[Geysers] Aurum and Doublet Pool

[Michael Frazier]

I spent some time this trip looking for a possible connection between  
the thumping of Doublet Pool (DP) and nearby Aurum geyser. I managed  
to catch 6 Aurum eruptions. Bob Leib and Scott Bryan each gave me one  
more data point. I also managed to find 2 more useable data points  
from when I looked at this in 2003.

Aurum eruptions:

Date, Aurum Start Time, Aurum Start Time from the start of Doublet's  
thumping in minutes (negative is before), Pass or Fail (see text  
below), Data Source (blank is me)

4/24/09	0939	0.5	Pass
4/25/09	0823	2	Pass
4/25/09	1828	-12	Fail		Bob Leib
4/26/09	0831	-1	Pass
4/26/09	1324	2	Pass
4/27/09	1038	-1	Pass
4/28/09	0825	-18	Fail
4/30/09	0824	?	Fail		Scott Bryan - time was several minutes after end  
of DP thumping

 From 2003:
6/13/03	1004	2	Pass
6/13/03	1809	?	Fail		 I didn't have a DP time but my notes say it was  
not thumping

I picked a reference point for DP that was the start of thumps that  
can be felt and heard. That was much easier to time than the beginning  
of DP's cycle which starts with the pool starting to pulse up and  
down. The pulsing is too vague at first to get a consistent time and I  
often was too distracted to notice the start that way. The thumps  
could be felt and heard even over talking so I am pretty sure the  
times are good. The reason I mention this is that it is obvious from  
the data that the Aurum eruptions start from 1 min before DP thumping  
to 2 min after DP a lot of the time. So, I am going to define this as  
the "pass" times and anything else as the "fail" time. Clearly, at 1  
minute before the thumping DP is pulsing strongly so this seems  
plausible. The interval of DP from start to start was from 34 to 37  
minutes with a duration of 7 to 9 minutes for the thumping. After 2  
minutes into the thumping Aurum was never triggered in the data I saw.  
Looking at the data, you can see that there are 6 pass and 3 fail.  
Again, 6 times between -1 and +2 minutes and 3 times during the  
remainder of the approximately 35 minute interval.

I was concerned that with so few data points it was hard to tell if  
this was just a coincidence or something real. It is easy to believe  
it was something real if it happened all the time. As you can see from  
the data, it only happens some of the time. If the two events were  
completely random you would expect the chance of Aurum erupting in  
that 3 minute window out of the 35 minute interval as 3/35 or 0.086.  
That seems pretty unlikely, but how does the actual data compare to a  
totally random set of events?

To try to get a feel for what you would expect to see from a totally  
random set of data I wrote a simple program that generated a set of  
random times and graded them on whether they fell in the 3 minute  
window or not. To compare against the actual data I generated 9  
uniform random times between -18 and +17 (35 minute interval). If the  
numbers were between -1 and +2 I called that pass and any other time I  
called that fail. The results were pretty surprising. Only about 1 out  
of 40,000 tries got data that had 6 or more "pass" events with 9  
random times.

In conclusion, the data is very unlikely to be represented by two  
random events and there probably is a connection between DP and Aurum.  
Does this information have any practical use? Maybe. If you are  
watching Aurum and have to leave the hill for a few minutes, it's  
probably best to not leave near the start of DP thumping. :)

Mike Frazier