中国山地起伏度计算中地形自适应滑动窗口获取与验证

发布时间：2018-03-13 09:43

本文选题：起伏度　切入点：地形自适应　出处：《地理与地理信息科学》2017年04期 　论文类型：期刊论文

【摘要】：在地形起伏度研究中,多采用固定大小的"最佳计算窗口"进行运算,尚未见随地形特征变化的自适应窗口。针对单值最优窗口固有的不确定性,该文尝试构建适应局部地形的窗口阵列。首先对均值变点法作拓展与调整,遍历栅格单元,并求每一栅格处最佳窗口取值,从而得到地形自适应的滑动窗口。最后,使用SRTM和ASTR GDEM在山地范围内进行实验与检验,结果表明:最佳计算窗口的尺寸随地形发生变化;与单值窗口相比,基于地形自适应窗口求得的起伏度与地表实际情况更为吻合、细节特征更加清晰、精度更高;地形自适应窗口较好地改善了单值计算窗口对高程数据源敏感的情况。中国山地自适应滑动窗口以3.57km2、4.28km2、5.06km2为主,其统计特征符合正态分布,总体均值可与全局单值窗口相互印证。
[Abstract]:In the study of terrain relief, the fixed size of the "optimal calculation window" is used to operate, but no adaptive window with the change of terrain features has been found. The inherent uncertainty of the single value optimal window is discussed. This paper attempts to construct a window array adapted to local terrain. Firstly, the mean variable point method is extended and adjusted, the grid element is traversed, and the optimum window value at each grid is obtained, thus the terrain adaptive sliding window is obtained. SRTM and ASTR GDEM are used to carry out experiments and tests in the mountainous area. The results show that the size of the optimal calculation window varies with the terrain, and the relief obtained from the adaptive window is more consistent with the actual situation of the earth's surface than the single value window. The detail features are clearer and the precision is higher, and the terrain adaptive window can improve the sensitivity of the single value calculation window to the elevation data source. The main feature of the adaptive sliding window in the mountain area of China is 3.57km2 / 4.28km2 / 5.06km2, and the statistical characteristics of the window accord with the normal distribution. The total mean value can be verified with the global single value window.
【作者单位】：中国科学院水利部成都山地灾害与环境研究所数字山地与遥感应用中心;成都理工大学地球科学学院;
【基金】：国家自然科学基金项目(41631180;41571373) 国家重点研发计划(2016YFA0600103) 中国科学院、水利部成都山地灾害与环境研究所“一三五”培育方向项目;中国科学院、水利部成都山地灾害与环境研究所数字山地与遥感应用中心自主部署基金项目
【分类号】：P208;P931

【参考文献】