Referring to PeterK's log, 51658, while the Supermoon will change the Earth tides, these effects will be a few percent not a factor of four.  Nor will these normal tidal forces produce a discontinuity as seen in JeffBs' plots in logs 51648 and 51650, and herein.  Even if the IFO were to remain locked for an entire year allowing the tertiary affects (seasonal) to 'drift' the DC level of the tides, the total range on the system would be less than +- 400u and certainly not 700u in 30 hours.  This world would be in a world of hurt if that were the case.  This extreme excursion of demand is causally related to the PSL incursion on Thursday.  And given that this caused a lock-loss, warrants an FRS.  Possibly the operation of or procedure of running the AC in the PSL area needs adjusting.  Alternatively, the locking of the RefCav with a 10-20% step in power may be the culprit.

I suspect that the temperature excursion of the PSL room while not getting to the PSL RefCav temperature sensor, the laser table is warmed up and this slowly flows into the reference cavity via its legs.  Alternatively, I see that the PSL-FSS_TPD_DC output jumps 10+% up after the Thursday activity.  I see the RFPD of the FSS also steps up so something upstream increased laser power.  Could this increase of power just be warming and increasing the RefCav length--seems plausible.  The attached plot shows the step and very slow trend back down of the FSS Slow loop (Ch3) which is putting a wavelength change on the 1u light.

So, I see the room temperature change and the increase RefCav power warming and increasing the length of the RefCav evidenced in the slow loop control. Could this be mitigated to reduce risk of lockloss? Maybe. Change AC operation, monitor & shunt 'excess' power going to RefCav to keep it heat load the same.

Attached plot shows 8 days of LASER room temp, 'tidal' drive to EX HEPI (EY similar,) FSS_NPRO_TEMP, and the FSS_TPD_DC.