中国地区SRTM数据高程误差的分布特征

发布时间：2017-12-27 09:29

本文关键词：中国地区SRTM数据高程误差的分布特征　出处：《西北大学》2015年硕士论文　论文类型：学位论文

【摘要】：SRTM的1arc分辨率(约90m)高程数据自发布以来,在很多地学研究中都得到广泛的应用,其数据质量也是需要重点关注的问题。本文首次对中国陆地范围SRTM数据高程误差的大小、分布情况及与地形、地表覆盖等影响因素之间的关系进行了系统性评价。在空间配准的基础上提取了覆盖整个中国陆地范围50多万个样本点位置上的地形特征信息,分析高程误差在不同地形特征条件下的变化规律。研究发现：SRTM数据高程以及坡度、坡向信息能够较好的表达中国陆地范围不同地区的地形特征；SRTM高程数据的平均误差为-0.35m,90%误差为7.4m,达到了绝对高程误差小于16m的设计要求,但是SRTM数据中存在范围较小、明显可见的数据异常区域；坡度是SRTM数据高程误差变化的关键因素,同时也对其他地表特征条件下的误差变化规律造成影响；随着坡度的增大,SRTM高程误差由正变负,误差绝对值增大,最大达到-0.86m；正误差集中出现在315-360°和0-450的北坡方向,负误差集中在45-315°的偏东以及偏西南坡向；受坡度影响较小的平坦地区,植被覆盖区造成的高程误差集中在0.18～1.67m,与植被覆盖度的变化成正相关关系；而误差的变化幅度随坡度和植被覆盖度的增加逐渐增大。在冰川、沙漠、湿地区的平均误差为负,误差大小分别为-1.05m,-2.03m,-2.43m,而城镇建筑区的平均误差为正,误差大小1.05m。整体上,由于复杂的地形分布和多样的地表覆盖类型,SRTM数据高程误差在中国地区的变化规律较为复杂。
[Abstract]:SRTM's 1arc resolution (about 90m) elevation data has been widely applied in many Geosciences since its publication. Its data quality is also a matter of great concern. For the first time, we systematically evaluated the relationship between the size and distribution of the height error of land SRTM data and terrain and surface coverage in China. On the basis of spatial registration, we extract the terrain feature information covering about 500000 sample points in the whole Chinese territory, and analyze the change rule of elevation error under different terrain characteristics. It is found that the SRTM data of elevation and slope information to terrain features in different regions of the land Chinese expression better; the average error of SRTM elevation data for -0.35m, the 90% error is 7.4M, meets the design requirements of absolute height error is less than 16m, but there is a small range of visible abnormal areas of data in SRTM data the slope is the key factor; SRTM data error of elevation change, but also the impact on the error variation of other surface features conditions; with the increase of slope, are changed by SRTM elevation error negative error absolute value increases, the maximum error is -0.86m; in 315-360 and 0-450 degrees north direction, negative error concentrated in 45-315 degrees east and west to the South; flat areas less affected by slope effect, the elevation error caused by vegetation area in 0.18 ~ 1.67m, and The change of vegetation coverage is a positive correlation, and the variation of the error increases with the increase of slope and vegetation coverage. The average error in Glacier, desert and wetland area is negative, the error is -1.05m, -2.03m and -2.43m respectively, while the average error in urban construction area is positive and the error is 1.05M. On the whole, due to complex terrain distribution and various types of surface coverage, the variation of SRTM data height error in China is more complex.
【学位授予单位】：西北大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P208;P207.1

【相似文献】