基于DEM提取流域特征影响因子的分析

Analysis of Influence Factors for Extraction of Hydrological Basin Features from DEM

基于DEM的流域特征提取,受数据源、分辨率、重采样技术等因子影响。为系统地评价不同地貌类型对各影响因子的敏感程度,本文提出一种多指标因子分析模型,综合分析不同因子对流域特征,如流域面积、河流分叉比、河网密度、流域长度、流域坡度、流域高程、流域形状等的客观影响。实验过程选取山地、丘陵、平原地形的中小流域为研究区域,以ASTER 30m重采样生成的40—100m分辨率DEM,ASTER 30m、SRTM 90m和GMTED2010 250m的3种数据源及4种重采样技术生成的DEM场景为实验数据。结果表明:分辨率对流域特征影响最敏感,数据源次之,重采样技术不太敏感。山地地形下的流域特征变化趋势性显著,适合山地中小流域的分辨率为最高分辨率,数据源为SRTM,重采样技术为最邻近分配法。丘陵、平原地形评价结果具有随机性。

The extraction of hydrological basin features from DEM is influenced by some factors, i.e., DEM data source, DEM resolution and DEM resampling techniques. To evaluate the influence of these factors, this paper proposes a multi-criteria factor analysis model. This model can evaluate the Sensitivity of basin features, i.e., basin area, river bifurcation ratio, river network density, river basin length, basin slope, basin elevation and basin shape, to DEM data source, DEM resolution and DEM resampling techniques. During the process of experiment, we using different resampling methods to get DEM at resolutions from 40 m to 100 m.Three data sources are used, ASTER 30 m、SRTM 90 m and GMTED 2010 250 m. The extraction of hydrological basin features from mountains, hills, plains terrain are tested. Results showed that, DEM resolution is the most sensitive factor compared to data sources and resampling technique for basin characteristics. The trend of basin characteristics for mountainous terrain is obvious, the highest resolution is the best resolution, SRTM is suitable for the basin of mountainous terrain for data sources, nearest method is the best resampling technique. The evaluation results of hills and plains are random.