摘要
土地利用变化数据,主要由不同时点的土地利用数据通过统计、对比、空间叠置分析等途径派生而来;而不同时点的土地利用数据主要来源于遥感数据、土地调查数据、历史图件数据以及统计数据。由于这些数据来源本身不可避免地带有不同程度的误差,由此派生的变化数据也就具有不同的精度。现有的文献中通常会给出时点数据的误差,但很少关注土地利用变化数据的误差。本文首先建立变化数据的误差分析方法,然后举例分析了土地利用类型变化幅度、时点数据误差及其不同组合对变化数据误差的影响。最后得出如下结论:时点数据的高精度并不能完全保证变化数据的高精度,还要看误差的方向是否一致;土地利用变化幅度越小,时点数据的误差对变化数据的精度影响越大,精度水平越难保证。这些结论提醒我们,土地利用变化研究中要对数据误差问题给以足够的重视,并应从数据源质量、数据获取方式、数据分析方法等方面提高数据的精度水平。
Land use change data are mainly derived from land use data of different periods through statistical analysis or spatial analysis in GIS environment, and those land use data usually came from remote sensing images, land survey, historical maps or regional statistical information. Because of the errors in those land use data, land use change data are in dissimilar precision, but the available papers on land use change have not paid more attention to the point, but gave some descriptions of the error of land use data. Aimed at the error characteristics of land use change data, error analysis method was firstly discussed in this paper, and then the impacts from influencing factors, including the percentage of land use change and the error of periodical land use data, were analyzed in detail to a certain extent. The result showed that high precision of land use change data cannot always be guaranteed by high precision of land use data, which also depends on the error direction and the percentage of changed area. The results imply that the error of land use change data have been ignored at present and should be improved in the process of land classification, image collection, processing, interpretation and analysis.
出处
《地理学报》
EI
CSCD
北大核心
2004年第4期615-620,共6页
Acta Geographica Sinica
基金
国家自然科学基金资助项目(40271008)~~
