14. Summary of the complete Einstein@Home S3 analysis results

After we initially analyzed the S3 data set, we made several improvements:

$\bullet$ The data was re-calibrated using the final data-quality assessment (called the version 5 data-quality flags) and instrument calibration (called the version 3 instrument calibration).

$\bullet$ More known line artifacts were removed from the data set to prevent spurious candidates.

$\bullet$ In addition to hardware injections, software injections were also included. For an explanation of these injections see the section on hardware/software injection.

$\bullet$ Segment pairs were re-grouped to maximize the number of pairs for which both ten-hour segments contain hardware-injected signals.

This final re-analysis of S3 was completed in August 2005 (then followed by a long period in which the results and analysis tools were reviewed and vetted). The following two plots in Fig. 14.1 summarize the results of YOUR Einstein@Home analysis of the LIGO S3 data.

Figure 14.1: All-sky map of the entire S3 results over all frequencies (50Hz to 1.5kHz) (a) including the hardware/software injections, and (b) with all injections removed.

(a)

(b)

Figure 14.1 (a) shows the all-sky map of all the LIGO S3 candidates for the entire search frequency band from 50 Hz to 1.5 kHz. This figure includes the frequency bands in which simulated pulsar signals were injected in software or hardware. Figure 14.1 (b) shows the corresponding all-sky map excluding the "fake signal bands" (of order 0.2 Hz) in which simulated pulsars were injected in software or hardware. Readers who are interested may find the parameters of the hardware and software injections in this table. The precise frequencies of these fake signal bands are given in the table's second and third column.

We see in this figure that, once we have removed the injected, fake pulsar signals, there are no significant candidates left outside of the "noisy belt", which was described in section 12. Candidates on the "noisy belt" have to be discarded, because such candidates would be produced by line-noise artifacts of the detector, as discussed in sections 12 and 13. There are still eight surviving low-significance candidates outside of the "noisy belt", even after removal of the injected pulsars. These low-significance candidates were ruled out either by a followup analysis using S4 data, or by subsequent identification as noise-artifacts (e.g. one candidate was found to be associated with a 70.12 Hz oscillator in one of the VME electronics crates connected to the interferometer instrumentation).

We therefore conclude that we have not seen any credible evidence for pulsar signals in the Einstein@Home search on S3 data. This is not surprising. The LIGO instruments were still undergoing commissioning to reach their design sensitivity at the time of the S3 science run. And even at design sensitivity, the level of noise in the instruments is high enough that it may obscure gravitational wave signals present in the data.

Whether or not we can detect something in the future depends upon nature, luck, and our ability to improve the detector and our analysis methods. Much progress has been made since the S3 analysis. This progress and our plans for the future are discussed in the next section.

Contents[1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16]Bibliography

Einstein@Home S3 Analysis Summary	Last Revised: 2007.03.28 08:59:23 UTC
Copyright © 2005 Bruce Allen for the LIGO Scientific Collaboration	Document version: 1.132