Did LIGO really detect Gravitational Wave signals or was it just noise?


After the Laser Interferometer Gravitational-Wave Observatory (LIGO) first observed Gravitational Waves in September 2015 and then again in December 2015 and for the third time in January 2017, a team of researchers from Niels Bohr Institute in Copenhagen, Denmark, calls into question whether the LIGO observatory had actually detected the gravitational wave signals.

“The detection of the gravitational wave events needs to be re-evaluated with more careful consideration of noise properties,” James Creswell from the Niels Bohr Institute & Discovery Centre and the first author writes in a paper (“On the time lags of the LIGO signals”) uploaded in the Cornell University Electronic Preprint achieve arXiv on June 13, 2017. The paper has not been peer-reviewed.

In September 2015, the two laser interferometers located in Livingston, Louisiana and the other in Hanford, Washington in Hanford recorded signals indicative of the merger of two black holes.  Since the two detectors are separated from each other by nearly 3,000 km, there was a time delay of 6.9 milliseconds for the detection of the signals between the two detectors. This time lag falls within the window of 10 milliseconds, which is clearly essential for validation of detection and provides some clue about the direction from where the waves had originated.

The time difference of about 6.9 milliseconds is vital as it is indicative of the difference in path lengths of the two stations with regard to the cosmic point source. Besides the time-lag, the pattern of the extracted signal should be the same for both the stations, i.e., there should be similar correlations in terms of amplitude and frequency.

But as far as terrestrial noise (signals that do now correspond to gravitational waves), which is produced by a wide variety of sources such as quantum noise and seismic noise, at the two detectors that are separated by a large distance cannot be correlated. And the time-lag of the noise cannot perfectly match the delay seen in the case of gravitational waves.

However, the researchers found that the noise was correlated and had the same time-lag. “The residual noise in the H and L [the two detectors] data also has the same time lag,” they write. “The more general presence of correlations with time delay would significantly reduce the ability to identify and measure gravitational wave signals with a similar time delay.”

“Similar effects are also found in the data for [December 2015 and January 2017 events], this suggests the presence of unwanted systematic effects,” the write.

“The purpose in having two independent detectors is precisely to ensure that, after sufficient cleaning, the only genuine correlations between them will be due to gravitational wave effects. The results presented here suggest this level of cleaning has not yet been obtained and that the detection of the gravitational wave events needs to be re-evaluated with more careful consideration of noise properties,” Dr. Creswell concludes.

“If the correlation properties of signal and the noise are similar, how is one to know precisely what is signal and what is noise?” Dr. Andrew Jackson, the leader of the Danish group, was quoted as saying in Forbes

Author responds

“We are in the midst of discussions with the LIGO people in an attempt to find common ground.  Until these discussions have been completed, we would prefer not to talk to journalists.  As you may imagine, this work has generated considerable interest.  We have raised questions regarding certain aspects of the analysis of the LIGO data. It is our desire to contribute to a scientifically sound (and amicable) resolution of these concerns. This must be our first priority,” Dr. Jackson from the Niels Bohr Institute and one of the authors of the “On the time lags of the LIGO signals” paper writes in an email to me in response to a list of questions I had asked the authors.

Countering the claims

Meanwhile, Ian Harry, postdoctoral physicist at the Max Planck Institute for Gravitational Physics, Potsdam-Golm and a member of the LIGO collaboration who had worked on the analysis of LIGO data has in a blog post responded to the concerns raised in the Niel Bohr authors’ paper.

“The conclusions I draw here have been checked by a number of colleagues within the LIGO and Virgo collaborations. We are also in touch with the authors of the article to raise these concerns directly, and plan to write a more formal short paper for submission to the arXiv explaining in more detail the issues I mention below. In the interest of timeliness, and in response to numerous requests from outside of the collaboration, I am sharing these notes in the hope that they will clarify the situation,” Dr. Hary writes.

“The authors [Dr. Creswell et al] take LIGO data made available through the LIGO Open Science Data from the Hanford and Livingston observatories and perform a simple Fourier analysis on that data. They find the noise to be correlated as a function of frequency. They also perform a time-domain analysis and claim that there are correlations between the noise in the two observatories, which is present after removing the GW150914 [September 2015] signal from the data. These results are used to cast doubt on the reliability of the GW150914 [September 2015] observation.

“There are a number of reasons why this conclusion is incorrect: 1). The frequency-domain correlations they are seeing arise from the way they do their FFT on the filtered data. We have managed to demonstrate the same effect with simulated Gaussian noise. 2). LIGO analyses use whitened data when searching for compact binary mergers such as GW150914 [September 2015 event]. When repeating the analysis of Creswell et al. on whitened data these effects are completely absent. 3). Our 5-sigma significance comes from a procedure of repeatedly time-shifting the data, which is not invalidated if correlations of the type described in Creswell et al. are present,” Dr. Hary writes.

“I find the claims of section 2 [the September 2015 event and its Fourier amplitudes and phases] are due to an issue in how the data is Fourier transformed, and do not reproduce the correlations claimed in section 3 [Correlations in the noise]. Even if taking the results at face value, it would not affect the 5-sigma confidence associated with GW150914 [September 2015 event]. Nevertheless I am in contact with the authors and we will try to understand these discrepancies,” Dr. Hary concludes.