October 19, 2005
By Stan Deyo


Many folks have asked for more detailed instructions on how to interpret my daily seismic/volcanic/ storm forecast maps. Hopefully the following will suffice in most cases.

HOW THEY ENABLE ME TO FORECAST QUAKES:
My seismic forecast maps are derived by finding the rate or speed at which sea surface temperatures vary from day to day over known fault lines.

As best I can hypothesize, the rapid temperature changes which precede the major seismic events under water or on nearby land are caused by heat being transferred from one side of the fault (under rupturing stress) to the opposite side. Pressures on crystalline structures like mantle rocks often generate piezoelectric charges which migrate up into the water and then into the moist air above the water and then back down into the water on the other side of the stress line. Since the moist air is a semiconductor I believe the transfer of heat is a form of the Peltier Effect - or, possibly, the Thomson Effect.

(The Peltier Effect: In 1834 the French physicist Jean C. A. Peltier discovered if a current passes through a thermocouple, the temperature of one junction increases and the temperature of the other decreases, so that heat is transferred from one junction to the other.)

(The Thomson Effect: The Scottish scientist William Thomson discovered in 1854 that if a temperature difference exists between any two points of a current-carrying conductor, heat is either evolved or absorbed depending upon the material)

HOW I DISCOVERED THE METHOD:
In 1995 the U.S. Navy at the F.N.M.O.C. in Monterey, California started publishing daily OTIS maps (now called NCODA) on the Internet. They showed daily anomalies in sea surface temperatures when compared to a 30-day, running average of sea surface temperatures. 

I discovered that I could use their data to determine where seismic and, sometimes, volcanic events would occur. I also found that certain severe electrical storms would produce a similar type of pattern to the seismic ones on my maps. The trick was how to separate the seismic/volcanic signals from the severe storm signals.

I found that a certain type of pattern appeared over known fault lines and tectonic plate boundaries 1-5 days before a significant (Richter5.0+) seismic event or a major volcanic eruption. The pattern is one in which a yellow spot is on one side of the fault line and a blue spot is on the opposite side of the fault line directly across from the yellow one. The yellow spot indicate a rapid heating while the blue ones represent a rapid cooling. I approximate the strength of the quake by the area of the signal multiplied by the area of the brightest portion of the wings.

THE "BUTTERFLY SIGNAL":
I have nicknamed these seismic signal patterns, "Butterfly Signals" because they are somewhat like the extended wings of a butterfly with its body positioned along a fault line.

When I see extremely bright "butterfly signals" over a large area I know that a major quake is imminent - usually within 3 days and often within 24 hours. However, when either one large or several smaller butterfly signals appear over a long fault line (say, 2000 miles) I estimate that a large seismic event (or events) may be 1-2 weeks away - possibly more. However, as such signals do not occur very often, I am uncertain how much lead time we get. I have only seen two of these in nine years and one produced the Great Sumatran Quake of 2004, which I warned about beginning December 22, 2004 as well as on radio broadcasts. The other did not produce any seismic events we could correlate.

FIRST MAP:
I produce two maps. The first map shows the continental (greens and browns) and sub-oceanic (shades of aqua) topography, the "base" map (yellows and aquas) and the tectonic plate boundaries (in yellow). The current day's warning circles are marked in white on this map for visual clarity. I place the circles at the center of the butterfly signal and extend their radius (sometimes radii) to the outer limits of the wings. 

SECOND MAP:
The second map shows the continents in black outlined in white. It displays the "base" map (yellows and aquas) combined with the actual sea surface temperature maps (shades of the rainbow spectrum from dark red to yellow to green, to aqua to dark blue) extending away from the equator and the tectonic plate boundaries (in yellow). This actual temperature spectrum is included to show where the shear boundaries are between the distinct thermal bands. It allows me to see where certain storm signals occur so that I can separate them from the butterfly signals. Also on this map are my warning circles of the previous 4 days (shades of red - the most recent - to light pink the oldest) including the current day's white circles.

This map is made to reveal which areas of the planet are showing consistent trends in butterfly signals or movements of a butterfly signal over the last week. Not all butterfly signals yield a seismic event - although all butterfly signals indicate increased strain on a fault line. It is like a weather forecaster saying, "It will be cloudy tomorrow but it may or may not rain; so take your umbrella." This map is not included in the archived maps.

ACCURACY:
I estimate my accuracy of forecasting by the ratio of the number of major seismic events within or close to oceans divided into the number of those major seismic events that fell within my warning signals of the last 5 days and multiplied by the error compensation rate (70-80%) for the Navy data…. (the latter being given to me by the Navy personnel at the FNMOC).

For those who are mathematically inclined, the formula would be:

.7(EF/ET) <= A <= .8 (EF/ET) 
where:
A = accuracy,
EF = number of major seismic events that fell within my warning signals
ET = number of major seismic events within or close to oceans.
So, with an EF/ET = 90%, 
63% <= A <= 72%

—Stan Deyo