Active Seismic Imaging of Axial Volcano

William Menke and Maria Tolstoy, PI's
Version 2; January 1999

Goal: Understand the relationship between the Axial Volcano magma chamber and volcanism and crustal genesis on the surrounding Juan de Fuca ridge.

Methodology: 3D seismic imaging based on P, PmP and Pn wave traveltime data from an active source (airgun-to-OBS/OBH) seismic experiment.

Introduction

Axial volcano, in the Northeast Pacific, is a large ridge-centered seamount associated with the Cobb hotspot (figure 1). Its position on the actively-spreading Juan de Fuca ridge (JdF, 60 mm/yr full spreading rate), its proximity to western North America, its shallow (1600 m) summit depth, its prominent (2x8 km wide) caldera, and it vigorous hydrothermal activity have led to its being the focus of numerous research efforts (e.g. special section in the August 1990 issue of J. Geophys. Res.).

In 1993 a large seafloor volcanic eruption occurred along the CoAxial segment of the JdF, immediately to the north of Axial Volcano. This eruption was detected during its early stages by SOSUS observations (Fox et al., 1995), and subsequently studied intensively (e.g. special section in January 15, 1996 issue of Geophys. Res. Lett.) The eruption appears to have been caused by the lateral propagation of a dike from the magma chamber of Axial Volcano to a site 25 km to the northeast (Dziak et al., 1995). The sequence of events seems to be similar to the 1974-1985 rifting episode in northern Iceland, which involved the lateral propagation of dikes away from Krafla Volcano (Brandsdottir and Einarsson, 1979). The Iceland rifting episode led to about 9 meters of spreading of the North American - Eurasian plate boundary. The amount of spreading associated with the CoAxial eruption is not known.

Unlike Krafla, whose magma chamber has been imaged by shear wave attenuation (Einarsson, 1978) and P wave delays (Brandsdottir et al. 1997), the presence of a magma chamber at Axial Volcano is less well determined. The 50-100 m deep caldera, interpreted as a collapse feature, indicates the presence of magma in the recent past. A 15 cm subsidence of the caldera floor detected by a sea floor pressure sensor over a two week period in 1988 is similar to those observed at Krafla and might represent magma chamber deflation (Fox, 1990). A joint seafloor and sea surface gravity study has indicates the presence of low density material beneath the caldera, which is interpreted to be a magma chamber (Hildebrand et al. 1990). In 1994, OBSs deployed in and around the caldera, following brief volcanic-type activity recorded on SOSUS (C. Fox, pers. comm.), recorded harmonic tremor and substantial seismic activity possibly associated with magma movement and eruption (Tolstoy et al. 1997). The propagation of the 1993 dike away from Axial volcano also implies a source of magma somewhere in its vicinity. There are two, sub-parallel neovolcanic zones (NVZ's) in this region, the NE rift zone (associated with "normal" plate spreading on the ridge axis) and the CoAxial NVZ (associated with Axial Volcano) (Embley et al. 1990). Sohn et al.'s (1996) local seismic study, conducted after the eruption, appear to indicate that both segments were involved: The dike began propagating along the CoAxial NVZ, but then bridged over to the NE rift zone at its northern terminus.

The role of an Axial Volcano magma chamber (presuming it to exist) in crustal genesis is also unclear. Two interrelated unresolved issues are its degree of connectivity to the mantle magmatic source and to the along-axis NVZ's, and the thermal state of the crust beneath the volcano. Two quite different scenarios have been examined elsewhere: The thin (<100 m) axial magma lens of the East Pacific Rise seems to be the top-most part of a broad zone of partial melt that extends down to the Moho, and spreads in a nearly 2D and steady fashion (Harding et al. 1989, Vera et al. 1990; Phipps-Morgan and Chen, 1993); The much thicker (750-1000 m) magma chamber of Krafla Volcano sits atop a solid, cool dome of gabbroic cumulates, and spreads in an 3D and episodic fashion (Brandsdottir and Einarsson, 1979; Menke and Sparks, 1995; Brandsdottir et al., 1997). Whether Axial Volcano is similar to either of these two scenarios is not known.

The NSF-funded Webb & Sohn experiment, to which our Piggyback Experiment is proposed, is a passive seismic experiment that will deploy 16 Ocean Bottom Seismometers (OBS's) and 4 Ocean Bottom Hydrophones (OBH's) around Axial Volcano for an 8 month period (figure 2). An estimated 1500 microearthquakes will be recorded. Traveltimes from these microearthquakes will be used to image Axial volcano and determine the volume and shape of its magma chamber (if any). Temporal and spatial patterns of seismicity and earthquake focal mechanisms will also be used to infer along and cross-axis variations in the stress field, and to determine the depth of the brittle/ductile transition zone along the ridge axis.

