Modeling complex systems
指出线性模型不足以描述生态系统，需采用复杂非线性动力系统建模，并考虑经济主体战略互动、非线性反馈、时空尺度等特征，否则政策设计可能出错。
Abstract Empirical observations suggest that linear dynamics are not an adequate representation of ecological systems and that a realistic representation would require adoption of complex nonlinear dynamical systems with characteristics encountered in complex adaptive systems (CAS). Adequate modeling should include and combine elements, such as strategic interactions among economic agents, nonconvexities induced by nonlinear feedbacks, separate spatial and temporal scales and modeling of spatiotemporal dynamics, and allowance of alternative time scales. Ignoring these characteristics might obscure very important features that we observe in reality, such as bifurcations and irreversibilities or hysteresis. As a consequence, the design of policies that do not take CAS characteristics into account might lead to erroneous results and undesirable states of managed economic–ecological systems.