基于高分七号卫星DEM的坡度坡长计算精度评价
DOI:
作者:
作者单位:

北京师范大学地理科学学部

作者简介:

通讯作者:

中图分类号:

基金项目:

国家自然科学基金项目(No.42177308)


Evaluation of the accuracy of extraction of slope and length based on DEM of GF-7 satellite
Author:
Affiliation:

Faculty of Geographical Science, Beijing Normal University

Fund Project:

The National Natural Science Foundation of China (No.42177308)

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    [目的]坡度和坡长作为重要的地形因子,其计算的准确性直接会影响到坡面土壤侵蚀模型的评价精度。高分7号卫星的高精度高程测量能力在地形因子的准确提取上具有较大潜力,期望能将其利用到坡度和坡长提取之中,实现对地形的准确表达。这需要对利用高分7号卫星DEM提取坡度坡长的结果进行精度评价。[方法]利用地形因子(LS)计算工具,对陕西省吴堡县4个小流域高分7号卫星、1∶10 000地形图、激光雷达影像生成的DEM进行坡度、坡长的提取。采用相关系数(r)、相对偏差和绝对偏差作为评价指标,评价高分7号提取地形因子的精度。[结果]与激光雷达和1∶10 000地形图的坡度和坡长相比,高分7号提取的流域平均坡度偏低7.50%~9.02%、坡长偏大1.83%~19.35%,但坡长和坡度分级的面积比例并不存在显著差异;高分7号提取的坡度结果偏差(16.46%~44.26%)明显小于坡长的偏差(75.25%~140.87%);高分7号所提取的坡度坡长在沟间地上的相对偏差(坡度15.48%~56.63%;坡长50.02%~130.79%)明显低于沟谷地(坡度21.28%~63.61%;坡长93.01%~192.51%)。[结论]利用高分7号DEM提取坡度坡长时,对于获取坡长坡度面积分级和对指定点的坡度提取时是可行的,而对于指定点的坡长提取时则结果不够可靠,在研究流域中坡长坡度的空间分布上具有较大优势,而提取具体某一位置处的坡长坡度时的优势则不明显。

    Abstract:

    [Purpose] Slope and slope length are essential topographic factors affecting soil erosion process, and thus should be accurately estimated for erosion modelling. The Gaofen-7 satellite is known for its capacity of high-precision elevation measurement, and therefore holds a great potential for a high-quality extraction of the topographic factors such as slope and slope length. This study aims to evaluate the performance and accuracy of Gaofen-7 satellite in terms of deriving slope and slope length. [Methods] DEMs generated from the GF-7 satellite images, 1:10 000 topographic map and LiDAR images were processed using LS calculation tool to derive slope and slope length for four small watersheds in WuBu County, Shaanxi Province. Performance indicators including correlation coefficient (r), relative bias (RE) and absolute bias (AE) were used to estimate the discrepancy between different data sources. [Results] The results show that: (1) the elevation data of the three sources have extremely strong correlations (r>0.956). Compared with the topographic map and LiDAR, the watershed-average slope derived from GF-7 is lower by 7.50%~9.02% while slope length is higher by 1.83%~19.35%. However, the area distributions of slope length and slope derived from the three data sources have no significant difference. (2) Comparing the results of slope and slope length, the discrepancy in slope (RE: 16.46%~44.26%) derived from different sources is significantly lower than that (RE: 75.25%~140.87%) in slope length. (3) The discrepancy between different sources is enhanced in gully areas (RE of slope: 15.48%~56.63%; RE of slope length: 50.02%~130.79%), while being decreased in inter-gully areas (RE of slope: 21.28%~63.61%; RE of slope length 93.01%~192.51%). [Conclusion] It is feasible to obtain areal grading characters of slope or slope length and slope of specified locations using DEMs generated from Gaofen-7 satellite, while the results of slope length for specified locations are less reliable. In conclusion, the Gaofen-7 satellite has a great advantage in describing the spatial distribution of slope length and slope at the watershed scale, while being less capable of capturing slope length and slope at a specific site.

    参考文献
    相似文献
    引证文献
引用本文
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2023-06-26
  • 最后修改日期:2023-08-12
  • 录用日期:2023-08-31
  • 在线发布日期: 2024-01-27
  • 出版日期: