Anomalously high b-values in the South Flank of Kilauea volcano, Hawaii: evidence for the distribution of magma below Kilauea's East rift zone

https://doi.org/10.1016/S0377-0273(00)00263-8Get rights and content

Abstract

The pattern of b-value of the frequency–magnitude relation, or mean magnitude, varies little in the Kaoiki-Hilea area of Hawaii, and the b-values are normal, with b=0.8 in the top 10 km and somewhat lower values below that depth. We interpret the Kaoiki-Hilea area as relatively stable, normal Hawaiian crust. In contrast, the b-values beneath Kilauea's South Flank are anomalously high (b=1.3–1.7) at depths between 4 and 8 km, with the highest values near the East Rift zone, but extending 5–8 km away from the rift. Also, the anomalously high b-values vary along strike, parallel to the rift zone. The highest b-values are observed near Hiiaka and Pauahi craters at the bend in the rift, the next highest are near Makaopuhi and also near Puu Kaliu. The mildest anomalies occur adjacent to the central section of the rift. The locations of the three major and two minor b-value anomalies correspond to places where shallow magma reservoirs have been proposed based on analyses of seismicity, geodetic data and differentiated lava chemistry. The existence of the magma reservoirs is also supported by magnetic anomalies, which may be areas of dike concentration, and self-potential anomalies, which are areas of thermal upwelling above a hot source. The simplest explanation of these anomalously high b-values is that they are due to the presence of active magma bodies beneath the East Rift zone at depths down to 8 km. In other volcanoes, anomalously high b-values correlate with volumes adjacent to active magma chambers. This supports a model of a magma body beneath the East Rift zone, which may widen and thin along strike, and which may reach 8 km depth and extend from Kilauea's summit to a distance of at least 40 km down rift. The anomalously high b-values at the center of the South Flank, several kilometers away from the rift, may be explained by unusually high pore pressure throughout the South Flank, or by anomalously strong heterogeneity due to extensive cracking, or by both phenomena. The major b-value anomalies are located SSE of their parent reservoirs, in the direction of motion of the flank, suggesting that magma reservoirs leave an imprint in the mobile flank. We hypothesize that the extensive cracking may have been acquired when the anomalous parts of the South Flank, now several kilometers distant from the rift zone, were generated at the rift zone near persistent reservoirs. Since their generation, these volumes may have moved seaward, away from the rift, but earthquakes occurring in them still use the preexisting complex crack distribution. Along the decollement plane at 10 km depth, the b-values are exceptionally low (b=0.5), suggesting faulting in a more homogeneous medium.

Introduction

The island of Hawaii is built upon the ocean floor by magmatic intrusive and extrusive material (Hill and Zucca, 1987). As the ocean floor moves over the hot spot, older volcanoes (Mauna Kea) are disconnected from the magma supply and new volcanoes (Loihi) are created. For this reason, the magma plumbing system under Hawaii may be expected to be more complex (Ryan, 1988) than under single volcanoes that remain above their magma supply for long periods. In such a case of strong complexity it is of special interest, but perhaps also more difficult than usual, to map the conduits and chambers containing magma. In this paper, we use the frequency magnitude distribution (FMD) of earthquakes to infer the presence of magma, concentrating on Kilauea's South Flank and in the Kaoiki-Hilea region between Mauna Loa and Kilauea. Because our method is based on seismicity, we can extract information only from parts of the crust and upper mantle that produce numerous earthquakes. In Hawaii, most seismic activity is concentrated in the volcanic edifice that rests on the depressed oceanic sea floor.

The crustal volumes beneath the South Flank and Kaoiki produce abundant earthquakes (Fig. 1), mostly at depths between 4 and 10 km within the volcanic edifice (Fig. 2). The earthquakes in these volumes are thought to be volcano-tectonic in the sense that they are generated by stresses exerted by the volcanoes and their rifts, but not directly related to opening of magmatic conduits or the flow of magma. The distances of the epicenters from the rift zones and the calderas range from about 1–10 km. The largest recent earthquakes were an M6.6 in 1983 and an M6.1 in 1989 in the Kaoiki and South Flank areas, respectively. However, both areas are underlain by a near horizontal decollement plane, the surface of the old sea floor, which can generate large to great earthquakes (Wyss, 1988). In 1975, an M7.2 earthquake ruptured most of the decollement plane under the South Flank (Swanson et al., 1976b), and in 1868, a M7.9 earthquake probably ruptured all of the South Flank, Kaoiki, Hilea and adjacent areas (Wyss, 1988).

