Reports 1-1 of 1 Clear search Modify search
Detector Characterisation (Glitches)
robinet - 14:08 Friday 12 April 2019 (45627) Print this report
Magnetic glitches coupling to h(t)

Looking at the loudest pycbc glitches, I found a population of glitches which seem to result from magnetic glitches. They are associated to the 50 Hz activity.

Looking at spectrograms, there is a very loud glitch detected in many magnetic sensors (Figure 1) which is also detected in h(t) (Figure 2). After the glitch, the 50 Hz line seems to be excited.

These glitches are caught by many BRMSMon data quality flags: DQ_BRMSMon_FLAG_ALL_MAG_*. 

Here is a command line to print a list of GPS times for these glitches:

source /virgoApp/Omicron/v2r3p8/cmt/
omicron-print channel=V1:ENV_NEB_MAG_V gps-start=1238976018 gps-end=1239062416 snr-min=50 freq-min=49 freq-max=52

It returns for yesterday:

1239019305.0391 51.01 513.23
1239024388.8516 51.01 701.82
1239025834.7734 51.01 79.71

In Figure 3, I present the spectrogram of a high-SNR glitch in h(t) resulting from a magnetic glitch. For this glitch there is even a dip in the horizon! I believe this example belongs to a separate family of magnetic glitches. This time, it does not seem to be linked to the 50 Hz activity.

ENV experts should probably have a look at these events...


Images attached to this report
Comments to this report:
swinkels - 17:23 Friday 12 April 2019 (45630) Print this report
All 3 glitches are seen very clearly in the signals of the external magnetometers and all the ones located inside the buildings, and hardly on the IPS and UPS lines. These are likely caused by lightning strikes.

According to the awesome Blitzortung (see the historical data in the menu), there was significant thunderstorm activity in Tuscany around that time. It would be great if we could join that network by installing one of their sensors, which should give you access to their database with lightning strike times and locations. Rosario should be on the waiting list to join, but unfortunately that seems to take years on their side. These glitches should anyhow be easy to veto using the magnetometers.
Images attached to this comment