Travel

Velocity structure from forward modeling of the eastern ridge-transform intersection area of the Clipperton Fracture Zone, East Pacific Rise

Description
Velocity structure from forward modeling of the eastern ridge-transform intersection area of the Clipperton Fracture Zone, East Pacific Rise
Categories
Published
of 18
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 102, NO. B4, PAGES 7803-7820, APRIL 10, 1997 Velocity structure from forward modeling of the eastern ridge-transform intersection area of the Clipperton Fracture Zone, East Pacific Rise Michael L. Begnaud and James S. McClain Department of Geology, University of California, Davis Ginger A. Barth Department of Geophysics, Stanford University, Stanford, California John A. Orcutt and Alistair J. Harding Scripps nstitution of Oceanography, a Jolla, California Abstract. In the spring of 1994, we undertook an extensive geophysical study of the Clipperton Fracture Zone (FZ) on the fast spreading ast Pacific Rise. The Clipperton Area Seismic Study to Investigate Compensation xperiment CLASSIC) included surveys o examine the deep structures ssociated with the fracture zone and adjacent northern idge segment. In this paper, we report he results rom five seis- mic profiles acquired over the eastern idge-transform ntersection RTI), including profiles over the RTI high, the northern idge segment, nd the eastern ransform region. The travel time data for crustal phases, Moho reflections, and mantle phases were modeled using two-dimensional ay tracing. Seismic profiles reveal that the crust s similar in thickness north and south of the Clipperton FZ, despite differences in axial topography hat have previously been nterpreted n terms of differences n magma supply. When compared o older crust, he northern idge axis is character- ized by lower seismic velocities and higher attenuation. n our model, a low-velocity zone exists beneath he ridge axis, probably associated ith a zone of partial melt and/or very high temperatures. Within the transform one, we find that the southeastern rough s underlain by nearly normal crustal structure. The crust s slightly thinner than the adjacent aseismic extension but not enough o compensate for the depths of the trough. Toward the RTI, the trough s replaced by an intersec- tion high which appears underlain by a thickened crust, and a thicker upper crustal section. Both characteristics ndicate hat the intersection igh is a volcanic eature produced y excess olcanism t the intersection. The volcanism cts o "fill in" the transform rough, creating he thicker crust hat extends under he eastern aseismic extension of the transform. Our results show that the northern ridge segment, often identified as magma-starved, isplays he crustal hickness nd apparent signal atten- uation characteristic f a plentiful, but perhaps pisodic, magma supply. Introduction Ever since technologies have existed to map the seafloor, mid-ocean idge racture ones ave been viewed as one of the more distinctive tectonic environments. Wilson [1965] first modeled racture zones and their vari- ous features (i.e., seismicity, geometry, distribution) as being he result of strike-slip aults hat offset mid-ocean ridge segments. However, subsequent tudies f fracture zones evealed hat their complicated morphologies ould not be explained y simple kinematic models i.e., Heezen et al., 1964; Van Andel et al., 1967]. The severe bathy- Copyright 1997 by the American Geophysical Union. Paper number 96JB03393. 0148-0227/97/96JB-03393 $09.00 metric variations at fracture zones suggest omplex defor- mational histories [Fox and Gallo, 1984]. Transform aults (defined as the section of a fracture zone that offsets a mid-ocean idge) are thought o affect nearby crustal structures on mid-ocean ridges thermally, tectoni- cally, and petrologically. Thermal effects result from the juxtaposition of cooler, older lithosphere next to young lithosphere. The colder plate could affect seafloor spread- ing and retard the formation of crust [Sleep and Biehler, 1970]. Consistent with this, the results of the majority of studies ndicate hat the crust s thinner