土地利用变化模拟研究进展

土地利用变化研究经历了近30年的快速发展,学者基于不同建模目标构建出多种土地利用变化模型,实现了从数量模拟到时空格局模拟,从单一模型向多种模型耦合的跨越。当前研究主要在元胞自动机(Cellular Automata,CA)模型和CLUE-S (Conversion of Land Use and its Effects at Small region extent)模型的基础上进行改进,马尔科夫模型、系统动力学(System Dynamics,SD)模型、Logistic回归和随机森林等均可计算CA模型和CLUE-S模型中所需的土地利用需求,多标准评价、地理加权回归、多主体模型以及人工神经网络等方法也多被用于CA模型的扩展,而CLUE-S的改进则存在模型本身系列的升级。这些模型广泛应用于各种区域和尺度土地利用变化预测实例研究并研发软件系统和数据集。驱动力分析主要从自然因素与人文因素两方面进行,人文因素是引发土地利用变化的主要因素。在目前的研究中,由于技术手段的限制,仍然存在时空尺度、数据误差、数据整合的不确定性等问题。未来土地利用变化模拟研究应进一步发挥大数据技术优势,推动土地利用变化模拟研究朝向精细化、多元化方向发展。结合生态环境领域实际问题,深挖土地利用变化与其生态环境效应之间的互馈机制,将研究视角从探究人类活动对土地利用变化的影响逐渐转向二者相互作用,最终促进人地关系协调发展。

Land use change simulation: progress, challenges, and prospects

Land use change research has achieved remarkable results through rapid development in recent three decades. Various models have been constructed based on different research objectives of land use change. These models have been developed from quantitative simulation to spatial pattern simulation, from single model to multiple models. Cellular Automata model and the CLUE-S (Conversion of Land Use and its Effect at Small region extent) model are combined with other models or methods to improve the performance. Markov Chain, System Dynamics (SD), Logistic regression, and Random Forest are all able to calculate the land use requirements in the CA (Cellular Automata) model and CLUE-S model. Multi-criteria evaluation, geographically weighted regression, multi-agent model and Artificial Neural Network are also used to optimize the CA model, and the improvement of CLUE-S includes the upgrade of the CLUE series. These combined models are widely used to the case study of land use change in various regions, and to develop some software systems and datasets. The driving force analysis is mainly carried out from two aspects:natural factors and human factors, and the latter is the main factor that causes land use changes. In the current research, there are still problems of spatiotemporal scale, data error, uncertainties in data integration. In the future, land use change simulation research should further utilize the technological advantages of the big data and promote it to become more refined and diversified. In addition, future research should combine relevant knowledge in the field of ecological environment and dig deeper into the mutual feedback mechanism between land use change and its ecological environment effects. The research perspective can shift from exploring the impact of human activities on land use changes to the interaction between land use change and human activities for promoting the coordinated development of human-land relationships.