LIGO and Virgo Data

Data Usage Notes   Please Read This First!

LIGO and Virgo data cannot be effectively used without understanding these features. See the links below for more details.

  • 3 LIGO detectors operated during LIGO science run 5 (S5, Nov 2005 to Oct 2007):
    • H1 and H2 at Hanford, Washington (LHO), and L1 at Livingston, Louisiana (LLO).
  • 2 LIGO detectors operated from 2009 onwards, during LIGO science run 6 (S6, Jul 2009 to Oct 2010) and LIGO observing runs 1 (O1, Sep 2015 to Jan 2016) and 2 (O2, Nov 2016 to Aug 2017):
    • H1 at Hanford, Washington, and L1 at Livingston, Louisiana.
  • The Virgo detector operated from 2007 onwards, with an overlap with the LIGO science runs 5 and 6, and the LIGO observing run 2 (O2):
    • V1 at Cascina, Italy
  • Detector positions are as recorded in LALSuite
  • Injections in the data are simulated versions of astrophysical signals.
    • If you 'discover' an interesting signal in LIGO data, it could be a known injection.
    • Lists of injection times are available for the S5, S6, and O1 data sets.
  • Instrumental glitches are in the data. They are not astrophysical in origin, but can mimic such signals. More information is available on the Data Release pages above.
  • Depending on the data set, the LOSC strain data are available in the original 16384 Hz sample rate, or downsampled from 16384 Hz to 4096 Hz.
  • Data quality categories are used to characterize the quality of data at a given time.
    • Data which fails CAT 1 flags should not be searched for astrophysical signals.
    • All data analysis by the LIGO Scientific Collaboration vetos data that fail CAT 1 or CAT 2 flags.
    • See the run documentation above for details
  • Instrumental lines in the data are seen at quasi-constant frequencies that are artifacts, and can mimic astrophysical signals.
  • Gaps in the data occur when the detectors were not in their standard configuration (e.g maintenance).
  • Short Segments: Standard signal processing methods applied to very short segments (<16 sec) must be used with caution and could produce spurious results.
  • Frequency dependent noise in the data is only roughly Gaussian.
    • The noise spectrum also varies with time.
    • See the plot gallery.
  • Advanced LIGO and Virgo strain data are not calibrated below 10 Hz. Advanced LIGO detector noise rises rapidly below 10 Hz; it is many orders of magnitude larger than any plausible gravitational wave strain signal. Further, the data made available on LOSC are aggressively high-pass-filtered at 8 Hz in order to avoid downstream signal processing problems, so it doesn't properly represent either signal or noise at those low frequencies. In particular, any DC offsets have no significance. In the Initial LIGO (S5 and S6) datasets, data below 40 Hz is similarly uncalibrated and suppressed. The 8 Hz roll-off is visible in this example low frequency spectrum.
Return to the Data Page to download LIGO and Virgo data.