Along Kilauea's east- and southwest rift zones, large numbers of small earthquakes occur during magmatic intrusions (Dvorak et al., 1986). These intrusions often initiate under Kilauea's caldera (Dzurisin et al., 1984, Klein et al., 1987), propagate down the rift zones, generate varying degrees of seismicity, and in many cases end up in eruptions from a subsidiary cone in the rift zone (Hardee, 1987). The seismic activity during rift intrusions occurs primarily between 2 and 4 km depth and follows closely the surface trace of the rift. Because of this, most early models of the east rift zone envisioned the active zone to exist down to 4 or 6 km. However, a model with magma bodies down to 10 km depth beneath the rift zones (Crosson and Endo, 1982, Ryan, 1988) received support from geodetic data (Delaney et al., 1990), which required an expanding magma body at 3–9 km depth to satisfactorily explain the surface deformations. In this paper we present data from the frequency–magnitude relation of earthquakes beneath the South Flank that support the model of magma bodies beneath the East Rift zone that may reach depths of 8 km. The data also suggest that the greatest effect on the South Flank earthquake magnitude distribution occurs at intermediate depths (4–7 km) adjacent to shallow (2–4 km) magma bodies under the rift that have been proposed based on analysis of deformation, lava chemistry and seismicity patterns. We do not analyze the seismicity of the rift zone itself because the earthquakes in it occur primarily during intrusions of magma, have a complicated time–space history and carry no information about crustal conditions in the South Flank.

The relative abundance of small earthquakes compared to large ones is measured by the b-value in the FMD (Ishimoto and Iida, 1939, Gutenberg and Richter, 1944)logN=a−bMwhere N is the cumulative number of earthquakes with magnitude M and larger.

A number of recent papers demonstrated that the constants a and b vary strongly over a few kilometers distance in seismogenic volumes (Ogata et al., 1995, Wiemer and Benoit, 1996, Wiemer and McNutt, 1997, Wiemer and Wyss, 1997, Wyss et al., 1997, Wiemer et al., 1998, Power et al., 1998, Murru et al., 1999, Wiemer and Katsumata, 1999). This means that the mean magnitude, m, varies as a function of space because the b-value is inversely proportional to the mean magnitude (Aki, 1965, Utsu, 1965)m=Mmin+1/2.3bIn this paper, we map m. We do not assume that the FMD is approximated well by a power law, nor do we make any extrapolations based on a straight line fit. Therefore, it is inconsequential if the straight line fit may not be a good approximation in some volumes, because we are simply comparing the mean magnitude in different volumes by mapping the b-value.

Along the San Andreas fault system, creeping segments show anomalously high values (b>1.3) (Amelung and King, 1997, Wiemer and Wyss, 1997), whereas asperities correlate with anomalously low values (b<0.6) (Wiemer and Wyss, 1997, Wyss et al., 1999). In general, the b-values decrease with depth in California (Mori and Abercrombie, 1997, Wiemer and Wyss, 1997, Wiemer et al., 1998, Gerstenberger et al., 2000). In aftershock sequences, great variability of b-values is possibly related to the amount of slip during the main shock (Wiemer and Katsumata, 1999). Beneath volcanoes, the volumes surrounding active magma chambers and conduits exhibit anomalously high b-values (Wiemer and McNutt, 1997, Wyss et al., 1997, Wiemer et al., 1998, Murru et al., 1999).