The Webb & Sohn experiment will use naturally-occurring earthquake sources, which have the advantage of producing both P and S waves, of occurring at a range of crustal depths, and of occurring free-of-charge. (Note that even though converted S waves can often be generated by airguns, earthquakes produce significantly more S wave energy). The main limitation of natural sources, however, is their limited geographical extent. Most of the microearthquakes are expected to occur along a narrow band that follows the ridge axis. The poor distribution of sources will limit the imaging to the immediate vicinity of the ridge axis. The use of discrete events, as contrasted to the lines of events possible with artificial (airgun) sources, will probably also make the construction of seismic record sections - and thus the identification of secondary phases - exceedingly difficult. Phases such as PmP and SmS (Moho-reflections) could potentially provide information about the structure of the lower crust and the role of Axial volcano in crustal genesis.

We are thus proposing that an active Piggyback Experiment, that uses an airgun source, be added to Webb & Sohn's passive seismic experiment. Our plan, discussed further below, calls for 11 airgun lines, collected over the period of 7 days, data from which will supplement the passive array deployment. Webb & Sohn have encouraged our efforts, because the data will complement their project nicely. They are therefore allowing us to shoot to their instruments for a week of time during their passive experiment, and will provide us with the airgun recordings upon recovery of their instruments. (They will collect data continuously at 128 samples/s, a rate sufficiently high to record an airgun source).

Cruise Plan

Our nominal plan is to collect 7 days (about 1230 km) of airgun data, shot with R/V Ewing's 20-gun 8300 cubic inch tuned airgun array. We have discussed the use of the R/V Ewing with the Mike Rawson (Ewing Ship Scheduler) and it appears that there is a high likelihood of being able to use the Ewing during the deployment window of the Webb instruments. However, if R/V Ewing should become unavailable, the portable Scripps airgun array could be used on the R/V Thompson instead). Two days of transit are also requested. We are also requesting one day of hydrosweep time to fill in the bathymetry gaps that exist at the far ends of some of the lines.

The airgun lines can be divided into four sets (figure 2):

  1. Set A consists of 5 along-strike lines, each 100 km long and spaced 5 km apart. The center track, A3, crosses the center of the caldera, and also approximately intersects the nominal locations of 6 OBS/OBH's.

  2. Set B consists of 4 cross-strike lines, each 100 km long and spaced 10 km apart. Track B2 intersects the center of the caldera, and also approximately intersects the nominal locations of 4 OBS/OBH's.

  3. Set C, consisting of a single 40 km long track, is chosen to provide a fan-shot geometry through the caldera (see discussion below).

  4. Set D, consisting of two 100 km long tracks spaced 20 km apart, have an orientation chosen to fill in the gaps between the A and B tracks, and also to provide some coverage of the part of the JdF that is southeast of Axial Volcano (i.e. the South Rift Zone).

We plan to use a relatively slow airgun repeat time of 120 s, so as to avoid interference between the seismic waves and the water column multiples of previous shots. At the nominal ship speed of 4 knots, this rate corresponds to an intershot spacing of about 250 m, which in our experience is sufficiently small to allow phase coherence between arrivals in adjacent traces in record sections (which is important in identifying secondary seismic phases).

Primary Scientific Goals of the Piggyback Experiment

  1. Determine the 3D compressional velocity structure of the crust in a roughly 40x40 km box centered on Axial Volcano, and especially assess lower crustal heterogeneity, variations in the depth to Moho, and their relationship to the Axial Volcano magma chamber.

  2. Determine the nature of the crust beneath the Axial Volcano magma chamber, and in particular determine whether it sits atop a partially molten "mush" zone (as has been proposed to the East Pacific Rise) or atop a dome of solid gabbroic cumulates (as has been proposed for Krafla Volcano).

  3. Understand the relationship between the Axial Volcano magma chamber and crustal genesis in its immediate vicinity.

Why Primary Goals are Important, and How the Experiment Addresses Them

A regional crustal model places the magma chamber in a structural context. A magma chamber on a ridge is not an isolated object, but rather is one element of the magma storage and delivery system of the spreading center. The surrounding crust, and in particular its thickness and the relative volume of extrusive and intrusive material, provide a record of the long-term influence of the volcano on the plate tectonic processes. If, for instance, the volcano has increased the magmatic productivity of the ridge, we might expect that the crust beneath it will be thicker than elsewhere along the ridge. Similarly, if dike injection events tend to originate from the volcano, we might expect a systematic along-strike variation in the thickness of the extrusive layer. Long term changes in the productivity of the volcano would have an expression in the spreading-parallel direction.

