Saturday, October 17, 2009

SciMon Diary: Day 7

So I've reached the halfway point in my scheduled shifts - 7 down and 7 more to go. Today was generally seismically quieter than previous days (maybe the workmen just don't work as hard on a Friday!) and there were no major earthquakes, so we were able to maintain lock for longer and get a better range. We managed to keep in science mode for about half of the shift.

At the start of each shift I have to talk to the outgoing SciMon, look at the electronic logbook (elog) that's kept, and then write an entry detailing the detectors' performance during the shift. We have a variety of figures of merit (FOMs) that are displayed in the control room, examples of which include: a running plot of the range; the detectors state vector (i.e. whether we're in science mode or not); the average power in various frequency bands; the time to which, at the current sensitivity, we could reach the Crab pulsar's spin-down limit; different frequency bands from seismometers located at various parts of the detector (i.e. the central station, or the ends of the two arms of the interferometer); and the glitchiness of the detector. At the end of each shift plots of these FOMs are automatically posted on the elog, and the entry we write basically describes what's shown in the plots, but giving details of things like why we lost lock (which may not be directly obvious from the FOMs), glitches that occurred, and events in the seismic channels. These notes can then either inform the people who work on the detector about potential glitches in the detector system (that they can hopefully fix), or be used if an interesting event is seen in an astrophysical analysis to check that there's no other more mundane reason for the event. Anyway that was a long winded way of me getting round to saying that the automatically generated elog FOM plots were not actually generated before my shift (the script that runs them had accidentally not been restarted following a change), so I had to try and generate my own FOMs to describe what had gone on. I didn't do quite as professional a job with my amateur effort as in the standard plots, but I at least displayed the science segments and range for the previous eight hours.

We again had a school group in the control room today, and again their first questions seemed to be about the two clocks - what is it with school kids and clocks? Have they never seen them before?

My shift ended with us increasing the laser power from the 8W of normal weekday running to the 14W normally used at night and weekends. More power means more sensitivity (see this nice tutorial on laser shot noise for why), but it can also make the detector harder to control and easier to knock out of lock, so that's why the higher power is generally only used when it's seismically quieter. With this higher power we were managing ranges of over 16Mpc. Hopefully that'll continue over the weekend and when I go back in on Sunday we'll have lots of nice data.

Now I just have to prepare myself for tomorrow's rafting trip. Here's a preview of what I'll be doing: