Establishing the types and durations of non-earthquake sources of energy in continuous waveform records is essential to developing the correct understanding and cataloging of different classes of sources of ground motion. Natural and anthropogenic noise recorded throughout seismic networks contains information from multiple sources. Seismic station coverage continues to improve with the expansion of permanent networks and temporary installation of dense deployments. This trend and improved techniques for detecting small earthquakes and tremor increases the need to properly decipher transient signals in the waveforms. As temporary dense seismic deployments of geophones with high sampling rates (e.g. 500 Hz) become more common in implementing scientific research objectives [e.g. Schmandt and Clayton, 2013; Inbal et al., 2015; Brenguier et al., 2016], the need to develop systematic tests that quantify the coupling of atmospheric and surface processes to the solid Earth are necessary to further the understanding of transient seismic signals. We aim to characterize the coupling of wind to a variety of common in-situ structures (house, machines, poles, fences) located in the study area containing a dense array of geophones. The goal is to quantify the meteorological measurements and types of structures producing seismic signals that appear without careful characterization as possible seismic activity. The outcome will advance the understanding of different classes of sources that contribute to recorded seismic waveforms and the near-surface mechanical properties that allow the propagation of wind-related signals in the damage structure of the SJFZ. The proposed work consists of a one year project plan that involves three components:
(1) Deploy 30-50 high frequency geophones and 2 meteorological stations to record the wind speed and direction, atmosphere pressure and temperature for ~30 days at the Sage Brush Flat site, south of Anza, CA in the trifurcation area of the San Jacinto fault zone [Allam et al., 2014; Ben-Zion et al., 2015]. The seismic sensors will be 3-component Zland geophones similar to those previously deployed to maintain data consistency between the studies. The deployment will require two trips to the SGB field location, one to install the equipment and one for removal. The instruments will be installed to simultaneously measure the ground motions and the motions of structures and equipment located on site (Figure 2b). Affixing geophones to in-situ structures and equipment will allow a careful analysis to quantify the meteorological impact on shallow crustal signals observed in the data. By placing instruments at increasing distance from the structures and equipment we can assess the degree to which the coupling to the solid Earth occurs. Additionally, the data set will be augmented with the PBO borehole station P946 containing a short period seismometer and strainmeter. Combining the borehole records with the surface geophones will aid in properly describing the surface coupling processes. The proposed small deployment will leverage on the existing infrastructure to efficiently combine the new data set with the continuously recording data.
(2) Perform detailed waveform analysis in time and frequency domain to characterize all periods of elevated winds. We will first use the meteorological time series to establish the background levels of wind speed and direction. This information will be used to identify anomalous periods of increased winds to inform us of the timing of possible coupling to the surface structures and the ground. With this information we can explore the characteristics and differences among the set of seismic recordings for the same time period. The waveform analysis will quantify different time domain features such as amplitude and duration to assess the effect of the coupling of the wind to the in-situ structures and solid Earth. The frequency spectrum will be examined to learn the different modes excited by the coupling of the atmosphere processes to the solid Earth. The analysis will be performed to assess the signals as a function of distance from the structures and equipment. If there is strong coupling of the wind through the structures and equipment we expect to find a coherent time domain signal in stations on the equipment and the adjacent seismometers that decays with distance. Similarly, we expect to find a distant dependence in the spectral domain for the excited modes during the wind related events. These tests will demonstrate that the observed signals are converted form the structures to the ground as a function of wind velocity.
(3) Evaluate the total fraction of a day that contains these atmosphere related signals and quantify the change in signal-to-noise ratio during the wind events. The goal here is to establish a clear diagnostic to distinguish between local microseismic events and these other non-tectonic sources of energy. Establishing these criteria will involve lowering the threshold for the wind velocity to examine the degree of atmosphere coupling at lower wind speeds. We will utilize techniques commonly used to identify tremor, i.e. emergent signals with long duration, to develop a catalog of possible wind related noise events. Using envelope cross correlation and template matching we can reliably identify features of the signals and relate this to a corresponding wind velocity. Correctly labeling events as atmosphere coupling processes or possible microseismic events will greatly increase the ability to image subsurface deformation at seismogenic depths. Knowing the total amount of time each day that contains an atmosphere coupling related signal will decrease false detections of other phenomenon occurring at depth.