The airgun lines that we propose are designed to provide sources of compressional waves that criss-cross a roughly 100x100 km region surrounding axial volcano. Ray coverage within the central 40x40 km part of this region will be particularly good, and suitable for 3D imaging (figure 3). We expect to be able to resolve features with horizontal length scales of 2-4 km and vertical length scales of 50 m. We focus on the P-PmP-Pn ray system, with P being the crustal refracted arrival, PmP being the Moho reflected arrival, and Pn being the refracted arrival sampling the top of the mantle. The OBS and OBH recordings of these rays will be analyzed in three ways: 1) P-wave traveltimes, especially for the shorter ranges for which P is the first-arriving phase, will be tomographically inverted to provide a 3D model of upper crustal velocities; 2) PmP traveltimes will be used to map the thickness of the crust and its variation with geographical location; and 3) Waveform modeling of the longer-range (i.e. secondary arrival) P data will provide estimates of lower crustal structure. The present uncertainty in crustal thickness impacts the experimental design, because the pattern of rays (and in particular the PmP critical distance and the P-Pn crossover distance) depend upon crustal thickness. We have designed the lines to be optimized for a maximum crustal thickness of 10 km or less, but the experiment can tolerate crustal thicknesses of up to about 15 km.

Other useful seismic phases may also be recorded. If Pn is well-recorded we will attempt to use it to map variations in upper mantle velocity and direction of anisotropy (e.g. Dunn and Toomey 1997). Such data would place useful constraints on mantle temperatures and flow directions. If converted shear waves, such as PS or PmS, are recorded, we will use them to search for the changes in Poisson's ratio that might signal partial melting in the crust.

A regional crustal model will facilitate Webb & Sohn's passive experiment, by allowing more accurate earthquake location. A key element of the passive experiment is the investigation of the thermal structure of Axial Volcano and its environs. The maximum depth of earthquakes will be used as a proxy for the depth of the brittle-ductile transition (roughly the 450 deg C isotherm in basalt). Accurate earthquake hypocenters are vital to this agenda. While joint tomography/hypocentral location methods can be used to simultaneously determine hypocentral parameters and 3D velocity structure, they have poor resolving power when the geographical distribution of earthquakes is not uniform. Prior constraints on the velocity structure, derived from even a coarse regional experiment such as we propose, can significantly improve the performance of these methods.

Magma leaving the magma chamber contributes to crustal genesis. While dike injection events and their related extrusive and intrusive rocks build the upper crust, other sources may contribute to the lower crust. Lower crustal material may crystallize within a partial melt zone beneath the magma chamber; or it may be formed by solid state flow of cumulates from the magma chamber; or it may be underplated onto the crust at the Moho. Each of these scenarios implies a different structure beneath the magma chamber. Solid state flow of cumulates, for example, is clearly important in the Krafla volcano, where it has built up a high velocity dome in the mid-crust (Menke and Sparks, 1995; Brandsdottir et al., 1997).

The Piggyback Experiment will allow us to use P and PmP data to "undershoot" the magma chamber, and thus to investigate the behavior of the lower crust and the Moho. In particular, we will be able to compare lower crustal velocities and crustal thickness measured directly beneath Axial Volcano to values measured nearby. Airgun lines A3 and B2 cross both the caldera and several outlying OBS/OBH's. This undershooting geometry (i.e. A3 to S1 and S19, and B2 to S6 and S13) contains P waves that both pass through and pass under the magma chamber, and PmP waves with bounce points directly beneath and to the sides of the caldera (figure 4). They will permit a detailed examination of lower crustal and Moho structure. The other A and B airgun lines provide "control" information. They sample the lower crust and Moho in outside of the caldera region (as well as providing the broad ray coverage needed for the 3D imaging).

The undershooting technique has proved very valuable in the study of Krafla volcano (Brandsdottir et al. 1997). The effect of both the magma chamber and the high velocity dome is clearly evident in the undershooting data (figure 5). In the case of Krafla, a dramatic rise in Moho beneath Krafla (in both the ridge-parallel and ridge-perpendicular directions) is also detected (figure 6).

Airgun line C1 can be used in conjunction with the southeastern OBS/OBH's (S18 AND S19) to form a "fan-shot" geometry. P waves sweep beneath the caldera, sampling the mid-crust. Whether these rays will pass through or below the magma chamber is uncertain (since the crustal structure is as yet unknown). Either way, they will provide useful information on the width of structures under the central part of the volcano. A similar fan-shot (figure 7) has been successfully employed to map the ridge-parallel extent of Krafla's cumulate dome (Menke et al., 1997).

