Non-double-couple microearthquakes at Long Valley caldera, California, provide evidence for hydraulic fracturing

https://doi.org/10.1016/S0377-0273(03)00420-7Get rights and content

Abstract

Most of 26 small (0.4≲M≲3.1) microearthquakes at Long Valley caldera in mid-1997, analyzed using data from a dense temporary network of 69 digital three-component seismometers, have significantly non-double-couple focal mechanisms, inconsistent with simple shear faulting. We determined their mechanisms by inverting P- and S-wave polarities and amplitude ratios using linear-programming methods, and tracing rays through a three-dimensional Earth model derived using tomography. More than 80% of the mechanisms have positive (volume increase) isotropic components and most have compensated linear-vector dipole components with outward-directed major dipoles. The simplest interpretation of these mechanisms is combined shear and extensional faulting with a volume-compensating process, such as rapid flow of water, steam, or CO2 into opening tensile cracks. Source orientations of earthquakes in the south moat suggest extensional faulting on ESE-striking subvertical planes, an orientation consistent with planes defined by earthquake hypocenters. The focal mechanisms show that clearly defined hypocentral planes in different locations result from different source processes. One such plane in the eastern south moat is consistent with extensional faulting, while one near Casa Diablo Hot Springs reflects en echelon right-lateral shear faulting. Source orientations at Mammoth Mountain vary systematically with location, indicating that the volcano influences the local stress field. Events in a ‘spasmodic burst’ at Mammoth Mountain have practically identical mechanisms that indicate nearly pure compensated tensile failure and high fluid mobility. Five earthquakes had mechanisms involving small volume decreases, but these may not be significant. No mechanisms have volumetric moment fractions larger than that of a force dipole, but the reason for this fact is unknown.

Introduction

The first convincing examples of ‘non-double-couple’ (non-DC) volcanic earthquakes, whose mechanisms are incompatible with simple shear faulting, were three moment magnitude (Mw) 5.5–6.2 earthquakes at Long Valley caldera, in eastern California, in 1978 and 1980. These were the Mw 5.5 Wheeler Crest earthquake of 4 October 1978 (Ekström and Dziewonski, 1983) and two of the four Mw ∼6 earthquakes of 25 and 27 May 1980 (Barker and Langston, 1983, Julian, 1983, Julian and Sipkin, 1985). Since that time, non-DC microearthquakes have been identified at many other volcanic areas (Miller et al., 1998a), but the physical processes that cause them remain imperfectly understood, and it is not clear even that non-DC events of widely differing magnitudes are genetically similar. Julian (1983) and Julian and Sipkin (1985) suggested that the large 1978 and 1980 Long Valley earthquakes involved tensile failure caused by fluid pressure, but the seismic data available for these events could not resolve the volume changes that would accompany tensile faulting, so alternative explanations such as simultaneous slip on differently oriented shear faults have remained tenable. Recently, Dreger et al. (2000) fitted complete waveforms recorded on regional ‘broadband’ seismometers for six earthquakes of Mw 4.6–4.9 in December 1997, and showed that four of them have mechanisms with significant volume increases, and therefore must involve processes other than shear faulting.

Because of instrumental limitations, it has remained uncertain whether any of the thousands of microearthquakes that have occurred at Long Valley caldera since 1978 have non-DC mechanisms similar to those of the larger earthquakes. The permanent seismometer network in the area consists primarily of vertical-component sensors and uses analog telemetry and recording techniques with low dynamic range, so the only data bearing on focal mechanisms that they provide are P-wave first-motion polarities, which contain little information and are of little use for determining full moment-tensor source mechanisms (Julian et al., 1998).

To resolve questions about the mechanisms of microearthquakes at the caldera, the U.S. Geological Survey and Duke University installed and operated a temporary network of 65 three-component digital seismometers in the area from mid-May to late September 1997. During this period, the caldera experienced a series of earthquake swarms, which intensified with time and were accompanied by accelerated ground deformation (Foulger et al., 1998). An unusually complete seismic data set was obtained for these swarms, covering earthquakes down to moment magnitudes less than 1.

In this paper, we describe results from analysis of this data set using methods based on arrival-time differences and amplitude ratios, which extract extra information from seismograms and are relatively insensitive to bias caused by wave-propagation anomalies. The derived focal mechanisms show that most of the earthquakes have volume increases, and therefore involve processes other than shear faulting. High-resolution relative earthquake locations further show that the failure geometry of some events is close to that expected for tensile faulting, but the mechanisms require an accompanying volume-compensating process, probably involving rapid fluid flow.

