Prompted by Michal, I verified if the occurrence of Scattered Light glitches from External Injection Bench motion associated with bad weather conditions has changed after the recent intervention and improvement of the EIB ground noise correction: #63763.
I examined the conditions during the last three main events of bad weather and Scattered Light glitches: February 25-26, March 26-27, and April 1-2.
Figure 1: time series of the V1:ENV_CEB_SEIS_W_50Hz_rms_0.1_1Hz during the three examined periods. In all three cases, the level of ground motion in the CEB is comparable.
Figures 2, 3 and 4: glitchgrams for Hrec during the three examined periods. The bottom time series represent the glitch rates for those triggers with frequency at peak between 20 and 2048 Hz, and SNR > 6.5. One interesting thing to notice is the presence of additional, very loud glitches with frequency at peak of about 90-100 Hz, associated with the last period of bad weather. Investigations are ongoing to understand their origin and if they are related or not to the ground motion.
Figures 5, 6 and 7: same as above but for the V1:LSC_SRCL channel, which in previous entries (#63761 and , #63420) was observed with Scattered Light glitches most correlated with the motion of the EIB, represented by the channel V1:SBE_EIB_GEO_H2_200Hz. In the period April 1-2, almost no glitches are present above 20 Hz, while in the other two periods, the rate is very high and correlated with the ground motion. The fact that a similar reduction in the rate of glitches is not visible in Hrec suggests a different origin for them.
Figure 8: a few Scattered Light glitches present in LSC_SRCL and their prediction with the SBE_EIB_GEO_H2_200Hz channel, as described in #63761.
Figure 9: the Scattered Light visible in Hrec is mostly explainable by the motion of the West End mirror. Using the technique described in #63761, the most correlated channel found is Sa_WE_F0_X_500Hz. In the attached text file, the list of all channels with correlation larger than 20%. The reconstruction is not perfect due to the sparseness and irregular shape of the glitches.