Models of melt generation and flow require knowledge of crustal structure. We see this experiment as one element of a larger effort (which is outside the scope of this proposal) to understand the processes by which magma is produced in the mantle, transported, and accreted to the crust. Such an effort would utilize crustal data (such as Webb & Sohn's and our proposed experiment would collect), measurements of mantle properties (such as might be gained from teleseismic tomography and marine gravity), and numerical simulations of heat and mass transport to quantitatively model Axial volcano and the JdF.

Data Processing and Responsibilities

Menke and Tolstoy will be jointly responsible for the overall management and timely completion of the project, and will each participate in all aspects of the project. Tolstoy, who has extensive sea-going experience, will lead the airgun deployment and forward modeling efforts. Menke, who has significant experience in seismic imaging projects in Iceland, southern California and New England will lead the data processing and tomographic inversion efforts. A graduate student will also participate in the data collection and processing.

The data will be collected in Year 1 of the project. Menke, Tolstoy and a graduate student will go to sea to shoot the air gun lines. The graduate student will do some initial preparation and set up for the project while the instruments are deployed, and will archive and begin the initial processing of the raw data upon instrument recovery. We have available (and have extensive experience with) a variety of data display and manipulation programs that allow rapid construction of seismic record sections and which facilitate the measurement of traveltimes.

The bulk of the analysis will be conducted in year two. The forward modeling will be done using the ray tracing code 'rtmod' (Caress et al. 1992) and a finite difference code for propagating P and S waves through a magma chamber (Brandsdottir et al. 1997, figure 12). The tomographic inversion will be conducted using the code of Hu et al. (1993). We shall make special effort to collaborate with Webb and Sohn during all phases of our project, and to share with them travel time data that might be used in their hypocentral locations. A trip to Scripps is included for this purpose.

The results will be present at the fall AGU in San Francisco in December of year 2, and will be published in a professional journal. We will also make preliminary results available through the LDEO web-server.

References

Brandsdottir, B. and P. Einarsson, Seismic activity associated with the September 1997 deflation of the Krafla central volcano in northeastern Iceland, J. Vol. Geotherm. Res. 6, 197-212, 1979.

Brandsdottir, B., W. Menke, P. Einarsson, R. White and R. Staples, Faeroe-Iceland Ridge Experiment 2 - Crustal structure of the Krafla central volcano, J. Geophys. Res. 102, 7867-7886, 1997.

Caress, D.W., M.S. Burnett and J.A. Orcutt, Tomographic image of the axial low-velocity zone at 12:50 N on the East Pacific Rise, J. Geophys. Res. 97, 9243-9263, 1992.

Dunn, R.A., D.R. Toomey, Seismological evidence for three-dimensional melt migration beneath the East Pacific Rise, Nature, 388, 259-262, 1997.

Dziak, R., C. Fox and A. Schreiner, The June-July 1993 seismo-acoustic event at CoAxial Segment, Juan de Fuca Ridge: Evidence for a lateral dike injection, Geophys. Res. Lett. 22, 135-138, 1995.

Einarsson, P., S-wave shadows in the Krafla caldera in NE Iceland: Evidence for a magma chamber in the crust, Bull. Volcanol. 41, 1-9, 1978.

Embley, R., K. Murphy and C. Fox, High-resolution studies of the summit of Axial Volcano, J. Geophys. Res. 95, 12785-12812, 1990.

Fox, C., Evidence of active ground deformation on the mid-ocean ridge: Axial Seamount, Juan de Fuca Ridge, April-June 1998, J. Geophys. Res. 12813-12822, 1990.

Fox, C., W.E. Radford, R. Dziak, T.K. Lau, H. Matsumoto and A.E. Schreiner, Acoustic detection of a seafloor spreading episode on the Juan de Fuca Ridge using military hydrophone arrays, Geophys. Res. Lett. 22, 131-134, 1995.

Harding, A.J., J.A. Orcutt, M.E. Kappus, E.E. Vera, R.S. Detrick, and T.M. Brocher, Structure of young oceanic crust at 13°N on the East Pacific Rise from expanding spread profiles, J. Geophys. Res., 94, 12163-12196, 1989.

Hildebrand, J., J.M. Stevenson, P.T.C. Hammer, M.A. Zumberge, R.L. Parker, C.G. Fox and P.J. Meis, A Seafloor and Sea Surface Gravity Survey of Axial Volcano, J. Geophys. Res., 95, 12751-12763, 1990.

