Systems analysis and operations research have long been the tools to study the spatial and temporal relationships in forest resource management (as reviewed by Bettinger & Chung, 2004; Minas et al., 2012; Thompson et al., 2019; Toth, 2020). Those models are capable of reporting the tradeoffs between alternative management plans or policy choices. They also suggest optimal or suboptimal solutions as references for decision makers. This special issue includes six referred papers by authors who have been associated with the symposiums on Systems Analysis in Forest Resources. Those papers represent some recent model developments in solving challenging forest management problems in timber production, wildlife habitat conservation, forest supply chain optimization, forest insect control, and wildland fire suppression. Papers in this special issue also represent certain exciting innovations and improvements in implementing mixed integer programming, dynamic programming, agent‐based simulation, multiobjective programming, stochastic programming, and metaheuristic searching algorithms to solve new forestry problems.

Forest decisions often require managers to consider spatial relationships between management units (e.g., forest stands, polygons, or raster cells, etc.). Modeling spatial relationships has long been a major challenge (Llorente et al., 2017) in developing state‐of‐the‐art forest systems models. Research efforts along this direction have also led to many innovations in model designs and solution methods. Several papers in this special issue addressed certain unique challenges in formulating and solving spatially explicit models. For example, Bushaj et al. (2020) modeled distance‐dependent emerald ash borer spread probabilities for urban forest; Henderson et al. modeled forest core area to preserve the Kirtland's warbler habitat; Scholz et al. (2020) tracked wood fuel travel distances based on a transportation network; Wei et al. (2020) studied how fireline holding and breaching would impact wildland fire spread; Yemshanov et al. (2020) modeled landscape forest connectivity to preserve woodland caribou habitats.

Uncertainties complicate the decision‐making of forest management (as reviewed by Pasalodos‐Tato et al., 2013; Thompson et al., 2019). Random scenarios were often built to reflect system uncertainties and their impact to forest management decisions. In the special issue, Bushaj et al. (2020) built a scenario tree to model both the default and the dynamically updated emerald ash borer infestation probabilities; Scholz et al. (2020) used an agent‐based model to simulate the behaviors of forest enterprises, heating plants, and trader under positive or negative market scenarios; Wei et al. (2020) compared different contingency fire containment plans through randomly generated fireline breaching scenarios.

Forest management often needs to simultaneously achieve multiple objectives. In this special issue, She et al. (2020) constructed efficient frontiers to demonstrate the tradeoffs between the two objectives of maximizing net revenues and saving greenhouse gas emission from the salvage harvest and utilization of beetle killed forest. Yemshanov et al. (2020) and Henderson and Hoganson (2020) both designed their models to achieve the timber harvesting and habitat conservation objectives.

Systems analysis models can be difficult to solve. All six papers in this special issue reported model solution time. Some papers also suggested methods to improve the solution time of those models. For example, Henderson and Hoganson (2020) extended the decomposition algorithms developed by Hoganson and Rose (1984) and Hoganson and Borges (1998) to progressively accepting intermediate dynamic programming model solutions when solving a large forest planning problem. She et al. (2020) designed a parameter adjustment mechanism to speed up the meta‐heuristic searching process.

I wish to thank the Editor‐in‐Chief, Dr. Shandelle Henson, for the opportunity to organize this special issue, and the coeditor of this special issue, Dr. Sandor Toth, to help manage the review of several papers. I would also like to thank for all the authors who contributed to this special issue and the referees who have provided rigorous reviews and constructive suggestions for this special issue.

长期以来，系统分析和运筹学一直是研究森林资源管理中时空关系的工具（如Bettinger＆Chung，  2004年； Minas等人，  2012年； Thompson等人，  2019年； Toth，  2020年））。这些模型能够报告替代管理计划或政策选择之间的权衡。他们还建议最佳或次优解决方案，以供决策者参考。本期特刊包括与森林资源系统分析专题讨论会相关的六篇由作者撰写的参考论文。这些论文代表了一些最新的模型开发，可以解决木材生产，野生动植物栖息地保护，森林供应链优化，森林昆虫控制和野火扑灭等森林管理难题。本期特刊中的论文还代表了在实现混合整数编程，动态编程，基于智能体的仿真，多目标编程，随机编程，

森林决策通常需要管理者考虑管理单元之间的空间关系（例如林分，多边形或栅格像元等）。在开发最新的森林系统模型方面，长期以来，对空间关系进行建模一直是一个重大挑战（Llorente等，  2017）。沿着这个方向的研究工作也导致了模型设计和求解方法的许多创新。该期特刊中的几篇论文解决了在制定和求解空间显式模型时遇到的某些独特挑战。例如，Bushaj等。（2020）对城市森林中与距离有关的翡翠灰bore虫传播概率进行了建模；亨德森等。模拟森林核心区域，以保护柯特兰莺的栖息地；Scholz等。（2020年）根据运输网络跟踪木质燃料的行驶距离；Wei等。（2020）研究了保持和突破火线将如何影响野火蔓延。Yemshanov等。（2020）对景观森林的连通性进行了建模，以保护林地驯鹿的栖息地。

不确定性使森林经营的决策复杂化（如Pasalodos-Tato等人，2013； Thompson等人，  2019所审查 ）。通常建立随机情景以反映系统不确定性及其对森林经营决策的影响。在特刊中，Bushaj等人。（2020）建立了一个情景树，对默认和动态更新的祖母灰bore虫侵袭概率进行建模。Scholz等。（2020）使用基于代理的模型来模拟森林企业，供热厂和贸易商在积极或消极市场情况下的行为；Wei等。（2020年）通过随机生成的违反火线的情景比较了不同的应急消防遏制计划。

森林管理通常需要同时实现多个目标。在本期特刊中，She等人。（2020）构建了有效的边界，以证明两个目标之间的权衡，这两个目标是最大化净收入和从打捞收获和利用甲虫杀死的森林中节省温室气体排放。Yemshanov等。（2020）以及Henderson和Hoganson（2020）都设计了他们的模型来实现木材采伐和栖息地保护的目标。

系统分析模型可能很难解决。本期特刊中的所有六篇论文均报告了模型求解时间。一些论文还提出了改善这些模型求解时间的方法。例如，Henderson和Hoganson（2020）扩展了Hoganson and Rose（1984）和Hoganson and Borges（1998）开发的分解算法，以便在解决大型森林规划问题时逐渐接受中间动态规划模型解决方案。她等。（2020）设计了一个参数调整机制，以加快元启发式搜索过程。

我要感谢主编Shandelle Henson博士有机会组织这次特刊，并感谢本特刊的编辑Sandor Toth协助管理多篇论文的审阅。我还要感谢为这个特刊做出贡献的所有作者以及为这个特刊提供了严格的评论和建设性建议的裁判。