Collaborative Research: Geologic Methane Emissions to the Atmosphere--Improving the Bottom-Up Estimates of Microseepage

甲烷(CH4)是一种强大的温室气体，会导致大气变暖和气候变化，并具有重大的社会影响。然而，其众多来源的相对大小仍然不受限制。在本研究中，收集了从自然地质特征排放的甲烷的实地观测数据，这些甲烷是大气中最不确定的甲烷来源之一。然后，机器学习技术被用来提高美国各地对这些排放的预测能力。本科生和研究生是野外采样和数据分析工作的组成部分，教育推广活动旨在通过与当地学区和罗切斯特博物馆科学中心建立的项目来进行。微渗漏是自然地质CH4进入大气的低强度通量，被认为是地质排放的单一最大但限制较差的成分。这项研究试图改进对大面积富含碳氢化合物的沉积盆地的微渗漏的自下而上的估计，总体假设是现有库存中的微渗漏排放量被高估了。在夏季和冬季测定了美国8个不同盆地的CH4通量，以覆盖气藏深度、构造/构造和气候/生态系统不同特征的大范围区域。这些测量使用了现场可展开的闭合动态磁通室技术。然后研究了微渗透通量大小与一组假设的通量增强参数之间的关系，并利用该关系开发了一种提高微渗透通量的机器学习方法。由此产生的对邻近美国微渗漏CH4排放的新估计将提高我们对这种重要温室气体的自然排放的理解，这种温室气体也是大气化学和全球气候的关键参与者。该奖项反映了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。

Methane (CH4) is a powerful greenhouse gas that contributes to warming of the atmosphere and climate change and has significant societal implications. Nevertheless, relative magnitudes of its numerous sources remain ill-constrained. In the present study, field observations are collected of CH4 that is emitted from natural geological features and represent one of the most uncertain sources of CH4 to the atmosphere. Machine learning techniques are then used to increase predictive power of these emissions across the United States. Undergraduate and graduate students are an integral part of the field sampling and data analysis components of the work, and educational outreach activities are intended through established programs with the local school district and Rochester Museum & Science Center.Microseepage is a low-intensity flux of natural geological CH4 to the atmosphere that is postulated to be the single largest, yet poorly constrained component of geological emissions. The research seeks to improve bottom-up estimates of microseepage from large areas of hydrocarbon-rich sedimentary basins, with the overarching hypothesis that microseepage emissions in existing inventories are overestimated. CH4 fluxes are determined during summer and winter in eight different US basins to cover a wide area of varying characteristics in gas reservoir depth, structure/tectonics, and climate/ecosystem. A field deployable closed dynamic flux chamber technique is used for these measurements. The relationship between microseepage flux magnitude and a set of hypothesized flux-enhancing parameters is then investigated and used to develop a machine-learning approach for upscaling microseepage fluxes. Resulting new estimates of microseepage CH4 emissions from the contiguous US will improve our understanding of natural emissions of this important greenhouse gas that is also a key player in atmospheric chemistry and global climate.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.