Hu, G., W. Menke and S. Rognvaldsson, A demonstration of the joint use of P wave polarization and travel-time data in tomographic inversion: crustal; velocity structure near the South Iceland Lowland Network, Geophys. Res. Lett. 20, 1407-1410, 1993.

Menke, W. and D. Sparks, Crustal accretion model for Iceland predicts 'cold' crust, Geophys. Res. Lett. 22, 1673-1676, 1995.

Menke, W., M. West, B. Brandsdottir and D. Sparks, Along-strike seismic structure of the Northern Volcanic Zone, Iceland, submitted to Geophysical. J. Int., 1997. Phipps-Morgan, J. and Y. Chen, The genesis of oceanic crust, magma injection, hydrothermal circulation and crustal flow, J. Geophys. Res. 98, 6283-6298, 1997.

Phipps Morgan, J., and Y.J. Chen, Dependence of ridge-axis morphology and geochemistry on magma supply and spreading rate, Nature , 364, 706-708, 1993.

Sohn, R., S. Webb and J. Hildebrand, Post-eruption microseismicity on the CoAxial segment, Juan de Fuca Ridge, EOS Trans. AGU, 77, 46, 1996.

Tolstoy, M., F. Vernon and J.A. Orcutt, A Seafloor Eruption Caught on Disk? Ocean Bottom Seismic & Tilt Data recorded on Axial Seamount, Abstract, RIDGE/NORDVULK/NorFA Summer School on Active Processes at Mid-Ocean Ridges, Iceland, Aug. 25 - Sept. 5, 1997.

Vera, E., J. Mutter, P. Buhl, J. Orcutt, A. Harding, M. Kappus, R. Detrick, and T. Broecher, The structure of 0 to 0.2 m.y. old oceanic crust at 9 degrees North on the East Pacific Rise from expanding spread profiles, J. Geophys. Res. 95, 15529-15556, 1990.

Figures

Fig. 1. Shaded relief map of Axial Volcano and its environs. Note the caldera just southwest of (46N, 130W). (Color Version, PostScript Version). From the Ryan et al. Ridge Multibeam Data Synthesis:

http://imager.ldeo.columbia.edu/ridgembs/ne_pac/html/home.html

Fig. 2. Webb & Sohn's OGS/OBH locations (numbered circles) and our proposed airgun tracks (line segments) shown together Axial Volcano bathymetry (contours). The 40x40 study area (dotted) is also shown. (PostScript Version, Freehand Version).

Fig. 3. Ray coverage for line A3, which crosses the caldera at X=50 km. (PostScript Version).

Fig. 4. Undershooting the magma chamber using line A3 and station S1, at the northeastern end of the line. (Top)Ray diagram for a hypothetcial velocity structure with a low velocity magma chamber underlain by a higher velocity dome. (Bottom) Corresponding reduced traveltime curve. Note that the Pg and PmP phases are delayed by about 0.2s as they pass through the magma chamber. (PostScript Versions of Top, and Bottom).

Fig. 5. Four records sections from the recent FIRE experiment in northeasten Iceland, in which the Krafla magma chamber is undershot by Pg waves (Brandsdottir et al., 1997). Note the low amplitudes - caused by defocusing by the high velocity dome - as the waves cross the magma chamber (labeled V), and the secondary arrivals due to waves traversing the melt. These data are similar to data expected for airgun lines A3 and B2, except that the role of shots and receiver are interchanged (i.e. FIRE had arrays of receivers crossing the caldera and a few shots; the Piggyback Experiment will have a few receivers and arrays of shots. (PostScript Version).

Fig. 6. 2D model of the compressional velocity structure of Krafla volcano, along a ridge-parallel line (Brandsdottir et al., 1997). The rims of the caldera are labeled R, the position of the magma chamber, V). Note the shallow magma chamber (white), mid-crustal dome of high-velocity cumulates, and rise in Moho beneath the caldera. (PostScript Version).

Fig. 7. Traveltime anomalies (left) and corresponding map (right) for a shot in eastern Iceland (i.e. east of the Krafla volcano) observed on a seismometer array to its west (Menke et al., 1997). This is a fan geometry, with rays from the shot traversing the three volcanos (shaded patches) of the neovolcanic zone (grey lines) at a depth of 8-10 km. Note that the traveltimes are anomalously early for rays passing beneath Krafla, because of the high-velocity cumulate dome. The other two central volcanos have no distinct dome, suggesting that Krafla is the most magmatically productive. These data are similar to data expected for airgun line C1 (except that the role of shots and receivers are interchanged). (PostScript Version).

Supporting Documentation