Field Testing, Installation, and Calibration of a new Data Acquisition System for the USGS-Stanford-Berkley Ultra-Low Frequency Electromagnetic (ULFEM) Array

Since 2006, Stanford University, USGS, and UC Berkeley collaboratively maintain five permanent stations, to measure electric and magnetic data from 0.01 to 40Hz. Each station consists of three orthogonal coil magnetometers and two orthogonal 100m electrodes. The acquisition of ULFEM data helps study possible correlations between electromagnetic fields and seismic events related to the San Andreas Fault system. The current data acquisition system uses a Quanterra Q330 analog-to-digital converter. In 2010, we began development of a new 24-bit digitizing system known as the ULFEM 2010 digitizer in order to replace the Q330. The design of the new recorder was to be more economical and better tailored to the ULFEM network by providing power, calibration, and improved protection from lightning. However, the prototype had many problems, including a daily phase shift, amplifying error, and a time delay of 15 seconds (Bowden, et al., AGU, 2010). Currently, comparative testing of an improved prototype, ULFEM 2013, and the Q330 is taking place at the Jasper Ridge ULFEM station. The ULFEM 2013 contains eight channels that record input from three coil magnetometers, four electrodes, and temperature. Testing is ongoing and involves comparing the coil magnetometer and electrode signals processed by the Q330 and ULFEM 2013 digitizer. Data from the two systems will be compared in the time and frequency domains, and analyses will include calculating error and cross correlations. The ULFEM 2013 digitizer provides power to the magnetometer sensors as well as a calibration coil system (CCS). Every 24 hours, the CCS sends a calibration signal to calibration induction coils fitted to each of the three orthogonal magnetometers with the aim of testing the sensors' sensitivity and accuracy. The CCS produces a frequency sweep of 0.08, 0.51, 5, and 10Hz, creating a field nearly ten times greater than the Earth's field. The CCS consists of open source hardware and an amplifying frequency generator. Another ongoing effort to calibrate the ULFEM stations uses ground motion produced by distant earthquakes. Because of our stations' relatively close proximity (approximately 70km apart), teleseismic earthquakes at epicentral distances produce near-identical long-period seismic arrivals at each ULFEM station. The ground motion generated by the surface waves of distant earthquakes causes displacement of the induction coils, inducing magnetic anomalies in the recorded data. Because the ground motion has nearly the same characteristics at each station, the magnetic anomalies observed should have similar amplitudes regardless of the station (though modulated by the local conductivity structure at each site). To identify these coseismic signals, magnetic data were compared against seismic data from each station. Magnetic signals are clearly visible due to passage of Love and Rayleigh waves from teleseismic earthquakes of magnitude >7.4 and their relative amplitudes provide additional confirmation of the stability of our coils and recording system prior to the installation of the CCS.