冰川密林陡谷区活动性滑坡InSAR精细识别-以澜沧江梅里雪山段为例

西南山区地质灾害频发,近年来,连续多起如茂县滑坡、白格滑坡、水城滑坡等大型滑坡灾害表明,灾难型滑坡事件的发生不仅与该地区易孕育地质灾害的条件相关,无法准确识别与监控滑坡灾害的发育发展情况也是主要因素之一。位于青藏高原东南缘的“三江”地区,地形高陡、地质灾害频发,其中澜沧江流域环境更为恶劣,海拔高、地形陡,夏秋季山高林密、多降雨,春冬季天气寒冷、多降雪,传统的地面式地质灾害调查手段在此效果有限,光学遥感技术也受到天气条件的诸多限制。而基于主动式微波遥感成像的InSAR技术被证明适合进行人类不易到达地区的地质灾害识别工作。本文以澜沧江流域梅里雪山段为研究区,使用PALSAR1、PALSAR2与Sentinel-1多源雷达遥感数据,结合D-InSAR与IPTA-InSAR技术开展冰川密林陡谷区的滑坡灾害精细识别工作。过程中发现长波长的PALSAR1和PALSAR2数据有更好的适应性,短波长的Sentinel-1数据通过数据量多的优势弥补了自身穿透能不足的缺陷;成功解译了研究区内92处活动性滑坡灾害;平面面积超过20万m_²的有26处;时间序列InSAR结果结合地面调查验证表明,澜沧江河谷典型的大型滑坡,梅里水滑坡和亚贡滑坡活动速率快,危险性高;德钦县成周边典型滑坡,丁羊丁滑坡与归巴顶滑坡整体处于蠕变状态,局部变形速率高;同时德钦县城周边泥石流与冰碛物运动较为活跃。

Accurate Identification of Active Landslides in Region Composed with Glacier, Forest, Steep Valley: A Case Study in the Lantsang Meili Snow Mountain Section

Geological hazards occur frequently in the southwest mountainous area. In recent years, successive large-scale landslide hazards such as Maoxian landslide, Baige landslide, and Shuicheng landslide have shown that the occurrence of catastrophic landslide events is not only related to the conditions that are prone to breeding geological hazards in the area, and failure to accurately identify and monitor the development of landslide disasters is also a major factors. Located in the "Sanjiang" area on the southeastern edge of the Qinghai-Tibet Plateau, the terrain is high and steep with frequent geological disasters. The condition of Lantsang watershed is more severe, with high altitude and steep terrain. The mountains and forests are dense in summer and autumn and there is a lot of rainfall. It is cold and snowy in spring and winter. However, the traditional ground-based geological hazard survey methods have limited effects here, and optical remote sensing technology is also subject to many limitations of weather conditions. InSAR technology based on active microwave remote sensing imaging has proven to be suitable for the identification of geological hazards in areas not easily reached by humans. In this paper, a segment of Meri Snow Mountain in the Lantsang Basin is used as the study area, and multi-source radar remote sensing data, such as PALSAR1, PALSAR2, and Sentinel-1 data, are used in combination with D-InSAR and IPTA-InSAR technologies to carry out accurate identification of landslide disasters in steep valleys of glacial dense forests. It was found that the long-wavelength PALSAR1 and PALSAR2 data have better adaptability. The short-wavelength Sentinel-1 data makes up for the shortcomings of insufficient penetration energy by virtue of the large amount of data. It successfully interpreted 92 active landslides in the study area, there are 26 sites with a plane area of ??more than 20×10_⁴ m_²; the results of time series InSAR technology combined with ground survey verification show that the typical large landslides in the Lantsang Valley, such as the Meili water landslide and the Yagong landslide have a high activity rate and high danger. As typical landslides around Deqin County area, the Dingyangding landslide and Guibading landslide are in the creep state as a whole, and the local deformation rate is high. At the same time, debris flow and moraine movement around Deqin County are more active.