eROSITA’s Window into the X-ray-Transient Sky: AGN Accretion in the Era of Big Data

关于活动星系核(AGN)的燃料机制仍然存在悬而未决的问题：活动星系核的平均占空比是多少？最里面的吸积流的结构和形态是什么？这种流动如何依赖于全球吸积率？最近的光学和X射线观测已经产生了诱人的线索：罕见的吸积开启/关闭事件和视线柱状密度的变化，这是由于经过离散的团块造成的。然而，由于观测到的事件数量很少，进展受到限制。我们将使用eROSITA X射线望远镜汇编一个新的、扩展的X射线监测变率统计数据库，其范围超过1e6AGN。我们将使用机器学习来识别AGN中发生的新的变化状态和变化遮挡事件，从而增强目前为数不多的已知事件。关于吸积点火/猝灭事件的累积统计数据将揭示活动星系核的占空比，我们预期的多波长运动将监测活动星系核点火过程中吸积盘、BLR和X射线日冕的演化。我们还将监测大约800个Seyfert的云食事件，确定云的物理性质，并使用这些事件的累积统计数据来确定云的分布和物理来源。我们还将测量约60000个活动星系核中随质量和红移的变化幅度。最后，我们将监测大约1600颗耀星的快速X射线可变性，并与同时代的费米-LAT数据相关联，以测试所有正常状态下的单区和多区发射模型，而不仅仅是在爆发期间。eROSITA将产生唯一的数据库(在可预见的未来)，将X射线监测与目标活动星系核的统计显著样本结合在一起，因此也是唯一大量的X射线识别新的瞬变事件。我们的工作将首次产生基于X射线观测的活动星系核占空比估计，测试统一方案，并限制活动星系核点火事件中吸积流演变的速度。我们还将限制块状物质的分布和观察到的遮挡转变的速率，帮助我们理解X射线/光学类型的失配。我们还将了解Seyferts中的X射线可变性机制是否随红移而演变。我们的blazar变化子集将有助于区分当前各种blazar发射模型。我们的工作还为2020年代中期的更多大数据研究奠定了基础，这些研究也使用机器学习来对瞬变进行分类。我们的合作结合了AGN和黑洞双星的计时和光谱方面的互补专业知识，以及跨多个带通的高红移AGN的大样本。为了加强合作，该项目的新博士生将是双边的，由两国的联合导师共同监督。

Open questions persist regarding the fueling mechanisms in Active Galactic Nuclei (AGN): What is the average AGN duty cycle? What is the structure and morphology of the innermost accretion flow? How does that flow depend on global accretion rate? Recent optical and X-ray observations have yielded tantalizing clues: rare accretion turn-on/turn off events and variations in line-of-sight columndensities due to transiting discrete clumps. However, progress has been limited by the low number of observed events.We will use the eROSITA X-ray telescope to compile a new, expanded X-ray monitoring variability statistics database for more than 1e6AGN. We will use machine learning to identify new changing-state and changing-obscuration events in AGN as they occur, boosting the currently small number of known events. The accumulated statistics on accretion ignition/quenching events will reveal AGN duty cycles, and our expected multi-wavelength campaigns will monitor how the accretion disk, BLR, and X-ray corona evolve during AGN ignition. We will also monitor about 800 Seyferts for cloud eclipse events, determine clouds' physical properties, and use the accumulated statistics on such events to determine the distribution and physical origin of clouds. We will also measure variability amplitudes in about 60000 AGN as function of mass and redshift. Finally, we will monitor rapid X-ray variability in about 1600 blazars and correlate with contemporaneous Fermi-LAT data to test single- and multi-zone emission models at all normal states, not just during outbursts.eROSITA will yield the only database (for the foreseeable future) combining X-ray monitoring with a statistically significant sample oftarget AGN, and thus the only bulk X-ray identification of new transient events. Our work will yield the first X-ray observation-based estimate for AGN duty cycles, test unification schemes, and constrain how quickly the accretion flow can evolve in AGN ignition events. We will also constrain the distribution of clumpy matter and the rate of observed obscuration transitions, helping us understand X-ray/optical type mismatches. We will also learn if the X-ray variability mechanism in Seyferts evolves with redshift. Our blazar variability subset will help distinguish various current blazar emission models. Our work also sets the stage for additional Big Dataresearch in the mid 2020s that also use machine learning to classify transients.Our collaboration combines complementary expertise spanning timing and spectroscopy of both AGN and Black Hole Binaries and large samples of high-redshift AGN across multiple bandpasses. To cement the collaboration, new PhD students in this project will be bilateral, with joint supervisors from both countries.