栅格化属性精度损失的评估方法及其尺度效应分析——以四川省1:25万土地覆被数据为例

Evaluating Methods and Scale Effects of Attribute Information Loss in Rasterization: A Case Study of 1:250 000 Land Cover Data of Sichuan

选择在600m^30km16个尺度上,在ArcGIS中利用常用的面积最大值法(Rule of Maximum Area,RMA)对2005年四川省1:25万土地覆被矢量数据进行栅格化,并采用两种属性精度损失评估方法:传统的常规分析方法和一种新的基于栅格单元分析方法,来对比分析在这两种评估方法下RMA栅格化的属性(这里是指面积)精度损失随尺度的变化特征。结果表明:(1)在同一尺度下采用基于栅格单元方法分析所得的研究区平均属性精度损失大于常规分析方法分析得到的平均属性精度损失,且二者之间的差异在1~10km内很明显,当栅格单元大于10km时,两种方法得到的平均属性精度损失的差值稳定,且其随尺度的变化曲线趋于平行;(2)基于栅格单元分析方法不仅能够准确地定量估计RMA栅格化的属性精度损失,而且能客观地反映属性精度损失的空间分布规律;(3)对四川省1:25万土地覆被数据进行面积最大值法(RMA)栅格化的适宜尺度域最好不要超过800m,在该尺度域内数据工作量适宜,且RMA栅格化属性精度损失小于2.5%。

Taking land cover data of Sichuan at the scale of 1：250,000 in 2005 as a case study, this paper selects 16 spatial scales from 600 m to 30 km, uses rasterizing method based on the Rule of Maximum Area （RMA） in ArcGIS and two evaluation methods of attribute accuracy loss, which are Normal Analysis Method （NAM） and a new Method Based on Grid Cell （MBGC）, respectively, and analyzes the scale effect of attribute （area） accuracy loss at 16 different scales by the two evaluating methods comparatively. The results are shown as follows. （1） At the same scale, mean attribute accuracy loss computed by Method Based on Grid Cell are significantly bigger than that computed using Normal Analysis Method. Moreover, this discrepancy is ranged between 1 km and 10 km, while the grid cell is larger than 10 km. Mean attribute accuracy loss computed by these two evaluation methods is stable, even tended to be parallel. （2） Method Based on Grid Cell can not only estimate RMA rasterization attribute accuracy loss accurately, but can express the spatial distribution of rasterization attribute loss objectively. （3） The suitable scale domain for RMA rasterization of land cover data of Sichuan at the scale of 1： 250,000 in 2005 is better, equal or less than 800 m, in which the data volume is favorable, and the attribute accuracy loss is less than 2.5%.