| Geological disasters in Nanjiang happen frequently, and more than 1000 geological disasters have been cataloged at present. The central and southern lithology of Nan Jiang is red layer, it belongs to slide-prone stratum, and in the role of external inducing factors a large number of geological disasters happened. In recent years, the probability of extremely heavy rainfall weather increases with the global warming. For example, “9.16”heavy rainfall in 2011 induced more than 1000 geological disasters in Nanjiang. In view of the huge number and size of the geohazards, it’s very necessary to conduct geohazards susceptibility mapping of Nanjiang for indicating the direction of the geohazards prevention and mitigation. The geohazards susceptibility evaluation is to predict the possibility of the occurrence of geological disasters. Having developed detailed field geological investigation of study areas based on Geological Survey of China’s geological survey project “the major geological disasters investigation, warning, and mapping of the southwest areas(12120113010100)”, and finished the geohazards susceptibility map of study areas.First of all, making an analysis on the regional geological conditions, the distribution law of geological disasters, the formation conditions, and inducing factors. Then according to the relationship between them and the geological disaster points, selecting slope, profile curvature, slope height, the types of rock and soil, the types of slope structure, topographic wetness index, distance to rivers, the precipitations from july to september, and diatance to roads as evaluation factors of the geological disasters susceptibility mapping. Then the evaluation indexes were divided into basic environmental factors and inducing factors based on the evaluation indexes characteristics. Slecting the Logistic regression model and the fuzzy comprehensive evaluation model to conduct the geohazards susceptibility mapping of Nanjiang. The geohazards susceptibility levels are divided into 4 grades, which are high susceptibility, moderate susceptibility, low susceptibility, and not susceptibility.Logistic regression model evaluation indexes enter the model by stepwise regression, according the results of regression analysis, ultimately selecting slope, slope height, the types of rock and soil, distance to rivers, the precipitations from july to september, and diatance to roads 6 evaluation indexes as evaluation factors of the geological disasters susceptibility mapping. Using SPSS software to determine evaluation index weights, geological disasters point density was adopted to define the targeted value of the evaluation indexes, applying ArcGIS to finish overlay analysis, finally the geohazards susceptibility map of Nanjiang based on Logistic regression model was completed.The fuzzy comprehensive evaluation model applied analytic hierarchy process(AHP) to define the weights of evaluation indicators, and two-level fuzzy comprehensive evaluation to conduct the susceptibility mapping. The evaluation data procession was based on programming by Python, and calculating the susceptibility level of the evaluation units, it simplified the complicated mathematical calculation process and improved the mapping efficience.ROC curve and Kappa coefficient were applied to evaluate the accuracy of two kinds of models, then selecting the figure by the Logistic regression model as the geohazard susceptibility map of Nanjiang by the contrast analysis.The results show that high susceptibility areas are mainly distributed in the southern and central areas of Nanjiang red layer areas and the low altitude valleys in the southern mountains, moderate susceptibility areas are distributed in red layer areas, low susceptibility areas are mainly distributed in the slopes the altitude is 1000~1500m, and not susceptibility areas are mainly distributed in high altitude regions whose altitude is more than 1500 m. Comparing with the actual survey results, geohazard susceptibility map is reasonable. |
PDF Full Text Request