Jenne, Robert

Movies of test masses showed light modulation that was likely produced by interference with light reflecting from nearby surfaces that were moving on micron scales relative to the test masses. Several such scattered light paths were investigated and a multi-reflection path between the ESD traces on the reaction masses and the HR surface of the test masses was found to be most consistent with microseismic-peak scattering noise in time, peak frequency, and amplitude, suggesting that this noise could be eliminated by driving R0 to minimize relative TM-RM motion (https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=54298). This tracking drive to R0 was implemented at LLO last Tuesday and Jenne implemented it at LHO this Tuesday. Anamaria and Corey tested it at LLO (spectrograms showing improvement here: https://alog.ligo-la.caltech.edu/aLOG/index.php?callRep=50897), and Jenne and I tested it at LHO (https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=54532 and below).

The plot shows

1) that scattering glitches from natural microseismic motion (about 0.5 um RMS of ground motion, X and Y, and 0.8 um RMS of L2-R2 motion at EX) were present in large numbers when the R0 tracking was off, and were not evident when the tracking was on,

2) when the motion was increased with EY ISI injections (to about 5 um RMS of L2-R2 motion at EX- where DARM gets offloaded), glitches were still not evident, and

3) no extra noise was observed when the R0 tracking was on.

I made a similar test the next day with similar results for the injection, but the microseism was lower and there were no large scattering glitches when R0 was off.

Of course this was a short test, and we ask that DetChar keep us informed on longer-term changes in the rate of scattering glitches. I note that detection FARs may be inaccurate until updated to this new era.