The heightened microseism has been messing with our seismic earthquake automation lately. First attached plot shows corner station ground right now, and the useism has slid down enough in frequency that it is bleeding into the band that peakmon uses for the detection of incoming earthquakes. One part of the peakmon code is the band pass filter used in the second attached plot, which I have been quickly working on a slight modification of. Red is the current set of filters, blue is what I would like to try overnight to see if it helps. I haven't had a chance to do any real analysis on this yet, but if the seismic automation starts misbehaving we can just pause or revert my filter change. Worst I can see this doing is I would need to maybe rethink the peakmon transition thresholds in SEI_ENV, this would tend to scale peakmon down for a given ground input, so we would transition later or only on larger eqs than the nominal filter. I should also check to see what LLO is using, I don't remember.
It seems like this bandpass is working like it should, attached trend shows peakmon before and after. Top plot is peakmon, with the white cross-hairs intersecting with the time I switched to the new filter and 900nm/s. Bottom plot is the microseism over this before/after window. The microseism stayed high after turning it on, even climbing a bit higher and peakmon stayed under 900nm/s for the rest of this window. I still need to look at some data and make sure this won't cause us to miss earthquake transitions but I think this is still probably an improvement.
