RUI: Rapid Acquisition and Analysis of Complex Spectra with High-Speed Digitizers

这项由化学系化学测量与成像（CMI）项目颁发的本科院校研究（RUI）奖，由计算机与数据支持科学与工程（CDS&amp；E）和行星天文学项目共同资助，支持佛罗里达新学院的Steven T. Shipman教授和他的学生开发新工具，帮助科学家分析来自行星大气层的数据，包括我们自己的大气层和空间中的大型分子云。新工具基于高速电子设备，使科学家能够收集到比以前更多的样本信息。正在开发利用这些高速数据收集器的软件，并使用新的计算技术来处理涌入的数据。所有的工作都是由本科生和首席研究员一起完成的。新软件将被公开共享。研究人员正在使用新的高速数字化仪和网格计算技术，在大约250到325 K的温度范围内，快速获取和分析与星际和大气化学相关的分子的旋转光谱。在这个温度范围内，由于大振幅运动和热填充激发振动和构象状态的贡献，旋转光谱非常复杂。与当前仪器相比，一种具有近8000倍速度优势的新型数字化仪正在与温度依赖和微波-微波双共振测量相结合，以自动确定观测光谱中大量峰的低状态能量和能级连通性。正在开发的基于遗传算法和其他方法的软件将利用这些信息大大简化频谱分配过程。这些算法将被移植到网格计算平台上，以最大限度地利用它们的并行性，并进一步缩短频谱分析时间。

This Research at Undergraduate Institutions (RUI) award from the Chemical Measurement and Imaging (CMI) Program in the Division of Chemistry, with co-funding from the Computer and Data-Enabled Science and Engineering (CDS&E) and Planetary Astronomy Programs, supports Professor Steven T. Shipman and his students at New College of Florida as they develop new tools that will help scientists analyze data from the atmospheres of planets, including our own and large clouds of molecules in space. The new tools are based on high-speed electronics that allow scientists to collect significantly more information about their samples than was previously possible. Software is being developed that takes advantage of these high-speed data-collectors and uses new computing techniques to process the influx of data. All the work is being performed by undergraduate students working with the principal investigator. The new software will be publicly shared.The investigators are using new high-speed digitizers along with grid computing techniques to rapidly acquire and analyze rotational spectra of molecules of relevance to interstellar and atmospheric chemistry at temperatures ranging from roughly 250 to 325 K. In this temperature range, rotational spectra are extremely complex due to contributions from large amplitude motion and thermally-populated excited vibrational and conformational states. A new digitizer with a nearly 8000x speed advantage over current instrumentation is being used in conjunction with temperature-dependent and microwave-microwave double resonance measurements to automatically determine lower state energies and energy level connectivities of a large number of peaks in observed spectra. Software that is being developed, based on genetic algorithms and other approaches, will use this information to greatly ease the spectral assignment process. These algorithms will be ported to a grid computing platform to take maximum advantage of their parallelism and to further reduce the spectral analysis time.