eMB: Collaborative Research: Mechanistic models for seasonal avian migration: Analysis, numerical methods, and data analytics

迁徙动物，包括许多鸟类，每年在它们的夏季繁殖地和越冬地之间旅行一段惊人的距离，以跟踪关键资源的出现，并最大限度地实现繁殖成功。然而，由于这些资源的时间受到全球气候变化的影响，这些迁徙周期可能会被打乱。为了解决上述知识差距，由数学家、生物学家和数据科学家组成的跨学科原理调查小组将开发新的数学模型，如随机动态规划和基于主体的模型，以从机械上理解迁徙动物如何响应环境线索确定它们的路线和时间，以及这些如何受到气候变化的影响。该项目将促进eBird数据库的使用，这是一项草根努力，使公民能够使用移动应用程序直接贡献鸟类观测数据。该项目是俄亥俄州立大学和俄克拉何马州立大学的合作项目，提供了宝贵的教育、培训和拓展机会。除了对博士生的培训外，PIS还建议建立一个K-12项目，包括一个为期一天的研讨会，以让公众参与并提高对鸟类保护工作和气候变化影响的认识，以及一个针对代表性不足群体的夏季本科生研究项目(俄亥俄州立大学的Romus项目)。在专业层面，专业人员将举办协同活动，以促进数学家和生物学家之间的交流。其中包括2024年10月在班夫国际研究站为100多名虚拟和面对面参与者举办的为期一周的研讨会。这项研究将提高我们对气候变化对候鸟种群影响的理解，并为未来野生动物的保护和管理提供信息。对于理解生态对全球变化的反应和在不断变化的世界中保护生物多样性来说，一个关键的问题是，迁徙的动物是否可以调整它们的迁徙路线和时间表，以跟踪关键资源，即使这些资源的物候随着气候变化而变化。就季节性迁徙路线穿越半球的鸟类而言，过去的研究发现，春季植被复苏与它们到达繁殖区之间存在不同步。这引发了人们的担忧，即候鸟可能会受到气候变化的负面影响。这个由数学家、生物学家和数据科学家组成的跨学科团队将通过开发用于迁徙动物种群的随机动态规划(SDP)模型和基于代理的模型(ABM)来解决上述知识差距。首先，在最优迁移问题的背景下，建立了一个在连续时间框架下具有切换成本的SDP模型。PIS将制定和分析由此产生的贝尔曼方程，并解决将个人迁徙决策与新出现的人口模式联系起来的理论挑战。此外，投资促进机构还将使用空间上明确的基于代理人的建模，与拟议的数学建模并行并加以扩展。最后，PIS将利用西半球春季复苏的遥感数据和eBird数据库中鸟类种群水平的迁徙数据，进行参数估计和模型结果比较。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Migratory animals including many bird species travel a spectacular distance annually between their summer breeding ground and overwintering site to track the appearance of key resources and maximize reproductive success. However, these migratory cycles are potentially disrupted as the timing of these resources are impacted by global climate change. To address the above knowledge gap, the interdisciplinary team of principle investigators (PIs) comprising a mathematician, biologist, and data scientists will develop novel mathematical models such as stochastic dynamic programming and agent-based models to gain mechanistic understanding of how migratory animals determine their routes and timing in response to environmental cues, and how those can be affected by climate change. The project will promote the use of the eBird database, which is a grassroots effort enabling citizens to directly contribute bird observations data using a mobile app. The project is a collaboration between Ohio State University and Oklahoma State University and offers valuable educational, training, and outreach opportunities. In addition to training of PhD students, the PIs propose to establish a K-12 program with a daylong workshop to engage the public and raise awareness of bird conservational efforts and impacts of climate change, as well as a summer undergraduate research program targeting underrepresented groups (ROMUS program at Ohio State). At the professional level, the PIs will organize synergistic activities to facilitate exchanges between mathematicians and biologists. These include a confirmed week-long workshop at Banff International Research Station in October 2024, for over 100 virtual and in-person participants. This research will improve our understanding of the effects of climate change on migrating bird populations and inform future conservation and management of wildlife. A critical question for understanding ecological responses to global change and conserving biodiversity in a changing world, is whether migrating animals can adjust their migration routes and schedules to track key resources even as the phenology of these resources shifts with climate change. In the case of birds with seasonal migratory routes crossing hemispheres, past studies have detected asynchrony between spring vegetation green-up and their arrival at breeding areas. This raised the concern that migrating birds may be negatively impacted by climate change. This interdisciplinary team of mathematician, biologist and data scientists will address the above knowledge gap by developing stochastic dynamic programming (SDP) models and agent-based models (ABM) for migrating animal populations. First, the aim is to develop an SDP model with switching costs in a continuous-time framework, in the context of optimal migration problem. The PIs will formulate and analyze the resulting Bellman equations and address the theoretical challenge of connecting individual migration decision to emergent population patterning. In addition, the PIs will also use spatially explicit agent-based modeling in parallel with and to extend the proposed mathematical modeling. Finally, the PIS will leverage remote sensing data on spring green-up across the Western Hemisphere with population level migration data for bird species from eBird database, for parameter estimation and model result comparison.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.