4.3 Article

Automated mapping of water bodies using Landsat multispectral data

期刊

LIMNOLOGY AND OCEANOGRAPHY-METHODS
卷 10, 期 -, 页码 1037-1050

出版社

WILEY
DOI: 10.4319/lom.2012.10.1037

关键词

-

资金

  1. Swedish National Infrastructure for Computing (SNIC) through Uppsala Multidisciplinary Center for Advanced Computational Science (UPPMAX) [s00111-69]
  2. Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning, FORMAS

向作者/读者索取更多资源

The assessment of the role of lakes and impoundments at regional and global scales, e.g., in biogeochemical cycles, requires good estimates of the areal extent and shape of water bodies. Upscaling to large regions, except in limited areas where precise maps are available, so far depends on statistical estimates of the number and size of lakes, which explains why estimates are poor. We present an automated procedure that allows mapping of the actual number, size, and distribution of lakes at large scale. Landsat 7 Enhanced Thematic Mapper Plus (ETM +) mosaics from the GeoCover Circa 2000 dataset covering the Earth land surfaces with 14.25 m spatial resolution were used as input data. We developed an approach called GWEM (GeoCover (TM) Water bodies Extraction Method) that combines remote sensing and GIS to extract water bodies and study their abundance and morphometry. All water bodies greater than 0.0002 km(2) were taken into account as lakes. The accuracy of the method was tested on Sweden, where detailed maps of lakes, based on in situ data and orthophotos, exist for the whole country. The proposed method produced accurate results. The largest sources of errors are shadows of mountains and clouds, since the GeoCover mosaics are not absolutely cloud free.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.3
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据