The emerging field of gravitational-wave (GW) astronomy has entered an exciting phase with large-scale interferometric GW detectors (LIGO, GEO600, TAMA300, and Virgo) taking data in either science or engineering runs. The LIGO project completed its landmark fifth science run (S5) in 2007 and will begin a new one (S6) in summer 2009. This award supports research in three critical areas relevant to LIGO data analysis: (i) searches for a cosmological or astrophysical GW background, (ii) a statistical search for GWs from gamma-ray bursts, and (iii) detector characterization.
The first project will improve our sensitivity to GW backgrounds that have a non-uniform distribution of power on the sky. The second will develop methods to optimally use the information provided by NASA missions (Swift and Fermi) in GW searches. The third will use advanced data mining methods to develop a better understanding of real data, which will benefit the sensitivity of all GW searches. Three graduate students will receive direct training in GW data analysis, thus adding to the growing community of researchers in this emerging field. Since the University of Texas at Brownsville (UTB) is an Hispanic serving institution, the research activities will expose students who are traditionally under-represented in the areas of science and technology to forefront scientific research. On-going major outreach programs at UTB will be able to leverage this research, creating awareness among high-school students about exciting scientific projects such as LIGO.
The direct detection of gravitational waves produced by extreme events in the universe is expected to revolutionize our understanding of astrophysics and cosmology. Gravitational waves will give us a picture of the universe complementary to that from traditional electromagnetic astronomy (e.g., visible light, X-rays, radio waves, gamma rays, etc.). Although gravitational waves have not yet been directly detected, with improved detectors like Advanced LIGO coming on-line within the next few years, detections should become common-place. NSF grant 0855371 supported the development of new tools and algorithms to help us in our search for gravitational waves. Specifically, we worked on projects related to: (i) stochastic gravitational-wave backgrounds, (ii) a population study of gamma-ray bursts, and (iii) the use of data-mining tools to better understand the noise produced by the detectors. (i) A stochastic background of gravitational waves is basically a gravitational-wave "confusion noise" produced by either a large number of independent and unresolved astrophysical sources, or by waves left-over from the Big Bang. The direct detection of a stochastic background of cosmological origin would give us a picture of the universe mere fractions of a second after the Big Bang---a "holy grail" of gravitational-wave astronomy. For this project we developed a data analysis pipeline that can search for a non-uniform distribution of gravitational-wave power on the sky, similar to the maps showing temperature fluctuations in the cosmic microwave background radiation. We also conducted searches for stochastic backgrounds with a uniform sky distribution using data from the two co-located LIGO detectors in Hanford, WA, as well as the Virgo interferometer in Cascina, Italy. (ii) An astrophysical trigger is an event from a non-gravitational wave detector (e.g., a gamma-ray burst observed by the Swift or FERMI satellite) that is expected to have a gravitational-wave counterpart. Population study of astrophysical triggers is an approach where gravitational-wave data across multiple triggers is combined to infer properties about the astrophysical source population as a whole. Its main feature is that the properties of the population can be inferred even if the gravitational-wave signals are not strong enough to be detected individually. This approach has already been applied to LIGO data for a multi-gamma-ray burst analysis. In the quest to increase the sensitivity of population study, we have developed a new approach in which data from both gravitational-wave and non-gravitational-wave detectors are optimally combined into one joint analysis. Several issues with earlier population study methods, such as missing information for some of the triggers and ad hoc selection criteria for triggers, were addressed systematically and nearly automatically in the new method. We also solved several computational and algorithmic problems in implementing this method, enabling its eventual application to current and future LIGO-Virgo data. (iii) The LIGO project is an example of "big science", with the detectors generating raw data at the rate of several gigabytes per week. With such large data sets, it is imperative to resort to machine learning to draw meaningful inferences from the data. One important aspect of LIGO data analysis is to understand and characterize the underlying noise, especially the sharp "glitch" features that are seen over very short time scales. If we can identify the source of these glitches, we can use them as detector diagnostics; the end result being a cleaner, high-quality data that will benefit any search for gravitational waves. For this project, we used the cutting-edge software called "SkyTree" that has the capability to handle gigabyte-scale data for the purpose of mining information. We wrote a pipeline that can process all the information about LIGO noise glitches as input and then classify them into different categories based on statistical similarities. Each group is then correlated to determine their possible origin by means of sophisticated statistical analysis and visualization techniques. Training: Each of these projects was led by one or more graduate students at The University of Texas at Brownsville (UTB), working under the supervision of the PI and co-PIs. Since UTB is a minority-serving institution serving a predominantly Hispanic population, these research activities exposed students who are traditionally under-represented in the areas of science and technology to forefront scientific research. Also, by taking advantage of outreach programs at UTB, such as "The 21st Century Astronomy Ambassador's Program" and "Monday Night Physics", we were able to convey the excitement of projects, like those listed above, to a broader audience of young students.
