Noise parametric identification and whitening for LIGO 40-m interferometer data

One of the goals of gravitational data wave analysis is the knowledge and accurate estimation of the noise power spectral density of the data taken by the detector, this being necessary in the detection algorithms. In this paper we show how it is possible to estimate the noise power spectral density of gravitational wave detectors using modern parametric techniques and how it is possible to whiten the noise data before they pass to the algorithms for gravitational wave detection. We report the analysis we made of data taken by the Caltech 40-m prototype interferometer to identify the noise power spectral density and to whiten the sequence of noise. We concentrate our study on data taken in November 1994; in particular, we analyze two frames of data: the 18nov94.2.frame and the 19nov94.2.frame. We show that it is possible to whiten these data, to a good degree of whiteness, using a high order whitening filter. Moreover, we can choose to whiten only a restricted band of frequencies around the region we are interested in, obtaining a higher level of whiteness.