within and adjacent to transform aults and their aseismic extensions long the Mid-Atlantic Ridge (MAR) [Detrick and Purdy, 1980; Cormier et al., 1984; Sinha and Louden, 1983; Tolstoy et al., 1993; Detrick et al., 1993a] and the East Pacific Rise (EPR) [McClain and Lewis, 1980; Ouchi et al., 1982; Trdhu and Purdy, 1984]. However, some studies on the 7803  7804 BEGNAUD ET AL.: VELOCITY STRUCTURE, CLIPPERTON FRACTURE ZONE EPR have suggested he crust may thicken near the inter- section of the ridge and the fracture zone FZ) [Barth, 1994; Barth et al., 1994]. The difference n crustal emperatures or ages has also been proposed o create nomalous opog- raphy by differential ubsidence r bending n either side of the FZ [Sandwell and Schubert, 1982; Parmentier and Haxby, 1986]. In addition o the effects of the age difference cross he fracture one, strike-slip ectonics will produce ubstantial bathymetric nd perhaps tructural ariability. The active transform is not necessarily a single strike-slip fault. More likely, it comprises everal smaller, en echelon aults that create ocalized compression r dilatation ones, ead- ing to faulting [Bonatti, 1978]. Furthermore, changes n plate motion may lead to compression/dilatation cross transform one. Carbotte and Macdonald [1992] suggest that such a change n plate motion occurred between he Cocos and Pacific Plates and hat compression s presently occurring along the Clipperton Transform. Fault slip within the transform zone (TZ) may also ead to fractures that act as conduits or water penetration nto the upper mantle, causing erpentinization. he low-density, erpen- tinized mantle would hen produce iapiric uplift within he FZ [Bonatti, 1976]. To date, most fracture zone structural studies have con- centrated on the slow spreading MAR rather han the fast spreading EPR. Geophysical studies on the EPR have concentrated on the ridge axis itself [e.g., Ewing and Purdy, 1982; Toomey et al., 1994; Kappus et al., 1995] and axial magma chambers or low-velocity zones [e.g., McClain et al., 1985; Detrick et al., 1987; Harding et al., 1989; Toomey et al., 1990; Detrick et al., 1993b]. The few studies that focused on the transform zones and/or their aseismic xtensions n detail provide evidence or both hin or normal crust [McClain and Lewis, 1980; Ouchi et al., 1982; Trdhu and Purdy, 1984; McClain and Wright, 1990] and thickened crust [Barth, 1994; Barth et al., 1994]. Crustal hicknesses t EPR fracture zones ange rom -4.0 to-6.5 km with widely ranging crustal and relatively ow Moho (7.8 km/s) velocities. During April and May 1994, a detailed geophysical study, the Clipperton Area Seismic Study to Investigate Compensation CLASSIC) experiment, was conducted n and around the morphologically complex Clipperton Fracture Zone (CFZ) on the EPR (Plate 1). The primary goal of the CLASSIC experiment as o examine rustal structure using seismic omography nd thus to answer various questions oncerning he relationship etween topography, nternal structure, nd the compensation mechanisms f the transform. These questions nclude 1) how does crustal structure hange n proximity o a fast spreading Z; (2) what s the nternal tructure ssociated with the complex bathymetry f the area, mainly, the trough and ridge morphologies bserved t the FZ; (3) what are the processes ontrolling he formation f the ridge-transform ntersection RTI) highs t fracture ones; and 4) what are the differences hat occur between rust generated n he idge egments orth nd outh f the rac- ture zone? The experiment tilized ocean bottom seis- mometers OBS), ocean bottom hydrophones OBH), and air gun sources n two separate eployments, he irst con- centrating n he central western ransform egion, nd he second on the eastern RTI (Plate 1). In this paper we present an analysis of data collected during he second deployment. We have completed wo-dimensional orward modeling on selected hot rofiles n and near he eastern TI area. We rely on this approach s a first step because t permits direct omparison f our esults with he orward modeling used in almost all previous fracture zone studies. Furthermore, xisting omographic nversion