中国农业科学 ›› 2013, Vol. 46 ›› Issue (15): 3266-3276.doi: 10.3864/j.issn.0578-1752.2013.15.022

• 农业经济与管理 • 上一篇    下一篇

基于农户行为的农作物空间格局变化模拟模型架构

余强毅, 吴文斌, 唐华俊, 杨鹏, 李正国, 夏天, 刘珍环, 周清波   

  1. 中国农业科学院农业资源与农业区划研究所/农业部农业信息技术重点实验室,北京 100081
  • 收稿日期:2012-09-20 出版日期:2013-08-01 发布日期:2012-11-27
  • 通讯作者: 通信作者唐华俊,E-mail:tanghuajun@caas.cn
  • 作者简介:余强毅,E-mail:yuqiangyi@163.com
  • 基金资助:

    国家自然科学基金项目(40930101,41271112)、全球变化研究国家重大科学研究计划项目(2010CB951504)、国际科技合作项目(2010DFB10030))、农业部农业科研杰出人才基金项目

An Agent-Based Model for Simulating Crop Pattern Dynamics at a Regional Scale: Model Framework

YU Qiang-Yi, WU Wen-Bin, TANG Hua-Jun, YANG Peng, LI Zheng-Guo, XIA Tian, LIU Zhen-Huan, ZHOU Qing-Bo   

  1. Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences/Key Laboratory of Agri-Informatics, Ministry of Agriculture, Beijing 100081
  • Received:2012-09-20 Online:2013-08-01 Published:2012-11-27

摘要: 【目的】农作物空间格局是农业土地系统中景观功能层面的核心特征之一,农作物空间格局变化较土地利用与覆被变化更为频繁,但并不易于监测与分析。本文拟提出一个基于农户(Agent)作物选择行为的农作物空间格局动态变化模拟模型(CroPaDy)的构建方案。【方法】参考Agent农业土地变化模型的建模思路,按照 “ODD标准化建模程序”和“一般性的模型计算化设计框架”进行模型的概念化设计与计算化设计。【结果】概念化设计方面,将模型设计成一个闭合的环路(驱动因素分析—决策过程分析—行为结果分析)。其中,驱动因素分析影响农户决策的内/外部因素,而非传统的自然/社会经济因素。模型模拟步长为一年,状态变量每年更新,并将行为结果作为反馈从而更新内/外部影响因素,由此体现农作物空间格局变化的动态过程。计算化设计方面,将模型设计成3个既相互独立又紧密联系的子模块,分别是:Agent生成模块、Agent简化与分类模块以及Agent决策分析模块。模型计算化设计过程应用蒙特卡洛以及效用函数等常见方法,此外,尝试使用因子分析方法对Agent进行分类与简化。【结论】经检验,CroPaDy模型充分考虑了自然环境与人类社会的交互作用,能够用以表达特定区域内的农作物空间格局及其动态变化过程。

关键词: 基于主体的模拟 , 农作物选择决策 , 农作物空间格局 , 模型框架设计

Abstract: 【Objective】 Crop pattern is a key element in agricultural land systems other than land use and land cover. Crop pattern dynamics take place very frequently, but they are not always easily observable, making many difficulties for analysis. In this paper the authors are trying to conceptualize an agent-based model to simulate crop pattern dynamics at a regional scale (CroPaDy). 【Method】Both of the conceptual model and the computational model of CroPaDy are designed strictly following the ODD Protocol proposed by Grimm et al. (2010) and the Generalized Framework for Parameterization of ABM proposed by Smajgl et al., (2011). 【Result】 The conceptual model of CroPaDy is designed as a closed-loop comprised by driving forces, decision making processes, and consequences. In driving force analysis, the authors focus on the internal and external factors that influencing farmer’s decision-making instead of the macro level biophysical–socioeconomic drivers for land use and land cover change. The state variables are set to be updated annually, incorporating feedbacks in any decision-making circle. The computational model links three sub-models named agents generating module, agent simplifying and classifying module, and agent decision-making module, respectively. Common methods including Monte Carlo and Utility Function are used in model parameterization. In addition, factor analysis is applied for replacing cluster analysis in forming farmer typologies. 【Conclusion】 The authors conceptualize the framework of CroPaDy model to present the interactions between human actors and their environment in agricultural land systems. Crop pattern dynamics, therefore, can be modeled by capturing farmer’s crop choice. However, due to the limited page space, model validation and the scenarios-based application are subjected to a series of companion papers that are about to be submitted independently.

Key words: agent-based modeling , crop choice , crop pattern , model framework