Last Updated on July 22, 2025 by The Editor
(Reposted from the original on QVSData)
There is much talk going round about the fact that earthquakes are getting deleted from the listings for Yellowstone. Is this a conspiracy? Are they trying to hide something from us?
The short answer is NO, they are not.
Having, at least as far as I am concerned, found a satisfactory explanation for the recurrence 47 times of the “set of 4” which was caused by two servers being out of synch, I would now like to offer an explanation as to why these earthquakes get added as ‘false positives’ in the first place.
This all relates to the ‘false’ 2.4 at 14:48 UTC (8:48 MST) on 08 Feb 2011, but exactly the same applies to the later ones.
This is the quake in question:
Date/Time UTC,Latitude,Longitude,Magnitude,Depth(Km),Location
2011-02-08 14:48,44.67310,-111.18000,2.4000,10.0000,western Montana
First take a look at these three slices of seismograph display from 3 recorders (Courtesy of the YVO)
Note that there is a gap in transmission on all three recorders at the same approximate time.
When the signal dies the computer is looking at zero. Why does the signal die? This is the telemetry problem that I have spoken about. Ice and snow build up on the equipment causes these.
The Yellowstone Volcano Observatory replied to me in an email:
As indicated earlier, the problem is ice and snow build up at the main telemetry node where all the data are gathered and put on a microwave tower for transport from the Henrys Fork area (outside the park) down to Salt Lake City.
Last year around the same time we had problems with telemetry in the park which again it turned out were due to ice build-ups on the equipment. This is the same phenomenon that gives, as an example, anomalous GPS readings in Iceland from time to time. It is a common thing in cold weather and climates.
The trace below is a section from the MCID station for which I downloaded the SAC waveform file.
The circles are telemetry spikes, and if you dive in deeper you can see more of them.
On the next one I have zoomed into the trace a bit, around the time of the FP and you can see myriad small spikes. (Some of these are telemetry errors and some are parts of the waveform.)
If you are a computer and you are looking at zero signal, and then it pops back up when the signal returns, this can be interpreted as an earthquake. This would especially be the case if the signal was reasonably high (wind, buffalo farts etc) when it came back on.
The YVO said as regards these:
Moreover, since the data problems differ slightly for each station, that results in a very “locatable” phenomenon for the computer. It would be possible to filter out this “bad behavior”, but would result in slower response time (a no-no for seismic networks) and possibly some missed earthquakes.
It is quite clear from my records of deleted quakes that these happen much more frequently in the winter months, and to a degree as an inevitable consequence of a large swarm. (I keep records of all deleted USGS, EMSC and GeoNet earthquakes.)
In addition the YVO also said:
Eventually, when the park receives another T1 line at Mount Washburn, we will get everything on the internet directly, and the system will improve (and be more redundant to problems).
(A T1 line is a broadband internet connection for those who are not aware.)
A most important fact has come out of this that what we see is not the detailed scientific record, not that I suspected it was.
YVO:
If you look at the public webicorders for 2/8, you won’t see the small data gaps below, but you will clearly see that there are all sorts of spikes from telemetry problems, especially on YMC and MCID, two of the stations that caused the earthquake to be picked.
I have no problems with this as a very reasonable explanation.