birches will only exceptionally be broken and uprooted at windspeeds of <9 m s −1 according to the models developed here. Since the general models give rise to somewhat greater residuals compared with the simulated data than do the single tree species models, it seems that the latter will give more reliable predictions of the risk of snow damage. The models could be useful when discussing the risk of snow damage in connection with alternative forms of stand management, especially in high risk areas, enabling high-risk trees to be removed during thinning.