GIS and Predictive Modeling: a Comparison of Methods Applied to Schima superba Potential Habitat and Decision Making

摘要

Recently, in order to alleviate the impact of environmental issues due to global warming, forestation, which may increase forest cover and thus enhance the quantity of carbon sequestration, has become one of the most effective measures for carbon reduction. However, most of the previous plantations could not have achieved the goal of carbon reduction, since they did not follow the principle of "planting the right tree at right place". We attempted to predict the potential habitat for planting Chinese guger-tree {Schima superba) in the Huisun study area in central Taiwan. We applied GIS to overlay the tree samples collected with GPS on the layers of elevation, slope, aspect, terrain position, and vegetation indices derived from SPOT-5 images for modeling the tree's habitat. We developed decision tree (DT), logistic multiple regression (LMR) and discriminant analysis (DA) models to predict the tree's potential sites in Huisun. Results based on the tree samples from Tong-Feng watershed in Huisun indicated that the accuracy of DT was slightly better than that of LMR, accuracies of the two models were much better than that of DA; and the three models were highly efficient in habitat modeling. DT and LMR models greatly reduced the area of field survey to less than 10% of the entire study area at the first stage, they were better suited for potential habitat modeling. Vegetation indices may not be useful for improving model accuracy for the widely distributed species. However, the reasons are still waiting to be further assessment.