Assessing The Landslide Susceptibility Of Watershed Units Of Fujian Province Based On GIS

In recent years, the frequent landslides has brought great threat to people's life and property, and seriously restricted the development of social economy. Fujian Province is a typical area severely afflicted by landslide in China, so study on the spatial distribution of landslides and its susceptibility are significance for landslide prevention. This paper summarizes the causes of the landslide in Fujian Province based on field survey and existing landslide research of Fujian. Then, statistical methods of frequency and density were used to analyze the quantitative relationship between landslide and 11 environmental factors. In the end, the method of calculating the contribution weight of assessment factors to landslide is proposed by using the Gene coefficient of the landslide environment factor, and the landslide susceptibility evaluation model for watershed unit is constructed. And the model was used to identify the landslide susceptibility of 7492 watershed units in Fujian province. In this study, the following 3 conclusions are obtained.Conclusion 1:This paper summarizes the causes of the landslide in Fujian Province based on field survey and existing landslide research of Fujian area. Landslide is one of the most serious geological disasters in Fujian Province, which is affected by the regional human activities, rainfall, topography, geological structure, vegetation and so on. Human activities are main reason to landslide, such as building house at slop toe, planting on steep slope, large-scale construction, and mining exploitation.etc. Meanwhile, rainfall accelerated the occurrence of landslidesConclusion 2:Altitude, slop, topographic relief, aspect, curvature, average annual precipitation, vegetation coverage, topographic wetness index, stream power index, land use and soil type were selected as the analytical factors based on the causes of the landslide in Fujian Province. Then, the frequency and density of landslides distribution in the above mentioned factors were calculated, combining with 5166 landslide sample points. And two results were obtained. Firstly, regions in precipitation zone of 1500-1800mm, vegetation coverage of 50-80%, altitude of 100-800m, slop of 10-25°, and topographic relief amplitude of 100-200m are prone to landslide incidents; The main distribution aspects of landslides location are south and southwest, north and northwest aspects are less distribution; From the slop type of the landslides, both concave and convex slopes are distributed, but the distribution density of convex slope is higher than that of concave slope; Landslide incidents mainly occur in the red soil and paddy soil, farmland is the most sensitive to landslide, followed by construction land and garden plot. Secondly, landslide incidents are prone to happen in the zone where altitude in 0-100m and slope greater than 10°, altitude in 100-800m and slope less than 20°, altitude greater than 800m and slope less than 25°; And human activities are the main driving factor of landslide incidents in the region of altitude less than 100m.Conclusion 3:River system was extracted by the hydrological tools in ArcGIS software based on ASTER GDEM data of 30 m resolution in Fujian Province. And then, 7492 watershed units were obtained based on the river system. Consideration the statistical result between landslide and environmental factors, altitude, slop, topographic relief, aspect, average annual precipitation, topographic wetness index, vegetation coverage, land use and soil type were selected as the assessment factors of landslide. The sensitivity values of each sub category of above factors were calculated by the landslide density ratio formula. The Gini coefficient of landslide environmental factor was put forward, which was used to calculate the weight of environmental factor to landslide. And then, landslide susceptibility assessment model of watershed unit was constructed to calculate landslide susceptibility in 7492 watershed units of Fujian province. The assessment results were verified by empirical probability model and the accuracy reach 75%. This new method can be used as reference for assessment of landslide susceptibility in large area. And the sensitivity map of watershed unit has important significance for landslide remediation, land-use planning.etc.