Observations and modeling of seismic background noise

Abstract

The preparation of this report had two purposes. One was to present a catalog of seismic background noise spectra obtained from a worldwide network of seismograph stations. The other purpose was to refine and document models of seismic background noise that have been in use for several years. The second objective was, in fact, the principal reason that this study was initiated and influenced the procedures used in collecting and processing the data. With a single exception, all of the data used in this study were extracted from the digital data archive at the U.S. Geological Survey's Albuquerque Seismological Laboratory (ASL). This archive dates from 1972 when ASL first began deploying digital seismograph systems and collecting and distributing digital data under the sponsorship of the Defense Advanced Research Projects Agency (DARPA). There have been many changes and additions to the global seismograph networks during the past twenty years, but perhaps none as significant as the current deployment of very broadband seismographs by the U.S. Geological Survey (USGS) and the University of California San Diego (UCSD) under the scientific direction of the IRIS consortium. The new data acquisition systems have extended the bandwidth and resolution of seismic recording, and they utilize high-density recording media that permit the continuous recording of broadband data. The data improvements and continuous recording greatly benefit and simplify surveys of seismic background noise. Although there are many other sources of digital data, the ASL archive data were used almost exclusively because of accessibility and because the data systems and their calibration are well documented for the most part. Fortunately, the ASL archive contains high-quality data from other stations in addition to those deployed by the USGS. Included are data from UCSD IRIS/IDA stations, the Regional Seismic Test Network (RSTN) deployed by Sandia National Laboratories (SNL), and the TERRAscope network deployed by the California Institute of Technology in cooperation with other institutions. A map showing the approximate locations of the stations used in this study is provided in Figure 1. One might hope for a better distribution of stations in the southern hemisphere, especially Africa and South America, in order to look for regional variations in seismic noise (apart from the major differences between continental, coastal and island sites). Unfortunately, anyone looking for subtle regional variations in seismic noise is probably going to be disappointed by the spectral data presented in this report because much of the station data appear to be dominated by local disturbances caused by instrumental, environmental, cultural, or surf noise. Better instruments and better instrument siting, or a well-funded field program, will be needed before a global isoseismal noise map can be produced. However, by assembling a composite of background noise from a large network of stations, many of the local station variables are masked, and it is possible to create generalized spectral plots of Earth noise for hypothetical quiet and noisy station sites.