Abstract
Efforts to detect magnetic precursors to earthquakes are often complicated by the presence of man-made noise. The problem is especially severe in urban areas where both the levels of interference and the potential damage due to an earthquake are the greatest. In this talk I will describe the results of an ongoing project to monitor ultra-low frequency (ULF; 0.01 to 10 Hz) magnetic fields along the Hayward Fault in the San Francisco Bay Area of California. The acquired data are almost completely contaminated by magnetic interference generated by the Bay Area Rapid Transit (BART) system. I will use both measurements from an array of multi-component sensors and electromagnetic modeling results to characterize the interference. I will then present a combined space-time approach for the identification and removal of the interference. A key component of the approach is a recently developed undecimated discrete wavelet transform based method for transient detection in the presence of 1/f noise.