基于光学影像的冰川面积参数提取研究

发布时间：2018-07-20 18:29

【摘要】：随着全球气候变暖,全球大部分冰川都出现加速消融与退缩的趋势。监测冰川变化对于计算冰川对全球海平面上升的贡献量,分析全球气候的变化规律以及评估区域淡水资源的储量变化具有重要意义。而冰川面积是研究冰川变化的重要参数。目前,利用多时相遥感数据提取冰川信息已经成为冰川面积变化监测的主要手段。论文首先提出一种改进的冰川信息提取方法,并将改进的方法应用于青藏高原格拉丹东地区的冰川研究中,获取了该地区多时相的冰川面积参数,并分析了冰川变化与气象数据的关系。首次使用我国自主研制的HJ-1A/1B卫星影像提取冰川的面积参数。论文的主要工作和创新点如下： (1)提出了一种综合波段比值和水体指数的高山区冰川信息提取方法,与单独波段比值法及人工目视解译比较的实验结果表明,改进方法提取的冰川边界更加准确,消除了地表水对冰川信息提取的影响,提高了冰川制图精度与效率。 (2)采用综合波段比值和水体指数的改进方法提取了格拉丹东雪山1992、1999、2006及2009年的冰川和冰川湖的面积参数。通过分析发现,在1992-2009年期间冰川一直处于退缩状态,而冰川湖处于不断扩张状态,两者紧密相连,呈非线性负相关。最后通过对气象数据分析得到如下结论,夏季气温的显著升高可能是造成该区域冰川持续退缩的主要因素。 (3)首次采用HJ-1A/1B数据提取冰川信息。并对比验证了最大似然法、支持向量机法及基于HJ-1A/1B数据的决策树法等三种遥感分类方法提取冰川信息的精度；试验结果表明,支持向量机方法提取的冰川信息精度最高。图14幅,表9个,参考文献80篇。
[Abstract]:With global warming, most glaciers around the world are accelerating melting and retreating. Monitoring glacial changes is of great significance in calculating the contribution of glaciers to global sea level rise, analyzing the regularity of global climate change and evaluating the changes of regional freshwater reserves. The area of glacier is an important parameter to study the change of glacier. At present, extracting glacier information from multitemporal remote sensing data has become the main means to monitor glacier area change. In this paper, an improved method of extracting glacier information is put forward, and the improved method is applied to the study of glaciers in Gradendong area of Qinghai-Tibet Plateau, and the parameters of glacial area in this area are obtained. The relationship between glacier change and meteorological data is analyzed. For the first time, the HJ-1A / 1B satellite image is used to extract the glacier area parameters. The main work and innovations of this paper are as follows: (1) A method of extracting glacier information from alpine region is proposed, which combines the ratio of wave band and water body index. The experimental results show that the method is compared with the ratio method of single band and artificial visual interpretation. The improved method for extracting glacier boundaries is more accurate, eliminating the influence of surface water on the extraction of glacier information. The accuracy and efficiency of glacial mapping were improved. (2) the area parameters of glacier and glacial lake were extracted from Glacier and Glacier Lake in 1992 / 99 / 2006 and 2009 by the improved method of comprehensive band ratio and water body index. From the analysis, it is found that the glacier retreated from 1992 to 2009, while the glacial lake is in the expanding state. The two are closely related and have a nonlinear negative correlation. Finally, through the analysis of meteorological data, it is concluded that the significant increase of summer temperature may be the main factor that causes the continuous retreat of glaciers in the region. (3) the HJ-1A / 1B data is used for the first time to extract glacier information. The accuracy of maximum likelihood method, support vector machine method and decision tree method based on HJ-1A / 1B data are compared to verify the accuracy of the three remote sensing classification methods for extracting glacier information, and the experimental results show that the accuracy of glacier information extracted by support vector machine method is the highest. There are 14 figures, 9 tables and 80 references.
【学位授予单位】：中南大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P237;P343.6

【参考文献】