Interdisciplinary Science Session: Can Machine Learning and Data-driven Science Lead to Breakthroughs in Earth System Modeling and Analysis?; Aspen, Colorado; June 7-11, 2021

跨学科科学会议：机器学习和数据驱动科学能否带来地球系统建模和分析的突破？；

• 批准号: 2038111
• 负责人: James Arnott
• 金额: $ 1.8万
• 依托单位: Aspen Global Change Institute
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-08-01 至 2022-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2038111&HistoricalAwards=false
• 关键词: Interdisciplinary; Science; Session; Machine; Learning

Recent years have seen an explosion in the use of machine learning (ML) and other data science techniques in applications from self-driving cars to targeted advertising to medical diagnosis. Meanwhile climate and earth system science have become increasingly data intensive, as the volume of data from observing systems and computer models has increased almost exponentially. There is thus considerable interest in finding ways to apply the tools of data science to the problems of climate and earth system science.This workshop brings together a group of researchers from the fields of climate, earth system science, statistics, data science, and related disciplines to seek novel and productive ways to apply data science tools to climate and earth system science. One application to be considered is the use of ML as a means of representing small-scale processes in climate and earth system models. Another is the use of ML and similar techniques for the analysis of large volumes of data. Particular challenges to be considered include the incorporation of physical constraints into ML algorithms and the extent to which results of ML-based analysis can be interpreted in physically meaningful ways.The workshop has broader impacts through its attempt to introduce new and powerful tools into climate and earth system science. Data science tools have the potential to enhance the societal value of research results, for example by allowing scientists to provide better guidance to planners and stakeholders facing threats posed by extreme weather, climate change, and other earth system phenomena. Public outreach will be performed through a keynote lecture and web-accessible videos of workshop presentations, and a perspective paper will be published as a result of the meeting. The workshop is planned for the week of 7 June 2021.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.

近年来，机器学习 (ML) 和其他数据科学技术在自动驾驶汽车、定向广告和医疗诊断等应用中的使用呈爆炸式增长。 与此同时，气候和地球系统科学变得越来越数据密集，观测系统和计算机模型的数据量几乎呈指数级增长。 因此，人们对寻找将数据科学工具应用于气候和地球系统科学问题的方法产生了很大的兴趣。本次研讨会汇集了一组来自气候、地球系统科学、统计学、数据科学和相关领域的研究人员学科寻求新颖且富有成效的方法将数据科学工具应用于气候和地球系统科学。需要考虑的一个应用是使用机器学习作为表示气候和地球系统模型中小规模过程的手段。 另一种是使用机器学习和类似技术来分析大量数据。 需要考虑的具体挑战包括将物理约束纳入机器学习算法，以及基于机器学习的分析结果在多大程度上可以以物理上有意义的方式解释。该研讨会通过尝试将新的强大工具引入气候和气候领域，产生了更广泛的影响。地球系统科学。 数据科学工具有潜力提高研究成果的社会价值，例如，让科学家能够为面临极端天气、气候变化和其他地球系统现象威胁的规划者和利益相关者提供更好的指导。 将通过主题演讲和研讨会演示的网络视频进行公众宣传，会议结果将发表一份观点论文。 该研讨会计划于 2021 年 6 月 7 日这一周举行。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。