通过土地利用情景变化模型反映多尺度的土地利用现象,是学者们研究的焦点问题。本文基于MAS理论,构建从地块到村庄的多尺度土地利用变化模型,揭示微观层面不同尺度上土地利用变化的机制。应用该模型,以陕西省米脂县孟岔村为例进行研究,定量分析该村农户个体、农户群体和农户整体等不同尺度上的土地利用决策,并通过与2008年实际土地利用变化数据对比和分析,对构建的多尺度土地利用变化模型进行检验。通过分析,认为:(1)基于MAS的多尺度农户土地利用变化模型可较好地反映微观土地利用决策和空间的多尺度变化;(2)构建用 连续数值表示的农户土地利用决策与实际状况比较符合;(3) 本文为不同农户群体间的转化方法、农户间相互作用,以及市场与政策因素的作用等方面的研究打下坚实的基础。

The multi-scale issue has been one of the hot-spots in the land use/cover change (LUCC) research over long time, and the recent focus is to understand the micro mechanism of LUCC at multi-scale by using the land use/land cover scenario model. In this paper, we constructed a multi-scale land use/land change model from field to village based on the multi-agent system (MAS) and aimed to reveal the micro mechanism of land use/land change under multi-scale. We firstly made a short review in the context of MAS research and proposed that three aspects as constructing clear mechanism, quantitative analysis of interaction and explicating multi-scale simulation model are the key problems when considering human adaptation to the natural and social-economic changes in the MAS research. In section 2, we made the flow chart and the scale transformation mechanism figure. The correspondence between different organization levels and different spatial scales were also given. Based on the prior conception framework, section 3 deals with the transforming architecture of the multi-scale decision-making model. The transforming methods from the field scale to the village scale were given according to the transforming architecture. A case study for Mengcha Village of Mizhi County of Shaanxi Province was explored by using the proposed multi-scale model in section 4, and the model was tested by the LUCC data of 2008. In the last section, we made a discussion on the potentials and drawbacks of the approach. The conclusions can be drawn as follows: (1) The multi-scale LUCC model constructed in the study can reflect the micro land-use decisions and multi-scale space transformation well. (2) The households land use decision presented by continuous numerical value was in consistent with the actual situation. (3) The approach suggested in this study laid a foundation for the future research on the interaction and transformation among different types of households, and was also useful to studies of the impacts of the market and policy on the household land use decision-making. (4) Larger spatio-temporal scale land use decision-making model is needed to bridge the LUCC data and the short-term household decision-making efficiently.