青藏高原格拉丹东冰川动态监测及其变化成因分析  

Dynamic monitoring and change cause analysis of Geladandong glacier in Tibet Plateau

作　　者：张浔浔 赵阳刚 杨斌[1,2] 吴淑莹 段阳海 刘玉 ZHANG Xunxun;ZHAO Yanggang;YANG Bin;WU Shuying;DUAN Yanghai;LIU Yu(Civil-Military Integration Geological Survey Center,Field Scientific Observation and Research Station of Natural Resources in the Middle Reaches of the Brahmaputra River,Tibet Autonomous Region,Sichuan Chengdu 610036,China;Key Laboratory of Coupling Process and Effect of Natural Resources Elements,Ministry of Natural Resources,Beijing 100055,China;Sichuan Hua Di Building Engineering Co.,Ltd.,Sichuan Chengdu 610081,China;Chengdu Center of Hydrogeology and Engineering Geology of Sichuan Provincial Geology and Mineral Resources Bureau,Sichuan Chengdu 610081,China;Aerospace Information Research Institute Chinese Academy of Sciences,State Key Laboratory of Remote Sensing Sciences Institute of Remote Sensing Applications,Beijing 100094,China)

机构地区：[1]中国地质调查局军民融合地质调查中心,雅鲁藏布江中游自然资源西藏自治区野外科学观测研究站,四川成都610036 [2]自然资源要素耦合过程与效应重点实验室,北京100055 [3]四川省华地建设工程有限责任公司,四川成都610081 [4]四川省地质矿产勘查开发局成都水文地质工程地质中心,四川成都610081 [5]中国科学院空天信息创新研究院,遥感科学国家重点实验室,北京100094

出　　处：《中国地质调查》2024年第6期121-133,共13页Geological Survey of China

基　　金：西藏自治区科技计划项目“雅鲁藏布江中游水资源与生态效应监测体系构建及研究(编号:XZ202401JD0024)”;中国地质调查局“青藏高寒区资源与环境调查监测与评价(编号:DD20220881)”;“青藏高原沱沱河源区多年冻土对气候变化的响应(编号:2022KFKTC007)”联合资助。

摘　　要：格拉丹东冰川群是长江正源沱沱河的发源地,位于青藏高原腹地的唐古拉山脉,探讨其冰川面积变化过程及变化成因对长江源头生态水资源的可持续利用和水循环过程等具有重要意义。研究采用随机森林法开展冰川遥感影像解译,同时结合地形和气候数据探讨冰川面积变化的成因。结果表明:①通过对比常用的7种计算机自动分类方法并经过精度评价,最终确定随机森林法在开展大量连续的遥感影像自动解译过程中具有明显优势;②1999—2022年,研究区冰川面积呈波浪式递减趋势,共减少64.25 km^(2),减少比例为10.22%,减少趋势为2.94 km^(2)/a;③研究区内悬冰川的冰川末端形状较为平滑规则、对称、单峰明显,冰川舌的末端形状较为不规则,且边缘形状变化大,外表形态受地形控制明显;④海拔、坡向对冰川面积变化的影响显著,在气候因素中,气温是冰川变化的主要因素,和冰川面积呈负相关关系。研究成果对明确长江源地区近年来的冰川变化特征及生态水资源的可持续利用提供了科学依据和数据支撑。Geladandong glacier complex is the source of Tuotuo River,the primary source of Yangtze River.It is located in the Tanggula Mountains,which is in the remote interior of Tibet Plateau.The processes and causes of glacier area change are crucial for the sustainable use of water resources and the functioning of the water cycle in the headwater ecosystem of Yangtze River.The random forest method was employed to carry out the interpretation of glacier remote sensing images,and topographic and climatic data were integrated to investigate the underlying drivers of glacier area change.The results are as follows.①The random forest method is particularly suitable for the automatic interpretation of a large number of consecutive remote sensing images,after the comparative analysis of 7 most commonly used automatic computerized classification methods and the evaluation of their accuracy.②From 1999 to 2022,the glacier area in the study area exhibited a wavy decreasing trend,with a total decrease of 64.25 km^(2),representing a 10.22%reduction and a decreasing trend of 2.94 km^(2)/a.③The glacier terminus in the study area exhibits a relatively smooth and regular morphology,with a distinct single peak and symmetrical outline.In contrast,the glacier tongue terminus displays a more irregular morphology,with a variable edge and a shape that is evidently influenced by topographic factors.④The change of glacier area is significantly influenced by elevation and slope direction.Among climatic factors,temperature is the primary driver of glacier change,ex-hibiting a negative correlation.This research could provide the scientific basis and data supporting for glacier change characteristics and sustainable use of water resources in the primary source area of Yangtze River.

关 键 词：格拉丹东冰川 冰川变化 遥感监测 随机森林法 气候变化 

分 类 号：P343.6[天文地球—水文科学]