Observing that the glitches reported in the previous post have the arc-like shape typical of Scattered Light, I used the usual code for studying this noise and it was quite easy to find the location of the scatterer that generates these arcs. By correlating all channels with position units, from which their instantaneous velocity was derived, it turned out that a large number of channels related to the Suspended North End Bench had a correlation larger than 80% with the arc frequency in Hrec. In the attached text file, I report the list of those with Pearson coefficient greater than 0.3.

Figure 1: spectrogram of Hrec during the period of excess noise with superimposed the arc frequency obtained from the most correlated channel, SBE_SNEB_SA_F0_diff_LVDT_z_500Hz, with Pearson coefficient of 0.83.

The highlighted period of excess noise and glitches appears to be concomitant with an earthquake recorded by the Guralp seismometer of the three corner buildings, North (blue), West (orange) and Central (green).

Figure 1: signals of the seismometers in the three buildings in the west direction in the hours around the period identified in the first post.

In the same time interval also monitors such as V1:ENV_SEISMON_N_WARNINGS, V1:ENV_SEISMON_MAG_1,  V1:ENV_SEISMON_GPS_LATEST_1 and  V1:ENV_SEISMON_GPS_EARLIEST_1 were active.