摘要：Soil erosion by water has become a serious problem in Pacific Islands mainly due to the tropical natural conditions and to the progressive and continuous human pressure (agricultural practices, bush fires, mining acti- vity¡). In the current study, an effort to predict potential annual soil loss has been conducted in Tahiti Iti island in French Polynesia. For the prediction, the Revised Universal Soil Loss Equation (RUSLE) has been applied in a Geo- graphical Information System framework. RUSLE-factor maps were made. The R-factor was determined from the average annual rainfall data. The K-factor was estimated using soil map available and granulometric analysis of samples of soils. The LS-factor was calculated from a 5 meters digital elevation model. The C-factor was calculated using remote sensing techniques and particularly supervised classification methods based on SPOT 5 satellite ima- ges. The P-factor in absence of data was set to 1. The mean annual soil loss predicted by the model is 2.4 t/ha/yr. The results show that some 85% of the region of interest has a potential erosion rate of less than 5 t/ha/yr but an extended part of the area is undergoing severe erosion with a potential soil loss up to 50 t/ha/yr, demanding the attention of local land managers.
关键词：Soil erosion; Revised Universal Soil Loss Equation (RUSLE); Geographic ; Information System (GIS); Integrated Coastal Zone Management (ICZM); Tahiti
