论文部分内容阅读
天山中段的山地针叶林带很大程度上影响了该地区整体卫星雪盖的识别精度,多角度卫星遥感技术的发展为林区积雪识别提供了新的途径。本文选取了2000年4月至2001年6月,10个时段研究区内无云覆盖的(Multi-angle Imaging Spectro Radiometer)MISR多角度数据,首先对红光波段不同角度观测结果组成的角度谱图像进行非监督分类,以确定天山林带的分布区域,然后在玛纳斯河中下游与那拉提山东部选取典型像元,分析这些像元红光波段各角度反射率在林区不同积雪覆盖状况下的表现差异。研究发现,若林区存在积雪,0°,±26.1°,±45.6°五个观测角度反射率的平均值大于0.1,在部分降雪月份,后向45.6°观测的反射率大于天顶方向观测的2.5倍。根据这一结论,给出基于MISR数据的研究区不同时段的积雪识别结果。结果表明,MISR红光波段对林区积雪反应敏感,不同角度观测的反射率在林区有雪和无雪时差异较大,故可利用多角度遥感信息进行林区积雪识别。
The mountain coniferous forest belt in the middle of Tianshan Mountain affected the recognition precision of the whole satellite snow cover in a large extent. The development of multi-angle satellite remote sensing technology provided a new way for identifying snow cover in forest area. In this paper, MISR multi-angle imaging Spectro Radiometer multi-angle data from April 2000 to June 2001 in 10 study areas were selected. Firstly, the angle spectral images composed of the observed results from different angles in the red band Unsupervised classification to determine the distribution of Tianshan forest belt, and then select the typical pixels in the middle and lower reaches of the Manas River and the eastern Nalati Mountain, analysis of these pixel red light reflectance at various angles in the forest area of different snow cover Under the performance differences. The results show that the average reflectance of the five observation angles of 0 °, ± 26.1 ° and ± 45.6 ° is greater than 0.1 in the Wulin area. The reflectance of the backward 45.6 ° observation in some snowfall months is greater than that observed in the zenith direction 2.5 times. According to this conclusion, the snow recognition results of different periods of the study area based on MISR data are given. The results show that the MISR red band is sensitive to the snow cover in forest area, and the reflectivity observed at different angles is quite different when there is snow and no snow in the forest area. Therefore, multi-angle remote sensing information can be used to identify snow cover in forest area.