遥感技术在现代冰川变化研究中的应用

The Remote Sensing in Research of Modern Glacier Changes

传统的现代冰川变化研究主要以实地观测和经验公式来获得冰川的面积变化、体积变化和冰川表面运动速度,从20世纪80年代以来随着航空遥感光学图像、数字高程模型、雷达等新技术数据的不断出现和发展,借助遥感手段研究冰川的性质和特征、监测冰川的动态变化成为冰川学研究发展的重要趋势,也有效解决了现代冰川研究中高山区资料受限等问题。冰川面积变化的计算机自动解译方法主要有阈值法和雪盖指数阈值法、监督分类非监督分类法、比值阈值法等。阈值法和雪盖指数阈值法操作简单,但是在阈值选取方面不好把握;非监督分类法操作简单但限制因素较多;如果训练区选择准确,监督分类法分类结果比较精确;波段比值阈值法操作相对简单,精度准确,是目前运用最多的计算机自动分类方法。数字高程模型和雷达数据在冰川体积变化研究中的应用,较好地提高了传统方法的精度和适用性;而且雷达数据具有不受大气传播和气候影响的特点,很好地弥补了遥感光学影像数据极易受云雪等影响的不足。高精度GPS和雷达数据的不断发展和应用在冰川表面运动速度的研究中起着不可替代的作用。

Traditional research methods in modern glacier changes mainly based on field observations and empirical formula to calculate changes in glacier area,volume and surface velocity,however,since 1980s,with the new technologies of aerial remote sensing images,digital elevation model,radar,etc.developed,the studies of nature and characteristic of glaciers and detection of dynamic changes in glacier have not only become the important trends of glacier research by means of remote sensing method,but also greatly resolved the problems such as limited information in high mountain areas.The computer-based automatic interpretation methods in the study of glacier area change including threshold and NDSI（Normalized Difference Snow Index）,supervised classification and non-supervised classification,threshold value of band ratio,etc.Although the operating of threshold and NDSI are simple,they are difficult to select the suitable threshold on such issues;Non-supervised classification method is simple but more constraints.With the accurate selection of training area,the result of supervised classification method is more accurate.The threshold value of band ratio method is the most widely used computer automatic classification because of its simple operation and accurate result.The application of Digital Elevation Model and radar data in the study of glacier volume change preferably improves the accuracy and applicability of traditional methods.And that radar data will not be affected by atmospheric propagation and climate change compensated for the weak script of remote sensing images data which is easily influenced by snow and clouds.With the development and application of high precipitation GPS and radar data,it plays an irreplaceable role in surface velocity studies.