Electromagnetic follow-up of gravitational wave candidates

Observations of astrophysical systems in different wavelengths can reveal insights in to systems which are not available from a single wavelength. The same can be expected from multi-channel observations of systems which also produce gravitational waves (GWs). The most likely source of strong, detectable GWs, which will also produce an electromagnetic (EM) signature, is the merger of compact objects containing neutron stars (NS) and black holes (BH), namely NS-NS and NS-BH systems. The focus of this thesis is to summarise current and past efforts to detect an EM counterpart of a GW event, with emphasis on compact merger sources. To begin, the formulation of GWs in general relativity is brie y discussed, as well as the main classes of GW sources. The global networks of GW interferometers in the recent past and near future are described, together with brief explanations of operational principles and the main challenges GW detectors face to make a confident detection. Current literature is reviewed to give a brief summary of the most promising sources which produce both GW and EM signals. Emphasis is given to gamma-ray bursts (GRBs), their afterglows, and kilonovae. In addition a brief description of GW searches triggered by an external source (such as a GRB) is given. A new form of search is then discussed in which GW events are used to point conventional EM telescopes, with emphasis on rapidly slewing, wide field of view optical telescopes. The main challenge in this form of search is that timing information from a network of GW interferometers yields large error regions for the source sky direction making it diffcult to locate an EM transient. Therefore a new statistic is presented in which galaxies (taken from a galaxy catalogue) within this search region are ranked. The probability of identifying the host galaxy of a GW signal from NS-NS and NS-BH systems is investigated and results presented for past and future GW detector configurations. The ROTSE-III telescope system took part in this first search for EM counterparts of GW triggers. With four identical robotic telescopes located across the world it responded to five GW events. Presented is an automation of the ROTSE image processing pipeline which allows large-scale processing and automated validation and classification of candidates. A background study was conducted to better understand the optical transient background and to determine the statistical significance of candidates. Pipeline performance is tested by inserting simulated transients following kilonova and GRB lightcurves in to images; an efficiency study is described. Finally the results of the images taken in response to the five GW events are presented and discussed