Study On Shallow Landslides Distribution And Rainfall Threshold Considering Vegetation Factors

To clarify the spatial distribution of shallow landslides induced by extremely heavy rainfall in the well-vegetated area on the western edge of the Loess Plateau,the relationship between vegetation and shallow surface landslides was quantitatively analyzed.In this study,the spatial distribution of shallow surface landslides induced by heavy rainfall in 2013 in the Niangniangba area of Tianshui,Gansu Province was studied by remote sensing interpretation and sensitivity analysis method,and the vegetation coverage was quantified by using the green index(GRVI).Then,the relationship between GRVI and landslide rainfall threshold was refined according to regional typical stands.The hydrological process and topographic conditions of the slope were comprehensively characterized by the area-slope equation.Finally,the rainfall threshold model was calculated for the typical small watershed in the study area,and the spatial heterogeneity of the parameters was quantified by vegetation index.The main conclusions are as follows:(1)Firstly,the non-cumulative coefficient was 1.771 with a landslide area density of 7.54%.Landslides are mainly distributed in the range of 1600-1750 m,with a slope of 20-25°.The slope is the most sensitive to the scale of a single landslide,and the sensitivity coefficient is 456.9.and landslide distribution had a close relationship with the distance to fault and river system,with a higher quantity on the south-facing slope which appeared to tend more landslide density on the south-eastern aspect.Secondly,landslide density or absolute quantity regardless of its aspect was found to have a non-monotonic change with GRVI increasing which had a threshold that the landslide density increase before GRVI to peak value of 0.10 to 0.20,then decrease to a constant.Thirdly,there were different ranks of sensitive factors according to their sensitivity factors which revealed that vegetation played a different role in different lithology.Vegetation can play both positive and negative effects on landslide occurrence.Lithology property and vegetation were combined to influence landslide distribution.(2)Hydrological and topographic conditions indicate that vegetation plays an important role in mitigating,marginal,and causing landslides.For the slope gradient below 32°,vegetation plays a role in slowing down and preventing the occurrence of the landslide.For the slope gradient of 32°–36°,vegetation plays a marginal role,and the difference in GRVI is not significant.Vegetation plays a negative role in the occurrence of landslides with a slope of more than 36°.Roots improve soil permeability,and the higher the vegetation coverage,the higher the saturated drainage rate.The saturated soil transmissivity of Pinus tabuleaformis(10.87 m~2/d)is higher than that of Larix kaempferi(2.47 m~2/d)because the root system is deeper and more distributed than that of Larix kaempferi.The rainfall threshold of Pinus tabuleaformis forest(83.33 mm/d)was higher than that of Larix kaempferi(53.29 mm/d),the root system can stabilize the slope by increasing the critical rainfall threshold.(3)Vegetation characteristic parameters can quantify spatial heterogeneity and play an important role in the distribution characteristics of shallow surface landslides.The correlation of soil saturation conductivity under different GRVI was fitted as y=1240 x~2-383 x+42(R~2=0.76,P<0.001).The vegetation coverage index can further accurately quantify regional geotechnical mechanics and hydraulics characteristics from the spatial scale.The Logistic regression model of integrated vegetation coverage and rainfall threshold proposed in this paper can provide a reference for analyzing the occurrence time of cluster landslides under the condition of insufficient data of rainstorm and can be used as an important index for disaster early warning and prediction.At the same time,GRVI and rainfall threshold have a high degree of explanation for the occurrence of landslides,and the regression coefficients of the two variables on the vulnerability of landslides are both negative,indicating their contribution to the vulnerability of landslides,and they have good practicability for the shallow landslide distribution area.