摘要
降雨及库水位涨落是引起库岸滑坡形变失稳的主要诱发因素,但滑坡位移速率对此类诱发因素的响应具有一定的滞后性,影响人类对滑坡所处运动状态的判断与预测。针对常规预测模型中未考虑时滞效应的问题,利用三峡库区新铺滑坡的GNSS位移监测数据、奉节气象站降雨数据以及三峡库区库水位涨落数据,通过对监测区内9个GNSS监测点的位移速率序列与降雨量、库水位高程序列进行时滞互相关分析,确定时滞参数,进而应用多变量灰色系统理论方法,建立了时滞GM(1,3)预测模型,并对滑坡位移速率进行预测验证。结果表明:三峡库区新铺滑坡位移速率与降雨量显著相关,对降雨量的响应滞后时间约为5 d,滑体中后部受降雨影响比前缘更明显;位移速率与库水位高程高度相关,对三峡库区库水位涨落的响应滞后时间约为31 d,滑坡前缘受库水位涨落影响更明显,且离长江越近,滞后时间越短;利用加入时滞参数的时滞GM(1,3)模型进行预测,模型拟合优度达到0.702,相比GM(1,1)模型和未顾及时滞因素的GM(1,3)模型,预测精度分别提升了53.8%和58.3%,平均绝对误差百分比分别降低了7.19%和7.47%,在滑坡位移速率预测及库岸滑坡防灾减灾领域具有一定的工程应用价值。
Rainfall and reservoir water level fluctuation are the main inducing factors of landslide deformation and instability,but the response of landslide displacement rate to such inducing factors is lagging,and affects the judgment and prediction of the movement state of landslide.In order to solve the problem that time-delay effect is not considered in the conventional prediction model,the GNSS displacement monitoring data of Xinpu landslide in Three Gorges reservoir area,rainfall data of Fengjie meteorological station and reservoir water level fluctuation data in Three Gorges reservoir area were used;through the time-delay cross correlation analysis of displacement rate,precipitation and reservoir water level elevation series of 9 GNSS points in the monitoring area,the time-delay parameters were determined;by using the multi-variable grey system theory,the time-delay GM(1,3)model was established to predict and verify the landslide displacement rate.The results show that the displacement rate of Xinpu landslide in Three Gorges reservoir area is significantly correlated with precipitation,and the response lag time to precipitation is about 5 days,the middle and rear parts of landslide are more affected by precipitation than the front edge,and the displacement rate is highly correlated with the reservoir water level elevation;the response lag time to reservoir water level fluctuation in Three Gorges reservoir area is about 31 days,the front edge of landslide is more obviously affected by the reservoir water level fluctuation,and the closer to Yangtze River,the shorter the lag time;using the time-delay GM(1,3)model,the goodness of fit of the model reaches 0.702;compared with GM(1,1)model and GM(1,3)model without time-delay factors,the prediction accuracy of time-delay GM(1,3)model is improved respectively 53.8%and 58.3%,and the mean absolute percentage error reduces by 7.19%and 7.47%,respectively.The time-delay GM(1,3)model has a certain engineering application value in the field of landslide displacement rate predictio
作者
黄观文
王家兴
杜源
白正伟
王铎
HUANG Guan-wen;WANG Jia-xing;DU Yuan;BAI Zheng-wei;WANG Duo(School of Geological Engineering and Geomatics,Chang an University,Xi an 710054,Shaanxi,China)
出处
《地球科学与环境学报》
CAS
北大核心
2021年第3期621-631,共11页
Journal of Earth Sciences and Environment
基金
国家重点研发计划项目(2018YFC1505105,2019YFC1509802)
国家自然科学基金项目(41731066,41941019)。
关键词
滑坡预测
库岸滑坡
互相关分析
时滞分析
时滞GM(1
3)
灰色模型
新铺滑坡
三峡库区
landslide prediction
reservoir bank landslide
cross correlation analysis
time-delay analysis
time-delay GM(1,3)
gray model
Xinpu landslide
Three Gorges reservoir area
