ATD: Collaborative Research: Computationally Efficient Algorithms for Detecting Anomalous Atmospheric Emissions

温室气体的大规模异常排放和大气污染对排放点附近的人类健康构成威胁，损害国家排放目标，威胁能源安全。最近两起备受瞩目的天然气井喷事故凸显了早期发现和干预的必要性。几颗新的和即将发射的卫星具有探测和监测温室气体排放的具体目的，最近的研究表明利用卫星数据探测这类事件的潜力。然而，由于需要处理大量卫星数据，并且需要在精细尺度分辨率下进行重建以进行威胁检测，因此在量化这些发射异常或超级发射器方面存在巨大的计算挑战。该项目旨在通过开发用于探测大气排放的改进计算方法来解决这些挑战。该项目每年资助三所大学各一名毕业生。该项目旨在解决反问题计算效率求解器发展中的基本问题，并通过使研究人员能够快速、准确地检测和监测异常大气排放，并具有可量化的不确定性，从而突破卫星威胁检测的传统界限。该项目的主要重点是：(i)有效地结合先验信息和参数选择，（ii）改进时空逆建模，采用多种随机成分和降低成本的非精确和抽样方法来处理昂贵的伴随模型，以及（iii）通过与合成卫星数据的案例研究评估，测试和集成开发的方法。该项目的目的是利用卫星帮助确定潜在的直接威胁（例如，石油和天然气井喷），这不仅在灾害响应和恢复方面，而且在尽量减少长期环境风险方面具有重大的广泛影响。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。

Large-scale anomalous emissions of greenhouse gases and air pollution pose threats to human health in the vicinity of the emissions, compromise state emissions targets, and threaten energy security. Two recent, high-profile natural gas blowouts underscore the need for early detection and intervention. Several new and forthcoming satellites have the specific purpose of detecting and monitoring greenhouse gas emissions, and recent studies have demonstrated the potential of detecting such events using satellite data. However, there are enormous computational challenges in quantifying these emission anomalies or super-emitters due to the massive amounts of satellite data to be processed and the fine-scale resolution at which reconstructions are needed for threat detection. This project aims to tackle these challenges by developing improved computational methods for use in detection of atmospheric emissions. The project supports one graduate per year at each of the three universities.The project aims to address fundamental issues in the development of computationally efficient solvers for inverse problems, and to push the traditional boundaries of threat detection via satellites by enabling researchers to detect and monitor anomalous atmospheric emissions quickly, accurately, and with quantifiable uncertainty. The main thrusts of this project are (i) efficient incorporation of prior information and parameter selection, (ii) improved spatio-temporal inverse modeling with multiple stochastic components and cost-cutting inexact and sampling approaches to handle expensive adjoint models, and (iii) evaluations, testing, and integration of the developed methods via case studies with synthetic satellite data. The aim of this project is to help identify potential immediate threats (e.g., oil and gas blowouts) using satellites, which have significant broader impacts not only in disaster response and recovery but also in minimizing the long-term environmental risks.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.