Abstract: Small watershed is a regional unit which has relatively intact natural ecological processes. Watershed vulnerability assessment has an extremely significant impact on the reasonable utilization of regional resource and sustainable development. Large-scale evaluation considering natural, social and even other factors is one of the difficulties encountered in current watershed vulnerability assessment study. The 8.0 Ms Wenchuan Earthquake has caused severe damage to the mountains of southwest China. Because of the earthquake, the mountain areas have been facing with debris-flow hazard and landslides more frequently and the mountain environment has been increasingly vulnerable. Secondary mountain disasters have seriously affected regional economic and social development and would constant influence on the production and daily life of local people in a long time. Therefore, it is of great essentiality to enrich small watershed vulnerability assessment, especially to coordinate the relationship among natural and social factors and the effects of catastrophic mountain environment. And it is also highly important for mountainous watershed disaster prevention, mitigation and regional sustainable development. Longmen Mountain of Pengzhou city, the case area in this study, is situated in the core impacted region and the characteristics of post-earthquake secondary mountain disaster are more obvious. The study area was divided into 77 small watersheds by GIS technology and remote sensing image with high resolution. With the comprehensive consideration of natural, social factors and disaster events, the evaluation index system of small watershed vulnerability was constructed with Analytic Hierarchy Process (AHP). And the evaluation of small watershed vulnerability was conducted from the aspects of disaster degree and social vulnerability in the combination of small watershed and grid as unit by the overlay tools of GIS. In the end, the degree of small watersheds vulnerability is classified into slight, moderate, high and extreme fragility, with the number of small watersheds being 12, 21, 27 and 17, respectively. And the corresponding area accounted for the entire study area is 11.53%, 38.63%, 40.46% and 9.38%, respectively. Taking the regional major disaster events from 2009 to 2013 to test and verify, the evaluation results are consistent with the actual situation. However, the influencing factors of watershed vulnerability are numerous and it is of certain difficulty to evaluate the watershed vulnerability quantitatively and exhaustively. There are also some deficiencies in social vulnerability analysis and socio-economic data digitization. Last but not the least, the methods applied in this paper are feasible for larger-scale watershed vulnerability assessment under catastrophic mountain environment, which can also provide a reference for regional vulnerability assessment.

Key words: catastrophic mountain environment, GIS, watershed, vulnerability, Longmen Mountain