地形起伏度最佳统计单元算法的比较研究

Comparison Study on the Best Statistical Unit Algorithms of Relief Amplitude

基于90m分辨率SRTM DEM数据,利用邻域分析法在不同大小窗口下(2×2,3×3,4×4,…,30×30)对山西省地形起伏度进行了提取,运用人工作图法、最大高差法、模糊数学法、均值变点分析法和累积和(CUSUM)分析算法分别计算最佳统计单元,对统计单元计算方法的准确性及适用性进行分析,提出山西省地形起伏度提取的最佳统计单元。研究分析表明:均值变点分析法和累积和分析算法是相对比较有效的方法;运用均值变点分析法与累积和分析算法计算所得的最佳统计单元分别为11×11(0.980 1km^2)和14×14(1.587 6km^2);均值变点分析法是计算最佳统计单元的一种最为理想的方法;对于同一地区而言,不同DEM数据类型、不同分辨率DEM,所使用的最佳统计单元大小也不同。

Based on the 90-m resolution SRTM DEM data,the relief amplitude of Shanxi Province was extracted with the neighborhood analysis method at different neighborhood windows(2×2,3×3,4×4,…,30×30),the best statistical unit was calculated by using the artificial drawing method,maximum elevation difference method,fuzzy mathematics method,mean change-point analysis method and cumulative sum analysis method,the accuracy and applicability of statistical unit calculation methods was analyzed,and the best statistical unit of relief amplitude in Shanxi Province was determined.The results indicate that the mean change-point analysis method and cumulative sum analysis method are more effective than the artificial drawing method,maximum elevation difference method and fuzzy mathematics method,the best statistical units of the mean change-point analysis method and cumulative sum analysis method are 11×11(0.980 1 km^2)and 14×14(1.587 6 km^2),respectively.The mean change-point analysis method is one of the best ideal methods to calculate the best statistical unit.The best statistical unit is different between different DEM data types and different resolution DEM for the same area.