I was digging around recently looking for some old geyser data when I came across something that might be of general interest.
Back in 1993 I was working with Rick Hutchinson on recording geyser activity. He let me use some of the TempMentor devices, and I wrote a Mac application so I could access and download and display the data. The TempMentor that I used was inside a self-contained cylinder, about a foot high and about four inches across that provided a waterproof place for the recorder. The sensor itself was in the bottom of the cylinder, so I had to place it directly in the runoff channel of the feature I wanted to record. Fortunately, even back then there was enough space under the boardwalks that I could easily place it, even during the daytime. (When the boardwalks were rebuilt, they were raised, mostly to prevent them from continuing to provide marmot and tree-rat habitat.)
I placed the devices at Grand, West Triplet, and Plume. I also have Castle data, but I think that was from a device Rick placed, as even back then we were not encouraged to go that far off trail.
I still have the data I collected, but it's in a format that I would have to reconstruct. Fortunately, part of the program that I wrote did some processing to look for eruptions. That's what I came across, and it's interesting to look at now, 28 years later. I've attached the text files to this posting, as they are a bit too big to just add as is.
Plume runs from 09 Jul 1993 through 12 Aug 1993, without any gaps in coverage. During that time not only was Plume "going to sleep" but there were also a couple of Giantess eruptions. So the data shows not only the overnight pauses in activity, but the shortened intervals in response to Giantess' eruption. Most of the non-Giantess eruption intervals were around an hour, but in the 36 or so hours afterward, that interval dropped in half.
Most of the time Plume went to sleep around midnight, and woke up around noon, so many of the sleeps were in the twelve hour range, But there are also some that lasted the better part of a day (19 hours in one case). The recovery from Giantess mode was pretty abrupt, too. In one case it there were just four transition intervals before it was back into the hour range. In the other case, there were just three, and then it went to sleep two more intervals later.
West Triplet was also interesting. I moved the Plume recorder there so there's a record from 13 Aug 1993 through 15 Sep 1993. I noted the Rift times in the file, but not sure how/where I got them. Sam Martinez was also doing some recording, and he may have been the source. In any case, the recordings clearly show that the West Triplet eruptions that preceded Rift lasted longer and Rift started about 1/2 hour in (at the time West Triplet would have normally quit.)
Grand activity is in several segments. The first runs from 09 Apr 1993 through 10 Jun 1993, while the second is from 04 Jul 1993 through 15 Sep 1993. The shortest intervals were just under seven hours, while there were no intervals over 12 hours. For back then, that was nice, consistent behavior for it.
Finally, Castle runs from 21 Aug 1993 through 11 Sep 1993. During that time most intervals were eleven hours, give or take. There were one to three minor eruptions a week, and several major eruption intervals between them. The intervals after the minor eruption varied from three to seven hours, and the major eruption interval after that was about an hour longer than usual.