With the lower noise seen in CARM, we can now see clear coherence of some higher frequency lines with DARM. The frequencies are centered at 556, 574, 835, 860, 1115 and 1148 Hz, but are not very high Q. The third set of twin lines looks like a straight harmonics of the first set, whereas the second set is 1.5 times the frequencies of the first set. This would indicate fundamentals at 278 and 287 Hz, which are barely visible. There is also a small indication of the 5th harmonics at 1393 and 1435 Hz.
This looks pretty bad on 1800 s DARM averaged SFTs (see four figures attached). From a CW/Stochastic perspective these things are not lines, they are huge bumps in the spectrum.
As can be seen here: https://ldas-jobs.ligo-wa.caltech.edu/~keithr/O3spectra_DARM/#
the bumps in DARM at ~270 Hz started to be noticeable the 17th of April. The bumps at the other three frequencies (500, 800 and 1100 Hz) only are noticeable since 1 of May, so I guess that something changed that day which made the noise or the coupling way worse.
I checked the magnetometer channels in order to find coherence as was done here (https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=47644). I found coherence only with some magnetometers in the corner station, mostly EBAY_LSCRACK, LVEA_OUTPUTOPTICS and LVEA_VERTEX, with INPUTOPTICS showing less coherence and SUSRACK no coherence at all at the frequencies where these bumps exist
The attached figure shows a comparison between the 30 April and today, showing how the coherence for LSCRACK MAG has increased
The coherence between PRCL and DARM is likely associated with ISI table suspension resonances (violin modes -like we damped in some HAMs). The GS13s are coherent with DARM at these frequencies (first page of the figure).
The excitation of the ISI tables does not appear to be driven from the ground, though, because the GS13s are coherent between distant chambers, like BS and HAM5, while the ground motion is not coherent at these frequencies (second page of the figure). Could there be a back-reaction from global control at these frequencies? Supporting this interpretation is the observation that there is a lot more coherence betweeen GS13s on the BS table and on HAM5 than with the GS13s on HAM6.