The troubles with the range are due to B1_PD1_Audio saturations (fig 1); high DARM corrections are a consequence (fig 2). I'm not able to find in the ITF something so wrong to justify the observed fluctuations of DARM.

Figure 1, 2 and 3 shows that the B1 PD min/max is changing over a time scale of hours/days by ~50%, but staying about a factor 2 from saturation most of the time. The issue that was happening on Thursday and that we have been attributing to squeezing could be related.

Figure 4 and 5. Looking both today and on Thursday in the 5-10min before the saturation, the line at 1.25Hz is a bit higher on B1, and also the broadband noise between 1Hz and 15Hz in between the lines.

Figure 6 including the data up to 10s before the saturation this is increase becomes more clear.

Yesterday at 10h39-UTC the oncal DAQ was call by the operator for missing images for DET and SQZ cameras . To recover  the situation the following servers were restarted

• on rtpc12: TproImgDET and DET_Img_CEB at 2024-05-18-10h47m55-UTC
• on rtpc7:  TproImgFDS and  FDS_Img at 2024-05-18-10h47m24-UTC

To ensure that the video cameras are correctly received the TelescreenLeft and TelescreenRight servers have been restarted too at  2024-05-18-10h51m28-UTC

A certain correlation between the BNS range and the max level of B1 PD Audio is visible not only when a saturation occurs, but also in condition of good data. In the latest 1e6 seconds, the data with BNS range > 50 Mpc has been selected and a low-pass filter has been applied to the range and to B1 PD1 Audio max, in order to have a cleaner trend. A 2D plot with the two signals (fig 1) shows that the range is never very good when B1 fluctuations are larger. Fig 2 shows the same data in time domain, with B1 rescaled, shifted and changed in sign, in order to have both in a similar scale.

One can guess that the two trends have a common cause: an increased motion of something produces a larger B1 fluctuation at low frequency and a higher noise in the detection band.

A second guess is possible: the larger fluctuation of B1 at low frequency produces itself an excess noise at higher frequency, for some non-linear up-converting effect.

The mechanism behind this variable fluctuation needs to be understood, but also a new DARM control filter, with more gain at low frequency, could be implemented, in order to see if the trend remains or disappear. I also think that there is margin for a general reduction of noise on B1 at low frequency: a large component comes from injected lines (fig 3) and I cannot exclude that they could do a good job also staying at a lower amplitude.

Figure 1 shows the BNS range as a function of UTC hour of the day, for data from May 1-19. There is a decrease of 1-2Mpc during day time. Changes in human induced seismic activity could be the reason for the BNS range decrease and the B1 PD max excursion increase. Human activity (road traffic on bridges, trains, etc) tend to introduce seismic activity at a few Hz, which is also the frequency at which see the increase in B1 fluctuations.

It would be interesting to see if the fluctuations shown by Paolo are also coherent with fluctuating in seismic noise at a few Hz.

Figure 2 and 3 from VIM. Looking at the red trace of ENV_CEB_SEIS_N (bottom right corner), by eye this looks correlated with the B1 saturation problems we had on May 16 and May 18. Most other days the anthropic seismic activity on that plot is below 2e-6 during the day (and lower during the night), while on those two days it exceeded 4e-6. It would be good to confirm that correlation, and understand the origin of that increased anthropic noise.

The correlation between B1 photodiode saturation and anthropogenic seism seems confirmed. A concurrent excess noise is clealy seen by CEB and MCB sensors (Fig.1).

One such event occurred on Thursday afternoon May 16, between about  16:30 and 17:30 LT (Figure 2). Another one occurred on Saturday between about 11:30 to 12:30LT (Fig 3).

Fig. 4 and Fig 5. show that the seismic noise frequency extends from 0.8 Hz up to approx. 10 Hz. More or less the same amplitude is measured in CEB and MCB (note that the MCB "SEIS" sensor measures accelerations: in Fig 4, red spectra of MCB sensors are in velocity units, for comparison with CEB SEIS*). These characteristics point towards a source that is external. By the way, no grass cutting was occurring at those times. In the past noise with somehow similar characteristics was produced for example by excavators: https://logbook.virgo-gw.eu/virgo/?r=48311, https://logbook.virgo-gw.eu/virgo/?r=46729 or tractors https://logbook.virgo-gw.eu/virgo/?r=30059.