这个工作是在我的博士后申申博士（2023年出站）所做工作基础上的一个方法扩展，刚刚发表在《Water Resources Management》上。

这个工作的前序，是之前指导申申博士设计了一套方法，把以往基于流域模拟做管理措施情景优化时只做空间方案优化的方法，做到能考虑分阶段资金投入限制、管理措施随时间的效益变化等因素，这样把某一个选定的流域管理措施空间方案进一步优化其分阶段的最优实施步骤（文章2023年发表在《Water Resources Research》上，见博文“流域精细治理决策支持——从情景到路线图”的介绍）。

之后我的博士生吴彤接手，进一步设计方法，把前面先空间优化、再实施次序优化的两阶段过程整合，从最开始对管理措施情景（空间方案）优化的过程中就同时考虑实施次序是否同时可最优，这样可避免某些空间情景可能很理想（甚至最优）、但实施次序怎么都难以满足分阶段资金投入限制等时变因素的问题。

这个思路其实我在申申做的时候就已经想好了，限于当时的实现进度，分成了两阶段的工作来做。后半截这个方法工作由课题组里朱良君副研究员具体指导直博生吴彤做方法设计细节实现和文章修改，完成也很久了，但在投稿过程中很不顺，不幸遇到了第一轮审稿阶段长达一年多的绝望情景（！不确定是找不到愿意审的审稿人还是编辑部内部调整切换耽误了）、进退两难，还好修改后接收很快，终于发表出来。

Wu T, Zhu L-J*, Shen S, Qin C-Z, Zhu A-X. Spatiotemporal optimization method for watershed management practice scenarios within a simulation–optimization framework. Water Resources Management, 2026, 40: 281.

Abstract: 

Optimizing watershed best management practice (BMP) scenarios is crucial for ecosystem protection and sustainable development. Existing studies have focused primarily on the spatial optimization of BMPs. Although recent research has attempted to optimize BMP implementation timing under predefined spatial arrangements, a comprehensive method that jointly optimizes BMP location and implementation time remains lacking. In this study, a flexible BMP spatiotemporal optimization method is proposed that combines alternative spatial and temporal configuration strategies to generate different BMP spatiotemporal configuration strategies. As a proof of concept, a straightforward spatiotemporal configuration strategy was implemented based on the Spatially Explicit Integrated Modeling System and the Nondominated Sorting Genetic Algorithm II. This strategy utilized spatial relationships between BMPs and slope position units and assigned BMP implementation years through a random temporal configuration process. Three optimization tasks aimed at reducing hillslope soil erosion were conducted in an agricultural watershed, including spatial-only optimization, temporal-only optimization on the basis of three selected BMP spatial arrangements, and joint spatiotemporal optimization. The results demonstrated that spatiotemporal optimization produces more practical and cost-effective BMP implementation plans. Specifically, under a moderate overall investment level, the proposed method reduced the initial investment by approximately 40%, with only a 1% loss in environmental effectiveness. This study provides a flexible basis for application in a wider range of watershed management contexts, where additional BMP spatial and temporal configuration strategies can be developed from local management expertise and incorporated into the optimization process.