Upon Irene's request, I had a closer look at the situation of scattered-light glitches. For this study I used the data of the last 24h when the whether conditions are not good (see low frequency glitches in plot 1).

I ran UPV to find correlation with auxiliary channels. There are clearly (at least) 2 families of scattered-light glitches:

• The first category results from a coupling at MICH level as all "central" channels (B2/B4/BS...) witness these glitches. MICH noise subtraction of HREC does a good job as many glitches are removed. However, there is still a significant amount of these glitches visible in h(t). UPV reports that the best channel to flag these glitches is LSC_BS_CORR (see plot 2).
• The second family is located at the B8 level. The sea activity is probably responsible for those. These glitches are less frequent but they are more powerful (see plot 3).

In both cases we see the "beautiful signature" of scattered-light glitches in spectrograms with a typical arch-shape structure: see plot 4 and 5.

Also note that the line of glitches at higher frequency on plot 1 are also due to a high micro-seimic activity coupling to the beam alignment.