基于空天地技术的滑坡识别与智能监测预警

Recently, Landslide prevention and hazard mitigation is not only an unsolved problem in the world, but a great demand in China. However, it is highly needed to overcome the following issues regarding the landslide analysis, such as landslide identification, monitoring, assessment and forecast. Nowadays, innovative space remote sensing and satellite navigation technologies have bring unprecedented chances to landslide research. This proposal will focus on the development of key theory and method of landslide hazards by using spaceborne, airborne, ground-based and even underground technologies and equipments. Firstly of all, regarding landslide identification, SAR interferometry (InSAR) techniques will be proposed, including determination of effective coherent targets, selection of fine functional models, optimization of deformation algorithms and assessment of landslide identification results. Secondly, as for landslide inventory mapping, one automatic/semi-automatic procedure will be pointed out by fusing high-resolution optimal remote sensing image, InSAR and LiDAR techniques. Thirdly, in order to solve the problems of landslide real-time monitoring, including high cost, low accuracy and instability, one landslide monitoring sensor and equipment will be developed with small size, low cost and high accuracy characteristics. Lastly, to solve the difficulties of landslide stability analysis, forecast and early-warning, a novel landslide stability analysis model combining high accuracy monitoring information and physical damage mechanism will be researched, furtherly, landslide early-warning system based on multi-source information will be constructed. This research can be expected to provide a great support to landslide prevention and control with modern technologies in geomatics and spatial information technologies, which will undoubtedly have scientific theory significance and practical engineering reference.

滑坡防控既是世界难题，更是国家重大需求。但是目前在滑坡识别、监测、评估和预警四大环节上均有待新突破。新兴的空间遥感和卫星导航技术给滑坡研究带来机遇。本项目拟利用空、天、地表、地下技术开展滑坡灾害的关键理论方法研究。首先针对滑坡识别问题，提出利用InSAR滑坡动态早期识别技术，攻克监测点优化选取、精细误差模型、参数优化算法和识别可靠性分析等关键技术；针对滑坡动态编目问题，提出利用高分光学遥感、InSAR和LiDAR融合技术的滑坡自动/半自动识别动态编目；针对滑坡实时监测中高成本、低精度和稳定性差等技术瓶颈，研发小型化低成本高精度滑坡实时监测传感器技术及设备；针对滑坡稳定分析和预警问题，研究高精度监测形变信息为先验约束与滑坡变形破坏物理机制相结合的滑坡稳定分析，构建基于多源信息的滑坡预警信息系统。项目预期研究成果可提升现代测绘科学技术和空间信息技术对滑坡防控的支撑能力，具有重要科学和工程价值。

滑坡防控既是世界难题，更是国家重大需求，本项目围绕空、天、地技术在滑坡识别、监测、评估和预警四个环节开展关键理论方法研究：第一，提出了多项InSAR技术的滑坡形变识别关键技术，提升了InSAR滑坡识别和监测的精度及可靠性，并开发了一套融合高分遥感、InSAR和LiDAR数据的滑坡动态早期识别数据处理系统，实现滑坡的自动/半自动识别及动态编目。第二，研发了小型化低成本高精度滑坡一体化监测传感器设备，解决了监测设备成本高的瓶颈；研制出北斗/多源传感器多尺度全天候融合处理云平台，实现了滑坡全方位实时高可靠性监测。第三，开发了多源监测信息约束下滑坡稳定性分析模型和顾及地下水位与降雨不确定性的WOA-GPR动态代理滑坡稳定性评价模型，提供了滑坡稳定性评价的新方法。第四，提出了包括基于时间序列与变态模式分解的混合核极限学习机的滑坡位移高精度预测模型，构建了甘肃黑方台黄土滑坡示范区的GIS实时监测预警可视化系统。.研究成果提升了现代测绘科学技术和空间信息技术对滑坡防控的支撑能力，社会效益和经济效益显著。项目成果已在全国10余个省市自然资源、应急管理、地矿勘察、交通运输、城市规划、市政建设等领域得到推广应用。研制的北斗新技术装备入选自然资源部普适型示范北斗监测设备，并已推广应用至陕西、四川、云南、贵州等十五个地质灾害严重的省份，完成了300余套设备的部署。2019年至今，五次成功预警滑坡灾害，均实现人员零伤亡。滑坡监测预警应用成果被西班牙媒体、北斗官网、人民网等国内外百余家权威媒体报道。