A Transportable System for Monitoring Ultralow Frequency Electromagnetic Signals Associated with Earthquakes

Claims that electromagnetic (EM) signals occur associated with some earthquakes, typically prior to but sometimes during seismic activity, have appeared in the literature for several decades (Parrot and Johnston, 1989; Park, et al., 1993; Park, 1996; Johnston, 1997). Such claims cover an exceptionally broad range of phenomena. For example, Gokhberg et al. (1982) and Yoshino (1991) observed increases in signal amplitude at 81 kHz from minutes to hours prior to earthquakes and up to hundreds of kilometers away from epicenters, and attributed these signals to seismoelectric emissions. Seismoelectric signals (SES) of a different nature have been suggested to precede some earthquakes in Greece (Varotsos et al., 1993a; Varotsos et al., 1993b). These SES are transients of amplitudes of 20 mv/km and durations of several minutes recorded on multiple dipoles of different length, and their mechanism of generation is subject to debate (Lighthill, 1996; Pham et al., 1999). Reported claims of EM anomalies associated with earthquakes also extend over a large frequency range, from megahertz down to quasi-dc. At the low end of the frequency range, Johnston and Mueller (1987) observed magnetic field offsets coinciding with the 1986 North Palm Springs earthquake, which occurred in Southern California close to the San Andreas Fault, and Johnston et al. (1994) also observed offsets at the time of the 1992 Landers earthquake in the same region. At the high-frequency end, radio emissions at 18 MHz were recorded on multiple northern hemisphere receivers for about fifteen minutes before the 1960 great Chilean earthquake (Warwick et al., 1982). In the ULF (0.01-10 Hz) frequency range, Fraser-Smith et al. (1990) recorded anomalous magnetic field fluctuations prior to the 17 October 1989 Loma Prieta Ms = 7.1 earthquake in central California. In particular, there was an amplitude increase in activity about …