The development and validation of a windthrow probability model for Sitka spruce in Ireland.

Abstract

An empirical windthrow probability model for Sitka spruce (Picea sitchensis) in Ireland was developed and validated. Data were collected from a range of Sitka spruce stands on different sites. Logistic regression was used to determine which site, stand and silvicultural factors significantly influenced the probability of windthrow. Top height, top height squared, the regional location of the stand, soil type, and altitude significantly influenced the occurrence of windthrow in Sitka spruce. Whether or not the stand had been thinned was also important. A model to predict the probability of windthrow occurring in a forest stand was developed. A validation exercise indicated that only 6% of the stands for which the model predicted the probability of windthrow to be less than 1% actually experienced windthrow.