Collaborative Research: Exploring the dynamics of nitrous oxide in the Southern Benguela Upwelling System

一氧化二氮是一种寿命长、强度大的温室气体。它也是平流层臭氧消耗的驱动因素。一氧化二氮的自然来源包括细菌通过称为硝化和反硝化的过程产生的一氧化二氮。海洋是大气层一氧化二氮的主要来源，而东部边界上升流系统等生物生产力高的地区被证明向大气排放的一氧化二氮特别高。南本格拉上升流系统(SBUS)可以说是最具生产力的东部边界上升流系统，但该地区还没有公布一氧化二氮的测量结果。此外，一氧化二氮产生的主要生物途径在SBU中的相对重要性还没有得到很好的确定。来自康涅狄格大学和南卡罗来纳大学的一组科学家将调查SBUS中的一氧化二氮循环，估计进入大气的一氧化二氮通量，并探索是什么驱动一氧化二氮循环在不同季节和不同地点发生变化。这项研究将增进人们对一种重要的温室气体及其循环可能如何随着海洋条件的变化而改变的理解。拟议工作的资金将资助康涅狄格大学的一名研究生和南卡罗来纳大学的一名本科生的培训。它还将为博士后提供在区域建模和机器学习中应用计算技能的机会，为在学术界或工业界从事职业生涯做准备。该项目将通过将研究方法纳入康涅狄格大学的本科服务学习课程，促进本科生教育和向普通公众和政策制定者传播气候变化科学。这项工作旨在提供区域一氧化二氮进入大气的通量的估计，检查季节动态，评估一氧化二氮生产和消费的生物途径，并询问区域和大规模的一氧化二氮动态。为此，该小组将(A)在季节性调查期间测量在SBUS收集的样本中的一氧化二氮浓度，持续测量海面一氧化二氮，并根据一致的风速得出一氧化二氮从海到空气的通量的可靠估计；(B)通过测量一氧化二氮和硝酸盐的氮和氧同位素比率，查询一氧化二氮的产生和消费路径，并调查一氧化二氮循环的环境相关性；(C)根据这些测量结果建立一氧化二氮的区域机器学习模型，以调查一氧化二氮季节性和年际动态的环境驱动因素，利用历史数据和区域海洋动力模型得出的水文场预测SBU中的一氧化二氮。这项工作将深入了解东部边界上升流系统中一氧化二氮的生产和消费途径，该系统以广阔的大陆架为边界--与从活动边缘获得的更多知识形成对比。机器学习的一氧化二氮统计模型最终将用于根据SBU的气候预测预测一氧化二氮，并预测一氧化二氮对全球海洋脱氧的区域响应。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。

Nitrous oxide is a long-lived and powerful greenhouse gas. It is also a driver of ozone depletion in the stratosphere. Natural sources of nitrous oxide include production by bacteria through processes called nitrification and denitrification. The ocean is a major source of nitrous oxide to the atmosphere, and areas of high biological productivity like eastern boundary upwelling systems have been shown to generate particularly high nitrous oxide emissions to the atmosphere. The Southern Benguela Upwelling System (SBUS) is arguably the most productive eastern boundary upwelling system, yet there are no published measurements of nitrous oxide from this region. In addition, the relative importance of the major biological pathways of nitrous oxide production in the SBUS is not well established. A team of scientists from the University of Connecticut and the University of South Carolina will investigate nitrous oxide cycling in the SBUS, estimate nitrous oxide fluxes to the atmosphere, and explore what drives changes in nitrous oxide cycling across seasons and in different locations. This study will improve understanding of an important greenhouse gas and how its cycling might change in response to changing ocean conditions. Funding from the proposed work will sponsor the training of a graduate student at the University of Connecticut and an undergraduate student at the University of South Carolina. It will also provide a postdoctoral fellow the opportunity to apply computational skills in regional modeling and machine learning, offering preparation for a career in academia or industry. The project will contribute to undergraduate education and communication of climate change science to the general public and policymakers through the incorporation of the study methodology into an undergraduate Service Learning course at the University of Connecticut.This work seeks to provide estimates of the regional flux of nitrous oxide to the atmosphere, to examine seasonal dynamics, to assess the biological pathways to nitrous oxide production and consumption, and to query regional and large scale forcings on nitrous oxide dynamics in the SBUS. To this end, the team will (a) measure nitrous oxide concentrations in samples collected in the SBUS during seasonal surveys, measure surface nitrous oxide continuously underway, and derive robust estimates of the sea-to-air flux of nitrous oxide from coincident windspeed; (b) query nitrous oxide production and consumption pathways from measurements of the nitrogen and oxygen isotope ratios of nitrous oxide and nitrate, and investigate environmental correlates of nitrous oxide cycling; (c) develop regional machine learning models for nitrous oxide from these measurements to investigate environmental drivers of seasonal and inter-annual nitrous oxide dynamics, predicting nitrous oxide in the SBUS with historical data and with hydrographic fields derived from a regional dynamical ocean model. The work will give insights into pathways of nitrous oxide production and consumption in an eastern boundary upwelling system bounded by a broad continental shelf – contrasting the greater body of knowledge obtained from active margins. The statistical nitrous oxide models borne of machine learning will ultimately serve to predict nitrous oxide from climate projections of the SBUS, and to anticipate the regional response of nitrous oxide to global ocean de-oxygenation.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.