适用于大河流域的栅格DEM洼地填充算法

发布时间：2018-07-17 07:10

【摘要】：针对地势平坦且空间范围较大的流域水系自动提取任务中,采用常规的洼地填充算法难以提取出完整水系,形成大量"断头河"的问题,提出了一种能够导出整个数字高程模型(DEM)所有像元被淹没的次序表的方法,由该次序表构成的矩阵代替原有DEM来实现流向的计算,进而提取出累积流向及分级河道。经用覆盖中心流域的高精度DEM以及覆盖黄河和长江等大型流域的DEM来测试,结果表明:用其他算法提取结果均发生了"断头河"错误,而该方法则能提取出完整的水系。这种方法能正确实施洼地填充和像元填平处理,可以适应于任意规模和精度的DEM填洼问题,具有较强的鲁棒性,克服了以往算法难以处理大河流域DEM洼地填充的不足。
[Abstract]:In order to solve the problem of automatic extraction of watershed water system with flat terrain and large spatial range, it is difficult to extract the whole water system by using conventional low-lying filling algorithm, thus forming a large number of "broken-head rivers". In this paper, a method is proposed to derive the order table of all the pixels of the whole digital elevation model (Dem). The matrix composed of the order table replaces the original Dem to realize the calculation of the direction of flow, and then the cumulative flow direction and the graded river channel are extracted. The high precision Dem covering the central watershed and the Dem covering the Yellow River and Yangtze River basins are used to test the results. The results show that the "broken head River" errors have occurred in the extraction results of other algorithms, and this method can extract the whole water system. This method can correctly implement the filling of depressions and pixel filling, and can be applied to Dem filling problems of arbitrary scale and precision. It has strong robustness and overcomes the shortcomings of previous algorithms which are difficult to deal with Dem filling in great river basins.
【作者单位】：河南理工大学资源环境学院;中国科学院遥感与数字地球研究所数字地球重点实验室;
【基金】：国家自然科学基金项目(41105074,41275108) 河南理工大学博士研究基金项目(B2011-038) 中国科学院数字地球重点实验室开放基金项目(2011LDE010)
【分类号】：P208

【相似文献】