基于生态系统服务权衡特征的流域土地利用多目标优化方法

复杂的生态系统服务权衡特征是生态系统管理实现多赢目标的瓶颈，尽管围绕生态系统服务权衡类型、特征与协调措施开展了大量研究，但如何通过有效的土地利用管理实现协同增益，仍缺乏定量优化方法。本项目拟基于生态系统服务权衡特征及驱动因素，耦合土地利用优化模型与生态系统服务定量模型，结合机器学习反复识别优化目标的优势，构建多目标函数和综合效益最大化的土地利用优化方法。并以生态系统服务权衡特征突出的洞庭湖流域为对象，采用上述方法，并结合野外观测、情景分析等手段，开展粮食生产、土壤保持、水源涵养、水质净化和洪水调蓄等多种生态系统服务协同提升实证与方法完善研究。最终，建立基于生态系统服务权衡协调目标、土地利用分配优化途径和多种生态系统服务协同增益的土地利用多目标优化方法，为区域土地利用优化提供定量手段。

Complex ecosystem service trade-offs are the bottleneck of achieving win-win objectives in ecosystem management. Although a lot of researches have been carried out on the types, characteristics and coordination measures of ecosystem service trade-offs, how to achieve synergistic gain through effective land use management still lacks quantitative optimization methods. Based on the ecosystem service trade-off characteristics and driving factors, this study intended to build a multi-objective function and propose a land use optimization method to maximize comprehensive benefits by coupling the land use optimization model with the quantitative model of ecosystem services, taking advantage of machine learning which identifies optimization objectives repeatedly. Taking the Dongting Lake Basin as an example, which has prominent trade-offs between ecosystem services, we carried out empirical studies on synergistic improvement of multiple ecosystem services ( such as food production, soil retention, water retention, water purification and flood regulation ) using the above-mentioned methods combined with field observation and scenario analysis to improve our optimization method. We finally established a multi-objective land use optimization method based on the objectives of balancing ecosystem service trade-offs, the way of optimizing land use allocation and the synergistic gain of multiple ecosystem services to provide quantitative means for regional land use optimization.