地理科学进展 ›› 2010, Vol. 29 ›› Issue (11): 1408-1413.doi: 10.11820/dlkxjz.2010.11.034

• 自然区划与土地利用 • 上一篇    下一篇

土地变化驱动力研究的尺度问题

严祥, 蔡运龙, 陈睿山, 李昊   

  1. 北京大学城市与环境学院地表过程分析与模拟教育部重点实验室, 北京100871
  • 收稿日期:2010-03-01 修回日期:2010-05-01 出版日期:2010-11-25 发布日期:2010-11-25
  • 作者简介:严祥(1987-)|男|四川绵阳人|硕士|主要从事土地变化、区域发展等研究
  • 基金资助:

    国家自然科学基金项目(40871047)。

Scale Issues in the Driving Force Analysis of Land Change

YAN Xiang, CAI Yunlong, CHEN Ruishan, LI Hao   

  1. College of Urban and Environmental Sciences, Peking University;
    Key Laboratory of Analysis and Simulation ofEarth Surface Processes, Ministry of Education, Beijing 100871, China
  • Received:2010-03-01 Revised:2010-05-01 Online:2010-11-25 Published:2010-11-25

摘要:

尺度问题是土地变化驱动力研究中的热点和难点。统计方法在土地变化驱动力研究中应用广泛,而目前对此类研究所需考虑的尺度问题尚未有针对性的论述。本文通过总结目前土地变化驱动力分析中主要的统计方法,识别出其中亟待处理的5 种尺度问题,即研究区空间幅度、空间粒度、时间幅度、时间粒度和土地分类精度5 个因素变化对分析结果的影响。并且认为此5 种尺度问题的本质都是数据在空间或时间上聚合对统计分析结果的影响,若缺乏妥善处理,土地变化驱动力分析的结果将表现出不真实性,可能带来2 方面的风险:一是土地变化驱动力分析结果无意义;二是引导出错误的对策。土地变化驱动力研究中有3 种思路可处理该问题:①研究基本实体的明确界定;②敏感性分析或变化速率分析;③寻找回避尺度问题的恰当方法。3种思路均可借鉴,但仍有待进一步探讨。

关键词: 尺度问题, 驱动力, 统计分析, 土地变化

Abstract:

Scale issue is a focus but also a difficult theme in the driving forces analysis of land change. While statistics is very popular in analyzing drivers of land change, there is no particular research on the impact of scale in this domain. This article summarizes the methods often used in the driving force analysis of land change, recognizes five scale issues needed to concern. They are the impact of 5 situations (the extent of the study area, the grain size, the time range, the time period of data, and the classification accuracy) on the analysis results when using the term of‘scale’, and all of these 5 situations can be concluded into data aggregation. If the impact of data aggregation is not treated correctly, the potential dangers exist in two aspects: (1) the results is meaningless, (2) the policies based on the analysis are incorrect and harmful. So far, there are 3 approaches to solve this problem: (1) basic entity approach, (2) sensitivity analysis approach, and (3) frame-independent analysis approach. These approaches can be applied and more researches are needed in the domain of the driving force analysis of land change.

Key words: driver, land change, scale, statistics