Earthquake Precursors
2006 Worlshop on Earthquake Precursors
2007 Worlshop on Earthquake Precursors
2010 Workshop on Earthquake Precursors

Predicting earthquakes seems impossible.  Many have looked for ways to predict earthquakes but no one has yet successfully predicted accurately the time and place of a future earthquake.  However, this doesn't mean we shouldn't try.   We just need to be very careful before we make any claims about predictions.  

A number of year ago a colleague of mine (Prof. Tony Fraser-Smith) of Stanford University accidently detected anomalous electromagnetic signals in the ultra low frequency (ULF: f<1 Hz)  range two weeks prior to the strong Loma Prieta (California) earthquake in 1989.  Tony was not looking for earthquake signals, and his magnetic induction coils were being used for another research project.  However, the instruments were only 7km from the epicenter (very close), and the magnetometers started observing high levels of ULF radiation weeks before the large earthquake that occurred on 18 October (see figure below).  The ULF anomalies (high magnetic field intensity) continued after the earthquake itself due to aftershocks, and eventually returned to the background levels months later (see image).
loma

Two weeks is a long time for a prediction!  Could this really be a precursory signal related to the earthquake?  Others started to look in other regions of the world (Russia, Japan, Taiwan,...) often with similar success.  However, the detectors needed to be within 100-200 km of the epicenter.  Not knowing where the next earthquake will hit makes this a little difficult for research purposes.  We also need large earthquakes to test our models and hypothesis regarding the mechanisms that may caues these ULF precursory signals.  These large earthquakes are few and far between.  Nevertheless, we decided to start looking for such ULF precursors in Israel, and more particularly along the Dead Sea Rift Valley.  This region is a very active seismic region (see plot below left) although the magnitude of the earthquakes are not very large.  In the plot below the earthquakes over the last century are shown for the region.  There is a large concentration along the Dead Sea rift valley and into the Red Sea.  Our field sites are located in a chain from Timna in the south to the Hermon in the north, with a middle station near the Dead Sea. 

israelquakes

We started this project in June 2003 collecting ULF data (0.01 Hz) using three magnetic induction coils (Hx, Hy and Hz).  The natural ULF magnetic signals are a result of fluctuations of the Earth's magnetic field (pulsations) and during high solar activity  (geomagnetic storms) the ground ULF signals are greatly enhanced.  We therefore need to understand both the natural magnetospheric ULF fields as well as the local fields (perhaps seismic).  This is not a trivial task and involves a lot of data analysis and understanding of the physical processes that influence the observed signals.  We now use three different indices to track the background ULF signals, in order to study changes before, during and after earthquakes.  Under normal circumstances these indices should be the same at all our stations (see below comparison between 2 stations separated 200km from each other).   These indices are the "polarisation" index, the "fractal" index, and the "Vd" index.  We expect (or hope) that seismic activity close to one station will influence these indices before the earthquake occurs.  Below is shown some of our first measurements for a short period showing the three indices at two of our sites.  We continue to monitor continuously.



References:

Zomer, A., C. Price, L. Alperovich, M. Finkelstein and M. Merzer, 2008:  ULF amplitude observations at the dawn/dusk terminators, J. Atmos. Elect., 28(1), 2-29.

Related websites:

NASA article
Radio waves below 22kHz