-
New Location Near the Yellowstone Caldera
01/06/2020 at 06:36 • 0 commentsI have placed a machine in the Madison Valley of Montana! It is near the Yellowstone caldera and is putting out data. I'll update with more information soon!
-
Yun program now added to downloaded files
09/15/2017 at 14:24 • 0 commentsI just added the "noisemachine.ino" program to the files section, so that the entire project is completely available to anyone who wants to replicate it. I also added noisemachine.txt so that the program can be viewed without the arduino IDE.
-
Mexico Earthquake last week
09/14/2017 at 11:53 • 0 commentsPeople have been asking me if my device registered the magnitude 8+ earthquake that happened near Guatemala, off the West coast of southern Mexico on September 8. There were at least 96 fatalities and thousands were left homeless. Here is a picture to keep a focus on the human cost of these events. Note the total collapse of this residence. It is not clear from the news report whether anyone died in this house. There was no warning whatsoever before this earthquake from traditional seismometer monitoring stations.
Detection of distant quakes is not what this device is for, but the "big one" registered very well on our machine here in Knoxville. The peak is about half full scale on our machine. Here is the vector magnitude data - this quake was about 1600 miles from the Knoxville machine.
Here's the probability statistic data, log scale, running averaged -
Here's the vector location data (radians ) pointing at the quake toward the Southwest. I am not sure, but the later vibrations appearing to come from the Northeast may be S-waves (transverse waves) travelling at about half the speed of the initial P-waves (compression waves). This is a nice illustration of the fact that transverse waves vibrate at a 90 degree angle to the compression waves.
-
Another small quake 413 miles away
09/14/2017 at 00:57 • 0 commentsA minor earthquake occurred on Wed Sep 13, 2017 at about 12:33:10 CDT (Sep 13, 2017 17:33:10 UTC) 11.43 km northeast of Peterstown, WV. The magnitude was 3.1.
Our main machine recorded this event 23 seconds later. The vector magnitude plot for that hour is:
The brief event early in the hour appears to be something very local. It showed up the local USGS affiliated "strong motion" machine 7 miles or so away even though the West Virginia event did not show up well (see below). Here is an isolated view of the vector magnitude data from the West Virginia event:
There appears to be no advance vibration from this distant event whatsoever. Here is the combined probability statistic data from the entire hour.
This shows a substantial deviation from normal statistics for about 1000 seconds after the main front of the West Virginia tremor, a hint of which can be seen in the actual vector magnitude data. I speculate that the deviation from normal statistics may be due to echoes from the nearby Smoky mountains to our South and from the Appalachian mountains farther away to the East.
This graph shows the horizontal location in radians of the data points for the hour, indicating a persistent Eastward trend in the average location of the noise for about 1000 seconds:
Here is the data from the Knoxville USGS machine, by way of comparison. It records (in black) the event early in the hour that showed up on my machine, but there is only the barest hint of the West Virginia event at 1:33 PM EST. Here is the data from all of this afternoon on that machine:
Here is a cutout view of this data showing the local early event (in black) and the barely perceptible West Virginia event. I added the time labels.
The take home message that seems to be evolving is that distant events are not associated with detectable anomalies of the noise before the event, but local events do seem to be associated with anomalies. More data points (and lots of them!) are needed.
-
Nearby Tremor in Knoxville
08/25/2017 at 21:40 • 0 commentsAbout an hour ago, a small 2.5 magnitude tremor event happened several miles to the Northwest of our machine. I felt and heard it - it seemed like thunder some distance away. The rumbling persisted for quite a while. Here is the recording from the local USGS machine - there was no significant activity prior to the event on the local machines.
Here is the location of the event, according to the University of Memphis earthquake center, in relation to our machine - I have added text to the Memphis center image to indicate the two locations.
Here is the vector intensity data from our machine - X axis is in seconds.
Here is a close up view of the seismic noise during the 500 seconds before the main tremor.
Here is the apparent location data, in radians, including the tremor itself - it starts at 500 on the X axis. Positive values are westerly, consistent with the Northwest location of the quake from our machine. Because of the processing limitations of the Arduino Yun, the location calculations are less sensitive to tremors arriving from the North and South.
Here is the log scale of the combined probability data, without averaging. The combined probability is the product of the probability of the location data and the vector magnitude data. There are definite statistical aberrations noted in the minutes preceding the main event. The lowest probability point in the precursor data is in the one-in-a-billion range! Because of the size of the initial data set after start up, the minimum possible combined log probability is -10.3. This explains the bottoming out of the data during the actual earthquake.
Here is the same data, with running averaging.
Once again, VERY interesting data. We need to profile dozens of tremors before any conclusions can be drawn about the statistical prediction of local earthquakes, but this is another confirmatory event that seems to say we are on the right track! The placement of a machine to the west of Yellowstone park will happen sometime in late fall if everything goes well.
-
Adding a little cool factor
07/11/2017 at 02:44 • 1 commentEarthquakes don't happen very often in East Tennessee, so while waiting to get a device installed near Yellowstone I have a little extra time on my hands at this stage in the project. Just to add a little somethin' to my "indoor" seismometer, I bought a inexpensive little solid state ultraviolet laser module and an inexpensive laser control module on eBay, Total cost about $20.
Fluorite contains yttrium, europium and samarium in a calcium fluoride matrix. The first three elements absorb ultraviolet light, hold on to the energy for a while and emit it later as visible light. This type of photoluminescence is a combination of fluorescence and phosphorescence. The entire ball glows nicely when illuminated from below by the laser through a small hole in the seismometer base - the laser light itself is invisible.
It's aliiiive!
