A new technique designed to automatically identify and characterize waves in three‐axis data is presented, which can be applied in a variety of settings, including triaxial ground‐magnetometer data or satellite wave data (particularly when transformed to a field‐aligned coordinate system). This technique is demonstrated on a single Pc1 event recorded on a triaxial search coil magnetometer in Parkfield, California (35.945°,−120.542°), and then applied to a 6‐month period between 1 June 2003 and 31 December 2003. The technique begins with the creation of a standard dynamic spectrogram and consists of three steps: (1) for every column of the spectrogram (which represents the spectral content of a short period in the time series), spectral peaks are identified whose power content significantly exceeds the ambient noise; (2) the series of spectral peaks from step 1 are grouped into continuous blocks representing discrete wave events using a “spectral‐overlap” criterion; and (3) for each identified event, wave parameters (e.g., wave normal angles, polarization ratio) are calculated which can be used to check the continuity of individual identified wave events or to further filter wave events (e.g., by polarization ratio).

Abstract. Several electromagnetic signal types were observed prior to and immediately after 30 October 2007 (Local Time) M5.4 earthquake at Alum Rock, Ca with an epicenter ∼15 km NE of San Jose Ca. The area where this event occurred had been monitored since November 2005 by a QuakeFinder magnetometer site, unit 609, 2 km from the epicenter. This instrument is one of 53 stations of the QuakeFinder (QF) California Magnetometer NetworkCalMagNet. This station included an ultra low frequency (ULF) 3-axis induction magnetometer, a simple air conductivity sensor to measure relative airborne ion concentrations, and a geophone to identify the arrival of the P-wave from an earthquake. Similar in frequency content to the increased ULF activity reported two weeks prior to the Loma Prieta M7.0 quake in 1989 (Fraser-Smith, 1990, 1991, the QF station detected activity in the 0.01-12 Hz bands, but it consisted of an increasing number of short duration (1 to 30 s duration) pulsations. The pulsations peaked around 13 days prior to the event. The amplitudes of the pulses were strong, (3-20 nT), compared to the average ambient noise at the site, (10-250 pT), which included a component arising from the Bay Area Rapid Transit (BART) operations. The QF station also detected different pulse shapes, e.g. negative or positive only polarity, with some pulses including a combination of positive and negative. Typical pulse counts over the previous year ranged from 0-15 per day, while the count rose to 176 (east-west channel) on 17 October, 13 days prior to the quake. The air conductivity sensor saturated for over 14 h during the night and morning prior to the quake, which occurred at 20:29 LT. Anomalous IR signatures were also Correspondence to: T. Bleier (tbleier@quakefinder.com) observed in the general area, within 50 km of the epicenter, during the 2 weeks prior to the quake. These three simultaneous EM phenomena were compared with data collected over a 1-2-year period at the site. The data was also compared against accounts of air ionization reported to be associated with radon emission from the ground (Ouzounov, 2007), and a series of laboratory rock stressing experiments (Freund, 2006(Freund, , 2007a to determine if field data was consistent either of these accounts. We could not find a data set with preearthquake radon measurements taken near the Alum Rock epicenter to compare against our field data. However, based on the Alum Rock data set example and another data set at Parkfield, the field tests are at least consistent with the lab experiments showing currents, magnetic field disturbances, air conductivity changes, and IR signatures. This is encouraging, but more instrumented earthquake examples are needed to prove a repeating pattern for these types of pre-earthquake EM signatures.

[1] We use search-coil magnetometer data from a low-latitude station in Parkfield, California (L = 1.77) to study the occurrence of Pc1 pulsations associated with geomagnetic storms. The Pc1 pulsations and storms are identified using automatic algorithms, and the statistical distributions are examined using a superposed epoch analysis technique, as a function of local time, time relative to storm main phase, and storm intensity. Results show that Pc1 pulsations are 2-3 times more likely (than normal) to be observed in the 2-4 d following moderate storms and 4-5 times more likely in the 2-7 d following intense storms. The Pc1 frequencies are higher in moderate storms than they are in nonstorm times and become even higher and occupy a greater range of local times as the strength of the storms increase. These results are consistent with the idea that the source of EMIC waves extends to lower L values as storm intensity increases.

Abstract. We use a relatively simple model of an underground current source co-located with the earthquake hypocenter to estimate the magnitude of the seismotelluric current required to produce observable ground signatures. The Alum Rock earthquake of 31 October 2007, is used as an archetype of a typical California earthquake, and the effects of varying the ground conductivity and length of the current element are examined. Results show that for an observed 30 nT pulse at 1 Hz, the expected seismotelluric current magnitudes fall in the range ∼10-100 kA. By setting the detectability threshold to 1 pT, we show that even when large values of ground conductivity are assumed, magnetic signals are readily detectable within a range of 30 km from the epicenter. When typical values of ground conductivity are assumed, the minimum current required to produce an observable signal within a 30 km range was found to be ∼1 kA, which is a surprisingly low value. Furthermore, we show that deep nulls in the signal power develop in the non-cardinal directions relative to the orientation of the source current, indicating that a magnetometer station located in those regions may not observe a signal even though it is well within the detectable range. This result underscores the importance of using a network of magnetometers when searching for preseismic electromagnetic signals.

No abstract