EAGER: ADAPT: Time-Domain Study of the Dynamics of Relativistic Jets

来自活动星系核的相对论性喷流是由星际气体吸积到活动星系中心的超大质量黑洞所产生的粒子束和辐射。从无线电波到伽马射线能量，整个电磁波谱都可以观测到喷流。它们在星系中的存在已被证明与新恒星的形成速度有关。在美国国家科学院最近的一份报告中，目前对活跃星系核中相对论性喷流的结构和演化的有限认识被描述为“理解星系演化的主要瓶颈”。喷流是非常动态的天体物理物体。从喷流中观测到的光发射通常会在数年内缓慢漂移。强度的快速变化可以持续数周、数天甚至数小时。随着时间的推移，喷流发射的变化携带着构成喷流等离子体的粒子（主要是电子、正电子和质子）能量的信息，这为理解使相对论性喷流成为宇宙中最强大的粒子加速器的物理过程打开了一扇窗。该项目将利用圣路易斯华盛顿大学统计学、观测天体物理学和理论等离子体物理学领域研究人员的综合专业知识，开发一个新的统计和人工智能工具框架，计算机模拟和分析物理模型，描述无线电、光学和伽马射线望远镜所看到的相对论性喷流随时间演变的特征。并将其与驱动喷气动力的物理过程联系起来。这项研究可以加强对相对论性喷流和撞击物理学以及其他学科的物理机制的理解。该项目将对本科和研究生教学、天文和数据科学培训和技能的人力资源开发，以及针对数学、统计、物理和天文学领域代表性不足的少数民族学生的外联活动产生更广泛的影响。在过去的十年中，欧文斯谷射电天文台和无线电频率的超大基线阵列，兹威基瞬变设施（光学）和费米- lat天文台（伽马射线）等设施已经产生并继续提供具有前所未有的高节奏和时间覆盖的光曲线，用于活动星系核的明亮喷流。与此同时，近年来，理论研究也取得了迅速进展，物理学家对粒子加速机制的理解有所提高，这种机制可以产生观测到的高能喷射，特别是通过等离子体模拟。尽管第一原理理论建模取得了进展，但理论模型的输出与观测数据之间存在显著差距。该项目旨在开发和利用新的统计和人工智能工具来了解差异的主要来源，并将观察到的光曲线与最先进的细胞内粒子模拟相结合，以缩小差距。作为第一步，将生成一个模拟光曲线库，并开发一套新的诊断工具，用于多变量、不规则间隔的时域数据的时间序列分析。接下来，将使用多波段时间序列分析的重要特征来表征模拟和观测光曲线的特性，这将有助于区分不同类别的理论模型。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。

Relativistic jets from active galactic nuclei are beams of particles and radiation powered by accretion of interstellar gas onto a supermassive black hole at the center of active galaxies. Jets are observed across the electromagnetic spectrum, from radio waves to gamma-ray energies. Their presence in galaxies has been shown to correlate with the formation rate of new stars. Limited knowledge of the structure and evolution of relativistic jets in active galactic nuclei at present has been described as “a major bottleneck in understanding the evolution of galaxies” in a recent report by the National Academy of Sciences. Jets are very dynamic astrophysical objects. The light emission observed from jets is often seen to drift slowly over years. Fast changes in intensity that can last weeks, days or even hours are also observed. Changes in the jet emission over time carry information about the energy of the particles (mostly electrons, positrons, and protons) that the jet plasma is made of, opening a window to understand the physical processes that make relativistic jets the most powerful particle accelerators in the Universe. This project will leverage the combined expertise of researchers in the fields of statistics, observational astrophysics, and theoretical plasma physics at Washington University in St. Louis to develop a framework of new statistical and artificial intelligence tools, computer simulations, and analytical physical models that characterize the evolution of relativistic jets over time as seen by radio, optical, and gamma-ray telescopes, and will connect it to the physical processes powering the jet dynamics. The research could enhance the understanding of the physical mechanisms underlying relativistic jets and impact physics and other disciplines. The project will have broader impacts on undergraduate and graduate teaching, the development of human resources with training and skills in astronomy and data science, and outreach activities targeting local students from underrepresented minorities in the fields of mathematics, statistics, physics, and astronomy.In the last decade, facilities such as the Owens Valley Radio Observatory and Very Large Baseline Array at radio frequencies, the Zwicky Transient Facility (optical), and the Fermi-LAT observatory (gamma rays) have produced and continue to deliver light curves with unprecedented high cadence and time coverage for bright jets from active galactic nuclei. Alongside, theoretical efforts have also progressed rapidly in recent years, increasing physicists’ understanding of the mechanisms of particle acceleration that can produce the observed high energy emission from jets, particularly through plasma simulations. Despite the progress in first-principles theoretical modeling, there is a significant gap between outputs of theoretical models and observational data. This project seeks to develop and make use of novel statistical and artificial intelligence tools to understand the main sources of the discrepancy and combine observed light curves with the state-of-the art particle-in-cell simulations to narrow the gap. As a first step, a library of simulated light curves will be generated and a novel set of new diagnostic tools for time series analysis of multivariate, irregularly spaced time domain data will be developed. Next, important features from the multi-band time series analysis will be used to characterize properties of both simulated and observed light curves, which will help to discriminate among different classes of theoretical models.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.