Section snippets

Long Valley caldera

The 17- by 30-km silicic Long Valley caldera is located at the northern end of Owens Valley, on the boundary between the Basin and Range province and the Sierra Nevada in eastern California (Fig. 1). The caldera floor lies roughly 2200 m above sea level, and contains a resurgent dome about 10 km in diameter, which rises about 300 m above the surrounding moat. The geology, geophysics and tectonic history of the area have been described in detail by many authors (e.g. Bailey et al., 1976, Hill et

Previous earthquake mechanism determinations for Long Valley caldera

The mechanisms derived for the three non-DC Mw 5.5–6.2 events of October 1978 and May 1980 were constrained to be deviatoric and are close to CLVDs with their largest principal moments extensional and oriented east-northeast. Julian (1983) and Julian and Sipkin (1985) suggested that they were caused by tensile failure associated with intrusion of fluids into NNW-striking vertical cracks and had intrusion volumes possibly as large as 0.07 km3. The active fluid was thought to be water or CO2,

1997 seismicity

We computed moment tensors for 26 earthquakes with NCSN duration magnitudes between 0.43 and 3.14 that occurred during the summer of 1997, in approximately the first half of the seismic episode described above. Earthquake swarm activity began just west of Casa Diablo Hot Springs on 7 July, about three weeks after inflation of the resurgent dome began to accelerate. During the summer the temporary network recorded more than 4000 locatable earthquakes, along with ∼18 deep long-period events per

Results

Table 1 gives hypocentral coordinates and magnitudes for the 26 earthquakes analyzed, and Table 2 gives the derived focal mechanisms, in the form of moment-tensor components, source-type parameters (see below), principal moments, and principal-axis orientations. The moment tensors are normalized arbitrarily, so only the relative values of the moment-tensor components and principal moments are significant. Fig. 5 shows the geographical distribution of the mechanisms and previously published

Comparison with previous results from Long Valley caldera

The earthquakes described here are the smallest at Long Valley caldera for which non-DC mechanisms have been reported, and the fact that they occur places constraints on the physical processes involved. The non-DC earthquakes of October 1978 and May 1980 had Mw ∼6 (Julian, 1983, Julian and Sipkin, 1985) and those of November 1997 Mw 4.6–4.9 (Dreger et al., 2000). These large earthquakes all had substantial CLVD components, and the opening of cracks with volumes of 6.4×104 to 7×104 m3,

Conclusions

  • 1.

    Most of the small (0.4<Mw<3.1) earthquakes that occurred in the summer and early fall of 1997 during a time of accelerating earthquake swarms and inflationary crustal deformation at Long Valley caldera have non-DC mechanisms with positive CLVD and isotropic (volume increase) components.

  • 2.

    The source types of most of these earthquakes are consistent with mixed shear and tensile faulting, but require an additional volume-compensating process, which might be provided by the flow of fluid into opening

Acknowledgements

The instruments used in the field program were provided by the Natural Environment Research Council (NERC loan 540/0197) and the PASSCAL facility of the Incorporated Research Institutions for Seismology (IRIS) through the Stanford Instrument Center. Data collected during this experiment will be available through the IRIS Data Management Center. The facilities of the IRIS Consortium are supported by the National Science Foundation under Cooperative Agreement EAR-9023502. We thank the U.S. Forest

References (56)

  • R. Console et al.

    Non-double-couple microearthquakes in the geothermal field of Larderello, central Italy

    Tectonophysics

    (1998)
  • D.L. Anderson et al.

    The anelasticity of the Earth

    J. Geophys. Res.

    (1964)
  • S.K. Arnott et al.

    The Krafla spreading segment, Iceland: 2. The accretionary stress cycle and non-shear earthquake focal mechanisms

    J. Geophys. Res.

    (1994)
  • R.A. Bailey et al.

    Volcanism, structure, and geochronology of Long Valley Caldera, Mono County, California

    J. Geophys. Res.

    (1976)
  • J.S. Barker et al.

    A teleseismic body wave analysis of the May, 1980 Mammoth Lakes, California earthquakes

    Bull. Seismol. Soc. Am.

    (1983)
  • D.J. Barton et al.

    Frequency-magnitude statistics and spatial correlation dimensions of earthquakes at Long Valley caldera, California

    Geophys. J. Int.

