ACTUALITY AND CASE STUDY ON MECHANISM OF POST-FIRE DEBRIS FLOW INITIATED BY SHALLOW LANDSLIDE

The initiation mechanisms of post-fire debris flow can be divided into two categories: surface runoff initiated(SRI) and shallow landslides initiated(SLI). To date, most researches focus on SRI post-fire debris flow, which normally occurs within the first two years after forest fire, causing serious surface erosion and civil destructions. However, there are relatively few researches on the SLI post-fire debris flow after a long time of the occurrence of forest fire, which restricts practitioner's cognition of the starting conditions and initiation mechanism of SLI post-fire debris flow. Compared with SRI post-fire debris flow, SLI debris flow is mainly attributed to the root-soil strength degradation. Hence, quantitative determination of the long-term decline of root and soil strength by supplying new experimental data is a necessity to understand the initiation mechanism of SLI post-fire debris flow. To achieve the objective, the current research was carried out in a source area of SLI post-fire debris flow with a mixed conifer forest dominated by masson pine in Muli County, Liangshan Prefecture, Sichuan province. Root quantity statistics, root ultimate tensile strength test, triaxial test on root-soil mixtures, and slope stability analysis were conducted from the samples of three different timeslots after forest fire. The results indicate that the root quantity and ultimate tensile force decrease with the time after fire. In addition, root-soil mixture experimental triaxial results revealed that cohesion can drop dramatically 9 years after forest fire. Accordingly, factor of safety of local slope is calculated to be less than 1. The main reason for the occurrence of shallow landslide after fire is the decrease of shear strength of root-soil mixture caused by tree mortality and subsequent decay of tree root networks. The research results provide scientific data to support the risk assessment of SLI post-fire debris flow.