ارزیابی روش‌های مختلف پهنه‌بندی خطر وقوع زمین‌لغزش در اکوسیستم‌های جنگلی

Nowadays, forest engineers design forest roads network by utilizing a variety methods of zoning, attempt to identify low-risk variants of forest roads as landslide, so reduce costs of roads maintenance. Because of association between landslides and routing forest roads, finding the appropriate method of landslide zoning that also have potential to use in different areas of the forest is one of the most important steps for the realization of an intelligent design by experts and forest engineers. The purpose of this study is to achieve an appropriate method for mapping landslide to design the best forest road network. There have been many different methods introduced for landslide hazard zonation in the world that are divided to three methods including statistical (e.g., bivariate, multivariate, logistic regression, and information value model), heuristic or empirical (e.g., Stevenson, Nielsen and Brabb, Anbalagan, and Mora-Vahrson) and combinitation (e.g., artificial neural network, and fuzzy logic). Among statistical models” bivariate”, empirical “Mora-Vahrson”, and combination methods “artificial neural network” are most common. The results suggest that the use of statistical models are more capable and better than empirical models in the landslide risk zonation, although the combination models (e.g., artificial neural network, and fuzzy logic) have the highest accuracy and precision among other methods. If the number of available data is low, certainly combination methods can be more accurate than other models with the landslide hazard in the forest.