I have analyzed the evolution of the rate of the annoying 25-minute glitches over the entire month of January. The criteria to identify this family of glitches among the various omicron triggers is the same as described in logbook entry #62965. Refer to the latter entry for a comparison of the rates during the mini-Engineering Run and those during January.

Figure 1: scatter plot of the distance of one trigger from the next, along with their moving median (orange line) covering 10 points (~4 hours) and aligned to the center of the stride. this thing is important for the considerations associated with Figure 3.

Figure 2: distributions of glitch distances before (blue) and after (red) the decrease of about 0.3 C of the North Input ring heater temperature done between January 26 and 28. The two distributions are clearly shifted, with medians 25 minutes and 3 seconds, and 26 minutes and 41 seconds respectively: -0.3 degC -> + 1'40 in glitch distance.

Figure 3: To answer a question asked by Maddalena at the last Commissioning meeting, I plotted together with the median time series, the temperatures reported by the channels V1:INF_NI_BOTTOM_TE1 (violet) and V1:INF_TCS_NI_RH_TE (green). Notice the upside-down y-axis. Also, notice that these two temperatures have no 24-hour periodicity and that they are shifted by about 8 hours. Refer to Figure 4 for the detail of 2 days of data from the two channels. Going back to the original question, I'm tempted to say that glitch distance median time series is "more correlated" to the temperature of the North Input ring heater, hence the etalon control, than the temperature at the bottom of the North Input tower.