NNA Track 2: PredictFest: To Build Capacity for Arctic Stakeholders in Need of Multi-Scale Predictions

导航新北极（NNA）是美国国家科学基金会的十大构想之一。NNA项目解决快速变化的北极地区的融合科学挑战。北极研究需要为国家、更大地区和全球的经济、安全和复原力提供信息。NNA支持从地方到国际规模的新研究伙伴关系，使下一代北极研究人员多样化，加强正式和非正式教育的努力，并在适当情况下整合知识的共同生产。该奖项通过支持具有明确潜力的规划活动，以开发新颖，领先的研究理念和方法来实现NNA目标，从而实现了该目标的一部分。它整合了自然环境和社会系统的各个方面，解决了重要的社会挑战，并与当地和土著社区合作。在提高短期天气预报和长期气候、海冰和其他环境预报的准确性方面取得了重大进展。然而，在两周到几个月的尺度上仍然存在预测差距，特别是在北极。这种分季节到季节性（S2S）的尺度对于规划自给狩猎、商业捕鱼、灾害应对和风险缓解以及其他活动非常重要。该项目通过协作设计流程和研讨会PredictFest支持改进S2S预测能力的开发，该研讨会与北极利益相关者一起确定此类预测的需求和应用。PredictFest的参与者将准备研究报告，阐明利益相关者的计划和决策需求，确定响应这些需求的技术和建模方法，确定相关的联系和应用，并制定持续发展的计划。PredictFest为北极居民和科学家提供了一个机会，共同开发海冰、降水和其他环境因素的季节性预报应用程序，从而支持海洋和沿海地区的传统和商业用途。此外，该项目促进了以团队为中心的应用研究、教育和推广方法，并建立了北极居民和科学家之间的关系，以支持未来的工作。科学家和其他北极居民，包括土著组织、社区成员、企业主和公共决策者之间的直接接触，鼓励人们对如何使用季节预报信息以及采取哪些步骤来实现既定目标进行新的思考。这项拨款支持一种创新的能力建设新方法，利用新兴的S2S平台和科学来确定研究项目，以支持北极利益相关者的多样化规划和决策需求。该项目包含三个阶段：(1)利益相关者的参与和分季节到季节性预测重点的开发；(2)开发团队的组成和PredictFest的实施；(3)对PredictFest输出的后评估和协作改进。共同学习和共同生产有助于快速识别季节尺度预测可能对支持自然、建筑和社会环境的决策最有利的领域。在PredictFest期间，每个参赛团队由自然和社会科学家、利益相关者和社区代表以及技术支持人员组成。重点主题“海洋环境和近岸海岸界面”产生了不同的预测，可用于支持以海洋哺乳动物为中心的社区生存，包括固定渔具参与者在内的商业渔业，以及灾害应对和风险缓解规划。从预测的角度来看，团队将采用许多方法。其中包括开发变量之间的新关系，通常使用线性模型或基于树的学习，使用时间序列预测技术模型，如自回归综合移动平均（ARIMA），以及模式识别。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Navigating the New Arctic (NNA) is one of NSF's 10 Big Ideas. NNA projects address convergence scientific challenges in the rapidly changing Arctic. The Arctic research is needed to inform the economy, security and resilience of the Nation, the larger region and the globe. NNA empowers new research partnerships from local to international scales, diversifies the next generation of Arctic researchers, enhances efforts in formal and informal education, and integrates the co-production of knowledge where appropriate. This award fulfills part of that aim by supporting planning activities with clear potential to develop novel, leading edge research ideas and approaches to address NNA goals. It integrates aspects of the natural environment and social systems, addresses important societal challenges, and engages with local and Indigenous communities.Significant progress has been made to improve the accuracy both of short-term weather forecasts and longer-term climate, sea-ice, and other environmental forecasts. Nevertheless, there remains a forecasting gap on the scale from two weeks to a few months, especially in the Arctic. This subseasonal-to-seasonal (S2S) scale is important for planning subsistence hunting, commercial fishing, hazard response and risk mitigation, and other activities. This project supports development of improved S2S prediction capacity through a collaborative design process and workshop, PredictFest, that engages Arctic stakeholders to identify needs and applications for such predictions. PredictFest participants will prepare research pitches that articulate stakeholder planning and decision making needs, identify technical and modelling approaches that respond to those needs, identify relevant connections and applications, and set out a plan for continued development. PredictFest provides an opportunity for Arctic residents and scientists to work together to develop applications for seasonal forecasts of sea ice, precipitation, and other environmental factors, which support traditional and commercial uses of marine and coastal regions. In addition, the project fosters a team-centered approach to applied research, education, and outreach and builds relationships between Arctic residents and scientists to support future work. Direct engagement between scientists and other Arctic residents, including Indigenous organizations, community members, business owners, and public decision makers, encourages new thinking about how seasonal forecast information may be used and what steps may be taken to realize stated goals. This grant supports an innovative new approach to capacity building that leverages emergent S2S platforms and science to identify research programs to support the diverse set of Arctic stakeholder planning and decision-making needs. The project contains three phases: (1) stakeholder engagement and development of sub-seasonal to seasonal prediction focus, 2) composition of development teams and implementation of PredictFest, and 3) post-assessment and collaborative refinement of PredictFest outputs. Co-learning and co-production contribute to the rapid identification of arenas where seasonal scale predictions might be most beneficial to decision making in support of natural, built, and social environments. During PredictFest each participating team is comprised of natural and social scientists, stakeholders and community representatives, as well as technical support staff. The topic of focus, Marine Environments and the Near-shore Coastal Interface, generates distinct predictions that can be used to support community subsistence centered on marine mammals, commercial fisheries including fixed gear participants, and hazard response and risk mitigation planning. From a prediction standpoint a number of approaches will be pursued by the teams. These include developing new relationships between variables typically using linear models or tree-based learning, using time-series forecasting techniques models such as Autoregressive Integrated Moving Average (ARIMA), and pattern recognition.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.