Time series analysis of optical remote sensing data for the mapping of temporary surface water bodies in sub-Saharan western Africa
Time series analysis of optical remote sensing data for the mapping of temporary surface water bodies in sub-Saharan western Africa
复制标题
DOI：
10.1016/j.jhydrol.2009.02.052
复制
发表时间：
2009-05-30
影响因子：
6.4
通讯作者：
Combal, B.
中科院分区：
文献类型：
--
作者：
Haas, E. M.;Bartholome, E.;Combal, B.
临时小水体（TSWB）在1公里分辨率的地图推导出的撒哈拉以南非洲西部和半干旱和亚湿润干旱地区的实际地表水发生的时空分布表现出很高的跨和年内的变化。水体和潮湿地区已被绘制和分析的10个日常小水体（SWB）地图的基础上SPOT植被（VGT）数据跨越1999年1月至2007年9月期间。对SWB时间序列的进一步分析提供了有关水体季节性重现及其水文功能的更多信息。从连续时间序列中得到的地图确保了包含临时特征，这与基于几个单一日期观测的其他数据集相比是一个明显的优势。该方法以不同时段、不同区域的海水SWB时间序列为基础，采用分层随机抽样方法和一阶段聚类分析方法，快速生成TSWB图，并利用高分辨率卫星数据对检测到的水体进行验证。仅考虑委托误差，整个研究区域的总体精度为95.4%，其中干旱和半干旱气候区的结果最好。绘制水体的方法提供了令人满意的结果，特别是对于植被稀疏的地区，以及平坦的地区的研究区域。在更潮湿、更多植被的地区和山区，由于表面特征，错误检测的可能性增加。(c)2009爱思唯尔有限公司版权所有。
A map of temporary small water bodies (TSWB) at 1 km resolution was derived for the and, semi-arid and dry sub-humid regions of sub-Saharan western Africa where the spatio-temporal distribution of actual surface water occurrence exhibits high inter- and infra-annual variability. Water bodies and humid areas have been mapped and characterized by the analysis of 10 daily small water bodies (SWB) maps based on SPOT VEGETATION (VGT) data spanning the period January 1999-September 2007. Further analysis of the SWB time series provided additional information about the seasonal recurrence of water bodies as well as their hydrological function. A map derived from a continuous time series assures the inclusion of temporary features, a clear advantage compared to other datasets, which are based on several single date observations. The method described in this paper targets at a rapid creation of TSWB maps based on the SWB time series for different time intervals and regions.An accuracy assessment has been carried out with a stratified random sampling approach and a one-stage cluster analysis that relies on high-resolution satellite data to verify the detected water bodies. The overall accuracy, considering only the commission error, is 95.4% for the whole study region, with best results in the arid and semi-arid climate zone. The method to map water bodies delivers satisfactory results, particularly for sparsely vegetated areas as well as flat areas of the study region. In more humid, more vegetated areas and in mountainous areas, the possibility of false detections increases due to surface characteristics. (c) 2009 Elsevier B.V. All rights reserved.