基于SAR影像的干旱区冲/洪积扇地貌面定量分期研究——以河西走廊西部沙漠区的疏勒河洪积扇为例  

Quantitative staging of alluvial fan geomorphic surfaces in arid areas based on SAR imagery:A case study of the Shule River alluvial fan in the western desert region of the Hexi Corridor

作　　者：杨勇忠 任俊杰 李东臣 YANG Yongzhong;REN Junjie;LI Dongchen(School of Emergency Management Science and Engineering,University of Chinese Academy of Sciences,Beijing 100049,China;National Institute of Natural Hazards,Ministry of Emergency Management of China,Beijing 100085,China)

机构地区：[1]中国科学院大学应急管理科学与工程学院,北京100049 [2]应急管理部国家自然灾害防治研究院,北京100085

出　　处：《地质力学学报》2023年第6期842-855,共14页Journal of Geomechanics

基　　金：国家自然科学基金项目(U2139201,41941016,U1839204);中国科学院重点部署项目(KFZD-SW-422);应急管理部国家自然灾害防治研究院基本科研业务专项(ZDJ2017-24)。

摘　　要：河流作用形成的洪积扇和河流阶地可以提供过去构造活动、气候变化和地貌演变过程的有效记录;而准确划分洪积扇地貌面的期次是开展环境变化及构造活动定量研究的基础。已有研究往往利用L波段数据SAR后向散射系数值作为地貌粗糙度替代参数,进行地貌面定量分期,但并未考虑不同时间数据源对分期结果的影响。以疏勒河洪积扇为研究对象,通过分析多时相L波段SAR数据后验统计指标以及大气评估条件,确定最佳数据源,并运用最大似然分类法对后向散射强度值进行分类,以实现地貌面的定量分期。结果表明:使用分期后验统计指标作为选取最佳时像影像数据的标准,可以获得更好的分期结果;L波段HH单极化数据可得到较好的分期结果,与C波段数据相比,对于不同年龄地貌面的划分更具优势,且数据更易获取,具备自动化分期潜力;SAR影像质量以及分期结果与成像时大气条件密切相关,而与季节相关性不大,因此建议优先选择成像时地表含水量较低的影像,例如,高蒸发强度的夏季。文章提出的这套对遥感数据质量分析并进行地貌面分期的方法可用于完成干旱地区大尺度冲/洪积扇的快速定量分期,为构造和气候的研究提供有价值的信息。The alluvial fans and river terraces formed by river processes effectively record past tectonic activities,climate changes,and geomorphic evolution processes.Accurately dividing the alluvial fan into stages is the basis for the subsequent research.Previous researchers used L-band SAR backscatter coefficient values as a substitute parameter for geomorphic roughness to achieve quantitative zoning of geomorphic surfaces.However,these studies did not consider the impact of different time data sources on the geomorphic surface results.This study selects the Shule River alluvial fan as the research object.It determines the optimal data source by analyzing the posterior statistical indicators of multi-temporal L-band SAR data and evaluating atmospheric conditions.The maximum likelihood classification method is used to complete the classification of backscatter intensity values and achieve quantitative staging of the geomorphic surface.The results indicate that the posterior statistical indicators of staging can be used as the standard for selecting the best temporal image data to obtain better staging results.L-band HH monopolarization data provides better staging results,demonstrating advantages in distinguishing landforms of different ages compared to C-band data.Moreover,L-band data is more accessible and holds potential for automated staging.SAR image quality and staging results are closely related to imaging atmospheric conditions but show minimal seasonal dependence.Therefore,the study recommends prioritizing images with low surface water content during imaging,such as in high-evaporation intensity summer seasons.The proposed method for analyzing remote sensing data quality and staging landforms can be applied to rapidly and quantitatively stage large-scale alluvial fans in arid regions,providing valuable information for studies on tectonics and climate.

关 键 词：SAR 洪积扇地貌面 定量分期 后向散射系数 粗糙度 最大似然分类 

分 类 号：P931.2[天文地球—自然地理学]