喜马拉雅地区冰湖变化及其溃决灾害风险研究

Glacial Lake Outburst Flood (GLOF) is one of the most catastrophic natural hazards in many high mountainous communities all over the world with a clear characteristics of sudden dam-break, extreme damage and widely social and economic effects. Himalayas, as the natural borders among China, Bhutan, Nepal and India, are known as one of the hardest hit regions by many documented historical GLOF hazards there. The continuous glacial retreats and rapidly expanding glacial lakes in Himalayas definitely imply an increasing risk of GLOF due to the global warming, which have been given more attention by local governments and scientists in recent years with considering their catastrophic destroy and fatalities. Till now, limited researches in the mechanism and key physical factors of GLOF result in the lack of more accurate risk assessing indexes and model. Both of these restrain the mitigation and hazard prevention of GLOF a lot. In this study, we try our best to reveal the glacial lake changes from 1990 to 2015 in the whole Himalayas at five time points based on remote sensing images from a regional insight to a specific lake analysis. The second objectives is to improve the probability model of risk assessment for GLOF based on mechanism and key factors of outburst from field survey, historical GLOF events and experimental observations as well as simulations in our laboratory. Then, high risk glacial lakes will be selected by using the new-built model with glacial lake data extracted from remote sensing images in 2015. The third objective is to identify the extents and intensities of GLOF by an improved outburst simulation model and objectively assess the potentially social and economic damages. The final steps is to give some usefully positive strategies of mitigation and hazard prevention , as well as specific measurements while emergences appear. Totally, this study aims to guarantee local people’s assets and safety and promote the awareness of GLOF from a scientific insight, serving for construction of coordinated anti-disasters among Himalayan countries.

冰湖溃决灾害是一种高山区严重的自然灾害，其爆发突然、破坏力大、影响范围广。喜马拉雅山脉是全球冰湖溃决灾害最严重的地区之一，历史上溃决事件多发。全球变暖背景下，该地区冰川消融，冰湖快速扩张，冰湖溃决灾害风险在增加，引起了地方政府和学者们的广泛关注。目前，对冰湖溃决的机理和关键参数研究非常有限，制约了当地的防灾减灾工作，区域尺度的溃决风险评价模型仍需要改进。本研究以整个喜马拉雅山脉为对象，从区域到局地，遥感监测1990-2015年冰湖的变化特征；结合野外观测数据、历史溃决事件和模拟实验，揭示冰湖溃决的机理和关键参数，改进溃决风险评价概率模型，完成冰湖溃决风险评价，筛选出高危的冰湖；利用优化的溃决模型模拟高危冰湖溃决可能影响的范围和程度，评估造成的损失，从而提出主动的防灾减灾适应建议，服务于跨境协同减灾抗灾，保障当地人们的社会经济利益和生命财产并提高冰湖溃决灾害研究的理论水平。

喜马拉雅山脉是全球冰湖溃决灾害最严重的区域之一，在全球变暖背景下，冰川消融，冰湖快速扩张，冰湖溃决灾害风险在增加。为了提高冰湖溃决灾害的认知水平，本项目规划了喜马拉雅地区冰湖变化、历史冰湖溃决灾害调查、冰湖溃决灾害风险评估及冰湖溃决模拟和影响分析四个主要研究内容。项目组基于卫星遥感数据、无人机遥感、野外考察数据、气象观测数据等，采用遥感技术、GIS、数值模拟等方法从多个尺度分析喜马拉雅山脉冰湖变化、构建灾害数据库、完成典型流域灾害风险评估和典型案例的溃决模拟与影响评估。研究结果表明：1990-2015年，喜马拉雅冰湖总数增加了401个，冰湖面积扩张了14.1%（56.4 km2），其中冰川末端相连的冰碛湖扩张最为显著，贡献了83.1%（46.9 km2）。在118个快速扩张的冰湖（具有较高的潜在危险）中，有69个位于我国西藏境内，气候变暖将导致部分冰湖持续的扩张，冰湖溃决风险增加。基于遥感技术和地貌特征分析方法可以有效的检验高山区历史冰湖溃决事件的真伪，之前报道的冰湖溃决事件存在一些不实报道，一些溃决事件的参数已被更正。吉隆藏布流域冰湖危险性评估结果显示：流域内大多冰湖属于极低和低危险性，6个冰湖处于高危险性，需要对这些高危冰湖进行重点监测。冲巴雍错流域的案例揭示：冲巴吓错溃决洪水爆发在2001年，是由冰崩涌浪触发的，冰川持续退缩、陡峭地形和上游的巨大冰储量及冰川运动是冰崩涌浪的形成条件，虽然下游的2个冰湖极大的缓冲了溃决洪水，降低了溃决洪水的破坏力，但是冲巴吓错溃决洪水仍导致了较大的损失。本项目完成的喜马拉雅地区冰湖调查数据和历史冰湖灾害数据集将为区域的冰湖冰川变化研究、冰湖灾害风险评估提供宝贵的数据和理论支撑。项目研究结果不仅提升了公众对冰湖灾害的认识和理解，而且获得了学术界和新闻媒体的广泛关注，亦可应用于当地的防灾减灾工作。