We know from laboratory experiments, observations in underground mines and near wells that the b-value increases with decreasing effective stress, and with increasing heterogeneity of the material (Mogi, 1962, Scholz, 1968, Wyss, 1973, Urbancic et al., 1992), and that it may also be sensitive to the temperature gradient (Warren and Latham, 1970). A recent study by Lahaie and Grasso (1999) showed a correlation between the loading rate and the b-value at the Lacq gas field in France: When the loading rate was reduced, the b-value increased by almost a factor of two.

We interpreted the low b-values in asperities as expressions of elevated stress levels. The anomalously high b-values near magma chambers can be explained as due to larger-than-normal heterogeneity, or, alternatively, as due to high pore pressure. The purpose of this paper is to map the distribution of b-values in Kilauea's South Flank and in the Kaoiki-Hilea area, and to construct a model for the generation of the related earthquakes.

Section snippets

Data

The newest version of the Hawaiian Volcano Observatory's earthquake catalog was used to estimate the mean magnitude (b-value). Among the several magnitudes available, we have most confidence in the homogeneity of the ‘preferred magnitude’ (Mp) because this magnitude furnishes stable b-value estimates and it is available for the largest number of earthquakes. Mp is selected by a hierarchical scheme defined in the data description prepared by F. Klein for distribution of the Hawaiian earthquake

Method

We estimate the mean magnitude, or b-value, by both the maximum likelihood and the least squares method at every node of grids with spacing of 1 km or less (Wiemer, 1996). At each node, we extract the nearest N events (50<N<300, but typically N=100) to estimate b. The radii of the cylindrical volumes that contain 100 events are typically r=1, 1.5 and 3 km on the South Flank, in Kaoiki and in Hilea, respectively, but they vary inversely proportional to the local seismicity rate. From the resulting

Average b-value as a function of depth

The average b-value is estimated as a function of depth by a moving sample window of N=300 events, starting with the 300 shallowest earthquakes and moving the window deeper in steps of 10 events (Fig. 3). In the Kaoiki-Hilea area, the b-value remains approximately constant (b=0.8) down to about 10 km depth, below which it decreases slightly (Fig. 3b). This is similar to the pattern observed along parts of the San Andreas fault (Eaton et al., 1970, Wyss, 1973, Mori and Abercrombie, 1997, Wiemer

Discussion and conclusions

A normal pattern of b-value as a function of depth (Eaton et al., 1970, Wyss, 1973, Mori and Abercrombie, 1997, Wiemer and Wyss, 1997, Gerstenberger et al., 2000) is exhibited by the crust beneath Kaoiki-Hilea (Fig. 3, Fig. 4). We interpret this to mean that there is nothing special about the distribution of cracks, or the state of stress, in Hawaiian crust, in general. However, the anomalously high b-values beneath Kilauea's South Flank (Fig. 3, Fig. 4, Fig. 5, Fig. 6) show that there the

Acknowledgements

We thank M. R. Ryan and G. Saccorotti for suggestions and rewording in reviews that led to many improvements and T. Wright for helpful comments. This work was supported by NSF under grant number EAR-9614783, the Wadati endowment at the Geophysical Institute of the University of Alaska, Fairbanks, and a fellowship from the overseas research scholarship fund number 8-waka-190 by the Ministry of Education, Science, Sports and Culture, Japan for K. N. This paper is contribution number 1133 of the

References (61)

  • J.J. Dvorak et al.

    Mechanical response of the south flank of Kilauea volcano, Hawaii, to intrusive events along the rift systems

    Tectonophysics

    (1986)
  • D. Dzurisin et al.

    Magma supply and storage at Kilauea volcano, Hawaii

    J. Volcanol. Geotherm. Res.

    (1984)
  • J. Kauahikaua

    Geophysical characteristics of the hydrothermal systems of Kilauea volcano, Hawaii

    Geothermics

    (1993)
  • K. Aki

    Maximum likelihood estimate of b in the formula logN=a−bM and its confidence limits

    Bull. Earthq. Res. Inst.

    (1965)
  • F. Amelung et al.

    Earthquake scaling laws for creeping and non-creeping faults

    Geophys. Res. Lett.

    (1997)
  • R.S. Crosson et al.

