Minute-long gravitational-wave (GW) transients are currently a little-explored regime, mainly due to a lack of robust models. As searches for long-duration GW transients must rely on minimal assumptions about the signal properties, they are also sensitive to GWs emitted from unpredicted sources. The detection of such sources offers exciting and strong potential for new science. Because of the large parameter space covered, all-sky long-duration transient searches require model-independant processing and fast analysis techniques. For my PhD thesis, I integrated a set of fast cross-correlation routines in the spherical harmonic domain (SphRad) [50] into X-pipeline [95], a targeted GW search pipeline commonly used to search for GW counterparts of short and long duration GRBs & core-collapse supernovae. Spherical harmonic decomposition allows for the sky position dependancy of the coherent analysis to be isolated from the data [40] and cached for re-use, saving both time and processing units. Moreover, the spherical harmonic approach offers a fundamentally different view of the data, allowing for new possibilities for rejecting non-Gaussian background noise that could be mistaken for a GW signal. The combined search pipeline, X-SphRad, underwent a thorough internal review within the LIGO collaboration, which I led. The pipeline good functioning was assessed by rigorous tests including comparing a test data set with a standard sky grid-based analysis. I have developed a novel pixel clustering method that does not depend on the amplitude of potential signals. By using an edge detection algorithm, I quantify each pixel in the spectrogram by its similarity with its neighbours then extract features of sharply changing intensity (or ‘edge’). The method has shown promising results in preliminary tests. A simplified version of the algorithm was implemented in X-SphRad and large-scale testings are currently being processed.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:738384 |
| Date | January 2017 |
| Creators | Fays, Maxime |
| Publisher | Cardiff University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://orca.cf.ac.uk/110245/ |
Page generated in 0.0019 seconds