LEAPS-MPS: A Systematic and Automatic Spectral Analysis of Physical Conditions of Circumgalactic Medium using Genetic Algorithms;

天文光谱学是我们了解宇宙及其组成部分(如恒星、行星、星系、星际介质)的基础，对于几代天文学家来说，光谱学数据的分析是由经验丰富和训练有素的最终用户--天文学家自己进行的。这一模式将受到新一代光谱分析仪器的严重挑战，这些仪器将以比分析更快的速度在相对较短的时间尺度和数量级上产生数十亿个高分辨率光谱。研究团队将通过开发基于遗传算法(GA)的开源光谱分析包来解决这一“数据消防水管”问题，该包可以在最少的人工干预下准确地自动分析天文光谱，然后作为概念验证，使用它来探索星系气态晕中的物理条件。这个项目将支持两名研究生的论文研究，并允许PI在他的家乡机构进一步努力，让学生参与计算天体物理学和其他STEM领域的学生。该项目是由已建立的激励竞争研究计划(EPSCoR)共同资助的。建立适用于光谱分析中快速和快速增长的数据集的快速和自动化的全局最优搜索方法是一个复杂而长期存在的问题。拟议的工作将促进多个领域的知识，方法是开发一种基于遗传算法的全局优化方法，用于分析适用于现代数据环境中快速增长的大型数据集的高分辨率光谱，然后将其应用于利用HST的宇宙起源光谱仪(COS)获得的背景源档案光谱，促进对星系的环星系(CGM)和星系间(IGM)介质的理解。项目成果将包括一个可扩展和开放源码的软件包(Astro-Neo)，它将使天文学家能够以更高的质量和重复性分析大量数据。该核心软件已成功应用于材料科学领域，并将扩展到本项目生命周期以外的其他领域。当数据以拍字节/年的速度获得时，项目成果将直接解决数据分析方面的挑战，并为加速发现提供一个新的研究平台。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。

Astronomical spectroscopy lies at the foundation of our understanding of the universe and its constituents (e.g., stars, planets, galaxies, interstellar medium), and for generations of astronomers, the analysis of spectroscopic data has been carried out by experienced and highly trained end-users - the astronomers themselves. This paradigm will be severely challenged by new generations of spectrographic instruments that will produce literally billions of high-resolution spectra on relatively short timescales and orders of magnitude faster than they can be analyzed. The research team will address this "data firehose" problem by developing an open-source spectral analysis package based on a genetic algorithm (GA) to accurately analyze astronomical spectra automatically with minimal human intervention, and then use it to explore the physical conditions within the gaseous halos of galaxies as a proof-of-concept. This project will support the thesis research of two graduate students and allow the PI to further efforts at his home institution to involve students in computational astrophysics and other STEM fields. This project is co-funded by the Established Program to Stimulate Competitive Research (EPSCoR).Establishing fast and automated global optimum search methods applicable to large and rapid growing datasets in spectral analysis is a complex and long-standing problem. The proposed work will advance knowledge in multiple areas by developing a GA-based global optimization method for the analysis of high-resolution spectra that is applicable to large and rapid growing datasets in modern data environments and then applying it to facilitate understanding of the circumgalactic (CGM) and intergalactic (IGM) media of galaxies using archival spectra of background sources obtained with the HST's Cosmic Origin Spectrograph (COS). The project outcomes will include an extensible and open-source code base software package (Astro-Neo) that will enable astronomers to analyze large quantities of data with improved quality and reproducibility. The core software has been applied successfully in materials science and will be extended to other fields beyond the lifetime of this project. The project outcomes will directly address the challenges of data analysis when data is obtained at Petabyte/year rates and contribute to a new research platform for accelerating discovery.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.