    Focal mechanisms and locations of earthquakes in the vicinity of the 1975 Kalapana earthquake aftershock zone 1970–1979: implications for tectonics of the South Flank of Kilauea volcano

    Tectonics

    (1982)
  • P.T. Delaney et al.

    Deep magma body beneath the summit and rift zones of Kilauea Volcano, Hawaii

    Science

    (1990)
  • J.H. Dieterich

    Growth and persistence of Hawaiian volcanic rift zones

    J. Geophys. Res.

    (1988)
  • J.J. Dvorak et al.

    Analysis of surface deformation data, Kilauea Volcano, Hawaii, October 1966 to September 1970

    J. Geophys. Res.

    (1983)
  • J. Eaton et al.

    Aftershocks of the 1966 Parkfield-Cholame, California, earthquake: a detailed study

    Bull. Seismol. Soc. Am

    (1970)
  • J.C. Eichelberger et al.

    Eruption of andesite triggered by dyke injection: contrasting cases at Karymsky volcano, Kamtchatka and Mt. Katmai, Alaska

    Phil. Trans. R. Soc. London A

    (2000)
  • R.S. Fiske et al.

    Orientation and growth of Hawaiian volcanic rifts: The effect of regional structure and gravitational stresses

    Proc. R. Soc. London, Ser. A

    (1972)
  • R.S. Fiske et al.

    A model of Kilauea volcano's rift-zone magma system

    EOS

    (1993)
  • M. Gerstenberger et al.

    A systematic test of the hypothesis that the b value varies with depth in California

    Geophys. Res. Lett.

    (2000)
  • D. Gillard et al.

    Stress and strain tensor orientations in the south flank of Kilauea, Hawaii, estimated from fault plane solutions

    J. Geophys. Res.

    (1995)
  • D. Gillard et al.

    Highly concentrated seismicity caused by deformation of Kilauea's deep magma system

    Nature

    (1996)
  • J.L. Got et al.

    Deep fault plane geometry inferred from multiplet relative relocation beneath the south flank of Kilauea

    J. Geophys. Res.

    (1994)
  • R. Gutenberg et al.

    Frequency of earthquakes in California

    Bull. Seismol. Soc. Am

    (1944)
  • R.E. Habermann

    Teleseismic detection in the Aleutian Island arc

    J. Geophys. Res.

    (1983)
  • Hardee, H.C., 1987. Heat and mass transport in the east-rift-zone magma conduit of Kilauea volcano. In: Decker, R.W.,...
  • Hill, D.P., Zucca, J.J., 1987. Geophysical constraints on the structure of Kilauea and Mauna Loa volcanos and some...
  • M. Ishimoto et al.

    Observations of earthquakes registered with the microseismograph constructed recently

    Bull. Earthq. Res. Inst.

    (1939)
  • Jackson, D.B., Swanson, D.A., Koyanagi, R.Y., Wright, T.L., 1975. The August and October 1968 east rift eruptions of...
  • F.W. Klein

    A linear gradient crustal model for south Hawaii

    Bull. Seismol. Soc. Am

    (1981)
  • Klein, F.W., Koyanagi, R.Y., Nakata, J.S., Tanigawa, W.R., 1987. The Seismicity of Kilauea's magma system. In:...
  • Lahaie, F., Grasso, J.R., 1999. Loading rate impact on fracturing pattern: lessons from hydrocarbon recovery, Lacq gas...
  • K. Mogi

    Magnitude–frequency relation for elastic shocks accompanying fractures of various materials and some related problems in earthquakes

    Bull. Earthq. Res. Inst.

    (1962)
  • R.B. Moore

    Distribution of differentiated tholeiitic basalts on the lower east rift zone of Kilauea Volcano, Hawaii: a possible guide to geothermal exploration

    Geology

    (1983)
  • J. Mori et al.

    Depth dependence of earthquake frequency–magnitude distributions in California: Implications for the rupture initiation

    J. Geophys. Res.

    (1997)
  • M. Murru et al.

    The location of magma chambers at Mt. Etna, Italy, mapped by b-values

    Geophys. Res. Letts.

    (1999)
  • Cited by (79)

    View all citing articles on Scopus
    View full text