Proposal for A Stochastic-Signal-Model-Based Search for Intermittent Gravitational-Wave Backgrounds

引力波天文学观测领域始于2015年9月14日的一声“巨响”，探测到两个相互碰撞的黑洞合并产生的引力波。这一探测打开了一扇通往宇宙“阴暗面”的窗户，提供了观察天体物理物体和事件的手段，否则用标准光学望远镜是不可能看到的。迄今为止，包括美国国家科学基金会的LIGO在内的几个大型引力波探测器已经探测到大约100个来自相对较近的黑洞和/或中子星的“响亮”事件。但是来自更遥远的黑洞对的联合信号还没有被探测到。这个项目是专门针对这种信号设计的，这种信号（用听觉来类比）听起来就像爆米花发出的声音。换句话说，该信号由持续时间短（~秒）的引力波的微弱爆发组成，中间间隔着相对静默的周期（~几分钟）。作为该项目的一部分，开发的数据分析工具将明确考虑到信号的爆米花性质，从而导致更敏感的搜索，并可能在未来几年内首次检测到这种类型的信号。该项目将为一到两名研究生提供数据分析方面的支持和培训，从而增加这一新兴领域不断增长的研究人员群体。学生将获得的计算和数据分析技能可以转移到引力波天文学领域之外，使学生在各种学科中都有市场——有可能成为未来的大学教授，也有可能在大学之外的研究实验室或高科技公司工作。拟议的项目包括三个主要活动，其范围和复杂性在提议期间有所增加：(i)第一，在一套相对简单的玩具模型的背景下，提供基于随机信号的搜索爆米花状（间歇性）引力波信号的“原理证明”示范。（ii）第二，扩展第(i)部分开发的数据分析管道，以便在更实际的数据集上运行，从而对拟议的搜索进行压力测试。（iii）第三，运行第（ii）部分开发的管道的生产版本，该管道将在第4次观测运行O4期间获得先进的LIGO-Virgo-KAGRA数据，该数据将在2022年底/ 2023年初开始。提出的基于随机信号的搜索有可能推动引力波天文学领域的发展，因为它是第一个探测宇宙中恒星质量黑洞对合并信号的搜索。这是可能的，因为搜索考虑了信号的间歇性，与当前搜索相比，通过增加恢复信号幅度的信噪比，可以缩短检测时间，而当前搜索假设信号一直处于“打开”状态。此外，通过使用随机信号模型，与基于确定性信号的搜索相比，搜索对源类型的鲁棒性更强，计算需求更少，后者被调谐到与双黑洞合并相关的特定波形。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。

The field of observational gravitational-wave astronomy began with a "bang" on 14 September 2015 with the detection of gravitational waves from the merger of two inspiraling and colliding black holes. This detection opened a window into the "dark side" of the universe, providing the means to observe astrophysical objects and events that would otherwise be impossible to see with standard optical telescopes. To date, approximately 100 "loud" events, from relatively nearby black holes and/or neutron stars, have been detected by several large-scale gravitational-wave detectors including NSF's LIGO. But the combined signal from the population of more distant pairs of black holes has yet to be detected. This project is designed precisely to target this signal, which (using the analogy of hearing) would sound like popcorn popping. In other words, the signal consists of weak bursts of gravitational waves of short duration (~seconds) separated by periods (~a few minutes) of relative silence. The data analysis tools developed as part of this project will explicitly take into account the popcorn-like nature of the signal, leading to a more sensitive search and a possible first detection of this type of signal within the next few years. The project will provide support and training in data analysis to one or two graduate students, thus adding to the growing community of researchers in this emerging field. The computational and data analysis skills that the students will acquire are transferable outside the field of gravitational-wave astronomy, making the students marketable in a variety of disciplines--potentially as future university professors or outside the university setting in research labs or high-tech companies.The proposed project consists of three main activities, which increase in scope and complexity over the period of the proposal: (i) First, to provide a "proof-of-principle" demonstration of a stochastic-signal-based search for popcorn-like (intermittent) gravitational-wave signals in the context of a set of relatively simple toy models. (ii) Second, to extend the data analysis pipeline developed in part (i) to run on more realistic data sets, thus stress-testing the proposed search. (iii) Third, to run a production version of the pipeline developed in part (ii) on the Advanced LIGO-Virgo-KAGRA data taken during the 4th observation run O4, which will start near the end of 2022 / beginning of 2023. The proposed stochastic-signal-based search has the potential to advance the field of gravitational-wave astronomy by being the first search to detect the signal from mergers of pairs of stellar-mass black holes throughout the Universe. This is possible because the search takes into account the intermittent nature of the signal, which should lead to a reduced time-to-detection by increasing the signal-to-noise ratio of the recovered signal amplitude compared to the current search, which assumes that the signal is "on" all the time. In addition, by using a stochastic-signal model, the search is both more robust to the type of source and less computationally demanding than a deterministic-signal-based search, which is tuned to the specific waveforms associated with binary black hole mergers.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.