首页 >  2020, Vol. 24, Issue (6) : 776-786

摘要

全文摘要次数: 1111 全文下载次数: 1105
引用本文:

DOI:

10.11834/jrs.20208330

收稿日期:

2018-08-15

修改日期:

PDF Free   HTML   EndNote   BibTeX
滑坡蠕变与遥感影像上植被异常关系
郭忻怡1,2,郭擎1,冯钟葵1
1.中国科学院空天信息创新研究院, 北京 100094;2.中国科学院大学, 北京 100049
摘要:

以滑坡蠕变阶段坡体的蠕变会引起环境条件的改变,进而影响植被生长状况的野外考察客观现实为依据,提出一种间接监测滑坡变化的新方法。利用高分辨率光学遥感技术,对滑坡蠕变阶段遥感影像上坡体上覆植被的异常特征进行判识,建立遥感影像上植被异常与滑坡蠕变的关系,反映滑坡的演化过程,弥补GPS技术、InSAR技术及部分地面监测手段在地势高、地形陡峭、植被茂盛等条件下监测工作的不足,为后续的滑坡预测研究提供帮助。以植被覆盖度较高的新磨村山体高位滑坡为例,首先,对研究区域进行分区;其次,计算各分区的植被覆盖度;最后,利用植被覆盖度分析遥感影像上的植被异常与滑坡蠕变的关系,并根据滑后遥感影像和实地考察情况进行验证。结果显示,2014年—2016年,滑坡的主要物源区、变形体上方细长局部崩滑区和泉眼及冲沟周边的植被覆盖度出现明显的下降,即随着滑坡发生时间的临近,植被受滑坡蠕变的影响变大,植被生长状况变差;而且随着距裸地等滑坡风险较大区域的距离增大,植被受滑坡蠕变的影响变小,植被生长状况变好。这表明,植被异常与滑坡蠕变存在明显的时空相关性,体现了滑坡蠕变阶段遥感影像上植被异常与滑坡蠕变的内在联系,反映了滑坡逐步失稳的演化过程,为进一步预测滑坡的发生提供依据。

Relationship between landslide creep and vegetation anomalies in remote sensing images
Abstract:

Landslides bring great peril to mountainous areas in China. In recent years, catastrophic high-position landslides often occur after the Wenchuan earthquake. The landslides occurred in the mountains that have high or steep terrain and dense vegetation coverage. A typical case in Xinmocun occurred in Diexi Town, Maoxian County, Sichuan Province on June 24, 2017. Given the characteristics of high position and concealment, this type of landslide is difficult to be detected by GPS, InSAR, and other traditional investigation methods. Certain technologies should be promoted and applied to detect and prevent such high concealed landslides at high positions. The optical remote sensing technology, owning a special capability with large-range, non-contact, periodic coverage, and abundant data, has great potential in making up for the limitations of the above methods, which is of great significance to disaster prevention and mitigation. The creep of landslide causes changes in environmental conditions, such as loosening of rock mass, change in soil nutrient, and uneven distribution of water. Changes in environmental conditions cause vegetation growth to vary. In situ investigation is conducted on the anomaly of vegetation growth before landslide. Therefore, on the basis of vegetation anomaly, this study establishes a new indirect landslide monitoring method to prepare for the study of landslide prediction. This method identifies vegetation anomaly on the landslide body using high-resolution optical remote sensing images, establishes the relationship between the creep of landslides and vegetation anomaly, and analyzes the evolution process of landslide creep at the stage of potential landslides.The Xinmocun landslide, which has high vegetation coverage, is selected as an example for conducting experiments. First, according to the comprehensive interpretation of optical remote sensing images and geological data, the Xinmocun landslide area is divided into upper landslide hazard, middle potential impact, and lower human activity. Second, vegetation coverage in each area is calculated using three-time series optical remote sensing images (June 18, 2014; June 21, 2015; and June 28, 2016) before landslides. Finally, the relationship between vegetation growth status and landslide creep is analyzed and verified using the remote sensing images and geological survey after the landslide.Experiments on high-resolution optical remote sensing images, which were conducted at the same period of three years before the landslide, detected changes in vegetation. In the upper landslide hazard area, the main landslide source and the narrow slump areas above the deformable body were affected by the creep of landslide. Hence, vegetation coverage declined evidently from 2014 to 2016. With the increase of distance from the edge of the bare land, the smaller the effect of landslide creep is, and the better the status of vegetation gradually becomes. With the time of landslide approaching, the greater the effect of landslide creep is, and the worse the status of vegetation in the same position becomes. In the middle potential impact area, the vegetation coverage around the springs and gullies declines with the greater effect of landslide creep as the time of landslide approaches. In the lower human activity area, the variation of vegetation coverage has no obvious regularity because of complex factors.Through experiments of optical remote sensing data, the conclusions are drawn as follows. The status of vegetation in the upper landslide hazard area and the middle potential impact area has significant temporal and spatial correlation with landslide creep. This result reflects the inherent relationship between vegetation growth and the creep of landslides, which can be used to predict the occurrence of landslides.

本文暂时没有被引用!

欢迎关注学报微信

遥感学报交流群