Study On Theory And Method Of Rainfall-induced Landslide Early-warning In Ailao Mountain

ABSTRACT:In this thesis, the author selected Xinping County in the east of Ailao Mountain as the working area, studied the theories and methods for rainfall-triggered regional landslide early warning, analyzed the susceptibility between landslide and geo-environment features, constructed landslide early warning indicators and multi-parameter prediction model, developed the landslide warning analysis and issuing system. The results were used in the landslide hazard mitigation work during2008-2010, and were proven to be effective.The research work in this thesis can be summarized as the following:(1) The rough landform and well-developed folds and faults of Xinping County determined it is a landslide-prone area. The extreme rainfall events in recent years triggered a large amount of landslide hazards. The study of landslide warning in Xinping County can overcome the defects of wide range and low accuracy in national and provincial early warning serves, and can yield good economic benefits.(2) The author analyzed the susceptibility between landslide and different geo-environment features with certainty factor adjustment (CF) model, the high landslide sensitive geological environment features are:1500-2000m elevation;800-1400mm average annual rainfall;20～30°slope; magmatic rocks;500m distance to main structure; erosion-depositing landform. Then the weights of different segments were calculated using the certainty factor (CF) model results.(3) According to results of landslide susceptibility of different geo-environment features, the author established the logistic regression prediction model, calculated the landslide susceptibility of the whole county, which is divided into17617grid cells. Then the susceptibility values of different cells were classified into five categories. The evaluation results reflect the characteristics of landslide hazards in Xinping County, and quantitatively provide the landslide probability of each cell, which is an important basis for landslide early warning.(4) Under the guidance of the of explicit statistical forecasting theory, the author presented the landslide warning thresholds and multi-parameter comprehensive warning model, which includes landslide susceptibility zonation, village maximum potentiality parameter, antecedent effective precipitation value and intraday induced precipitation value, antecedent effective precipitation and slope, etc. Then, the author developed the determination method and the determination form. When antecedent effective precipitation is0or very little, the intraday induced precipitation is110mm for early warning,220mm for alarm. While when the antecedent effective precipitation is33mm, the intraday induced precipitation is30mm for early warning.(5) The work of landslide early warning system consists of warning program system development, consultation, issuing and response. The criterion for monitoring sites selection and instruments layout principle are summarized. In the paper, the author gave an introduction of the landslide monitoring point selection, monitoring equipments, the monitoring network, and monitoring data-analyzing, main function design, information issuing and so on. Monitoring and early warning system also includes the early warning consultation and response.(6) Based on GIS, Xinping County landslide monitoring and early warning system was developed. According to the checkout of landslide early warning service from2008to2010, the warning model and system are proved to be feasible and practical. Three warnings were issued in2008-2010. The Xiaotuanshan landslide was successfully predicted by this system. And there were still2empty warnings; so the early-warning model and system should be gradually improved though detailed study.