    (1999)
  • W.F. Brace et al.

    A note on brittle crack growth in compression

    J. Geophys. Res.

    (1963)
  • J.F. Claerbout et al.

    Robust modeling with erratic data

    Geophysics

    (1973)
  • R.P. Denlinger et al.

    Deformation of Long Valley caldera between August 1982 and August 1983

    J. Geophys. Res.

    (1985)
  • D.S. Dreger et al.

    Dilational processes accompanying earthquakes in the Long Valley caldera

    Science

    (2000)
  • G. Ekström et al.

    Moment tensor solutions of Mammoth Lakes earthquakes

    EOS Trans. AGU

    (1983)
  • C.D. Farrar

    Forest-killing diffuse CO2 emission at Mammoth Mountain as a sign of magmatic unrest

    Nature

    (1995)
  • G.R. Foulger

    Hengill triple junction, SW Iceland; 2. Anomalous earthquake focal mechanisms and implications for process within the geothermal reservoir and at accretionary plate boundaries

    J. Geophys. Res.

    (1988)
  • G.R. Foulger et al.

    Non-double-couple earthquakes at the Hengill-Grensdalur volcanic complex, Iceland: Are they artifacts of crustal heterogeneity?

    Bull. Seismol. Soc. Am.

    (1993)
  • Foulger, G.R. et al., 2002. Three-dimensional crustal structure of Long Valley caldera, California, and evidence for...
  • G.R. Foulger et al.

    Anomalous focal mechanisms; tensile crack formation on an accreting plate boundary

    Nature

    (1984)
  • G.R. Foulger et al.

    Implosive earthquakes at the active accretionary plate boundary in northern Iceland

    Nature

    (1989)
  • G.R. Foulger et al.

    Seismic monitoring and activity increase in California caldera

    EOS Trans. AGU

    (1998)
  • K. Fuchs et al.

    Computation of synthetic seismograms with the reflectivity method and comparison with observations

    Geophys. J. R. Astron. Sci.

    (1971)
  • D.P. Hill

    Earthquakes and carbon dioxide beneath Mammoth Mountain, California

    Seismol. Res. Lett.

    (1996)
  • D.P. Hill et al.

    Active tectonic and magmatic processes beneath Long Valley caldera, eastern California: An overview

    J. Geophys. Res.

    (1985)
  • D.P. Hill

    The 1989 earthquake swarm beneath Mammoth Mountain, California: An initial look at the 4 May through 30 September activity

    Bull. Seismol. Soc. Am.

    (1990)
  • J.A. Hudson et al.

    Source type plot for inversion of the moment tensor

    J. Geophys. Res.

    (1989)
  • S.-H. Hung et al.

    Non-double-couple focal mechanisms in an oceanic, intraplate earthquake swarm: Application of an improved method for comparative event, moment tensor determination

    J. Geophys. Res.

    (1996)
  • B.R. Julian

    Evidence for dyke intrusion earthquake mechanisms near Long Valley caldera, California

    Nature

    (1983)
  • B.R. Julian

    Analysing seismic-source mechanisms by linear-programming methods

    Geophys. J. R. Astron. Sci.

    (1986)
  • B.R. Julian

    Volcanic tremor: Nonlinear excitation by fluid flow

    J. Geophys. Res.

    (1994)
  • B.R. Julian et al.

    Earthquake mechanisms from linear-programming inversion of seismic-wave amplitude ratios

    Bull. Seismol. Soc. Am.

    (1996)
  • Cited by (124)

    • Moment tensor inversion of microearthquakes along the Santorini-Amorgos zone: Tensile faulting and emerging volcanism in an extensional setting

      2021, Journal of Volcanology and Geothermal Research
      Citation Excerpt :

      The use of P-wave polarities and amplitude ratios of SV/P, SH/P, or SV/SH to determine the source mechanism is considered as one solution to this problem. This method also eliminates the instrument response owing to the use of amplitude ratios rather than absolute amplitudes (Julian et al., 1998; Foulger et al., 2004; Jechumtálová and Šílený, 2005; Guilhem et al., 2014). In a previous work, Andinisari et al. (2021) investigated the role played by fluids in the Santorini-Amorgos zone by using raypath Vp/Vs ratios and by calculating various petrophysical parameters (Poisson's ratio, crack density, fluid saturation).

    View all citing articles on Scopus
    View full text