The Arduino code to make it dim and brighten is bone simple:
******************************************************************************************************
int laserPin = 5;
void setup() {
pinMode (laserPin, OUTPUT); //actual pin is 5
pinMode(6, OUTPUT); //we will use pin 6 as a ground for the electrodes in digital pin 5
digitalWrite(6, LOW); //pin 6 is used as ground for the electrodes, so the program never changes this.
}
void loop() {
int x = 1;
for (int i = 0; i > -1; i = i + x) {
digitalWrite(laserPin, HIGH);
delayMicroseconds(i * 20); // vary the brightness of the fluorite ball
// with this number; 20 is "subtle" in the daytime
digitalWrite(laserPin, LOW);
if (i == 255) x = -1; // switch direction at peak
delay(10);
}
}
********************************************************************************************************
The code can be added on to the main seismic noise program as well, so that the ball also changes intensity in response to vibrations in the home, but that delay(10) statement makes it a time hog. It is best to have a separate Arduino Uno controller for the laser and the Yun for the seismic noise device.
-
Aberrant noise detected 32 minutes before local quake
06/25/2017 at 21:05 • 2 commentsThis morning at 10:00:57 UTC (6:00:57 Eastern), a magnitude 2.6 tremor originating near Lenoir City, TN was recorded on all regional seismometers in the USGS networks.
Here is the USGS summary page of that event.
This event was 21 km in depth and was approximately 20 miles from our seismic noise device. Here is the signal as recorded on the Knoxville strong motion machine - a few miles from here. Note the absence of any signal before or after the event.
Here is the event as recorded on our machine - the data shown is from 4:00 to 7:00 AM. with data points every 1000 milliseconds. The x-axis is number of seconds from 4 AM and the Y axis is vector magnitude units (in machine voltage units, not the magnitude units used to refer to earthquake strength). This data is from the <15 Hz device. Unfortunately, the low frequency machine was not recording data during this event due to a bad clock module.
32 minutes before the main event, small increases in noise magnitude, associated with directional anomalies were noted. After-tremors were recorded on our device which did not show up on the Knoxville machine, but which did show up on some of the more sensitive regional machines.
Here is a zoomed in view of a scatter plot of the vector magnitude data from the aberrant noise precursor before the main event. Once above 33 units, the data is outside the normal bell curve of the noise.
Here is a line plot of the 5 second running average of the logarithm of the combined (magnitude and location) probability parameter. There is strong evidence of a prolonged statistical anomaly beginning about 32 minutes before the main event (red arrow), although of course it cannot be certain that the anomaly is actually related to the following event.
Here is the same data without the running averaging.
Notice that the lowest point in the precursor is around 10^-7. In other words, there is about a one in ten million chance that the magnitude and direction of that time point's data is part of the usual seismic noise. This is not even taking into account the tremendous improbability of having a cluster of "unlikely" measurements closely related to one another in a short time period. One of the time points for the actual quake reaches a peak at about a one in 10 billion chance.
These numbers are all the more impressive if one takes into account the meaning of the logarithmic data. If 50-50 odds was a line one inch long as on this graph, one in 10 million odds would be a line 158 miles long and one in 10 billion odds would be a line 158,000 miles long. That's almost a light second!
Seismic events are happening across the globe constantly. These far away events are what make up the majority of seismic noise. It's clear that some kind of cluster analysis will be helpful to deal with random and isolated (one point only) deviations from normal statistics and will allow a machine to identify local precursor events with extreme clarity. Far-away (weak intensity and brief apparent duration) seismic events would be experienced almost equally by the individual members of a local network of machines, while the machines will give very different responses to nearby precursor events.
A central computer responding to data from a local network will then need to quantify the heterogeneity of the probability data coming in from its individual machines in response to weak intensity signals. If the local devices are seeing a weak (low amplitude) but real ( a cluster of very low probability time points) event in widely different ways, it then can assume that a local seismic event may be happening.
Then decide whether to issue a warning.
-
Much less seismic noise intensity at lower frequencies
05/18/2017 at 00:14 • 0 commentsThe overall vector magnitude of the seismic noise decreased by about 90% when the 1 Hz filter was substituted for the 15 Hz filter. Very interestingly, the magnitude distribution is much more gaussian than the output from the 15 Hz device, as well. Will this be good or bad for what the device is intended to do?
I have increased the amplification of the second stage of the amplifier by a factor of 2 to partially compensate for the diminished noise intensity. The overall signal response to small local events seems to have decreased, but there is expected to be no change in the probability data output. Too soon to tell.The software variables are still being tweaked to deal with the new arrangement. It will be interesting to see how the data from the 1 Hz and 15 Hz devices compare while running side by side! Hopefully I will have some data to post in the next week or two.
-
Looking at lower frequencies
05/13/2017 at 19:51 • 0 commentsThe charge amplifier on one of the seismometers has been changed to respond to 1.5 Hz and below. It was a simple resistor change, of the rightmost resistor on the handwritten schematic down below. The resistor was 1000 Ohms and now it is 10,000 Ohms.
Up until now the amplifier has been using a 15 Hz low pass filter. It's going to take a couple of days to collect enough data to get the device's sensors calibrated. The plan is to run the 1.5 Hz device and a 15 Hz device side by side until "something seismic" happens, then compare and contrast the results.
-
Thanks for the prize!
05/08/2017 at 22:40 • 0 commentsI just found out my project was selected to be one of the prizewinners in this round.
Wow! I am humbled by the other winners and proud to be included with them. There were many awesome projects that were not selected, as well. I hope that these people will not get discouraged. Hang in there and make your project better and better!
I hope that this recognition will bring more "lookers" who may be interested in working together with me to create networked seismic noise monitoring systems in seismically active regions.
This thousand dollars is going to be given to Doctors Without Borders because of their fantastic record of dealing with earthquake emergencies.
Thanks!