schemes re inadequate n regions f complex athymetry, uch s he CFZ. The resulting wo-dimensional odels will be used as starting models or a future hree-dimensional elocity inversion of the entire data set. Regional Geology The Clipperton Fracture Zone is located on the Pacific- Cocos plate boundary t 10ø15'N atitude along he EPR, a fast spreading enter with a spreading ate during he last 0.75 m.y. of 106 mm/yr north and 112 mm/yr directly south of the FZ [Carbotte and Macdonald, 1992] (Plate 1). The transform offsets he EPR by 85 km in a right-lateral sense: he only right-lateral offset along he northern PR. This offset corresponds o a -1.5 m.y. age difference, x- posing he EPR axis to a -12 km thick ithosphere Gallo et al. [1986], using model of Parker and Oldenburg [1973]). The CFZ extends 4000 km across he Pacific Plate and has persisted n its present configuration or 9 m.y. [Klitgord and Mammerickx, 1982]. The Clipperton transform zone is dominated by a median ridge (with minimum depths of-2300-2500 m), with troughs to the north and south (-3850 m depth) (Plate 1), and is as much as -25 km in width between trough edges. This ridge and trough morphology s common along arge-offset ransforms f the EPR, typi- cally generating hundreds o over a thousand meters of relief [Fox and Gallo, 1984]. The adjacent EPR ridge segments, ffset by the CFZ, have different chemistries Langmuir et al., 1986; Batiza and Niu, 1992] and morphologies Macdonald t al., 1984; Gallo et al., 1986]. The East Pacific Rise to the north of the FZ is relatively narrow and deep, with a width of 1 km at the 2900 m contour and an axial high that reaches a depth of only 2800 m (Plate 1). Gallo et al. [1986] and Kastens t al. [ 1986] suggest hat the relatively great depth of the northern xis s a result of its magma-starved ature. In addition, Detrick et al. [1987], during a seismic eflec- tion survey, did not detect an axial magma hamber eflec- tor for 70 km under he EPR axis north of the RTI high. In contrast o the northern segment, he southern EPR segment s broader nd shallower, aving a width greater than 10 km at the 2900 m level and a bathymetric high that reaches depth of less han 2600 m (Plate 1). The segment s smooth nd flat, varying n relief by only -20 m over lateral distances of tens of kilometers [Gallo et al., 1986]. This distinctive morphology suggests hat the southern egment s in a major phase of basaltic melt genesis Gallo et al., 1986; Kastens et al., 1986]. An axial magma chamber eflector was detected eneath he southern egment y Detrick et al. [1987] at approximately 1.5 km depth, xtending orthward o within several ilo- meters of the CFZ [Barth et al., 1994; Kent et al., 1993].  BEGNAUD ET AL.' VELOCITY STRUCTURE, CLIPPERTON FRACTURE ZONE 7805 , -25ON PLAIE I oBozCO PACIFIC PLATE s•c3jm•os 110øW 105øW ,8, 7 104 ø 40'W 104 ø 20'W , 4 0 104 ø O0'W 10 ø 50'N 10 ø 40'N 10 ø 30'N 10 ø 20'N 10 ø 10'N 10 ø O0'N 9 ø 50'N D•p•h(m) Plate 1. Bathymetric map of the Clipperton Fracture Zone, East Pacific Rise (EPR) at 10ø15'N dis- playing seismic efraction shot ine tracks and nstrument ocations gridded bathymetry was obtained by merging data sets rom Macdonald et al. [1992] with data acquired during the CLASSIC experiment). Contour interval of 100 m, color change every 50 m. Solid squares designate ocean bottom seis- mometer (OBS) locations. Solid circle designates ocean bottom hydrophone (OBH) location. Instrument abbreviations re as follows' S, OBS-Sharyn; , OBS-Janice; P, OBS-Phred; K, OBS-Karen; L, OBS-Lynn; Jy, OBS-Judy; nd N, OBH-NightTrain created sing generic mapping ools Wessel nd Smith, 1991]). Inset is a regional map of the northern EPR highlighting study area for the CLASSIC experiment adapted rom Gallo et al. [1986]). A volcanic eruption occurred n 1991 along the EPR [Haymen et al., 1991], south of the CFZ. All of these observations re consistent with the hypothesis hat the southern segment s in a magmatically obust phase of activity. Topographic ighs are observed n the older crust adja- cent to each ridge transform intersection. Suggested hypotheses or formation of these RTI highs nclude (1) expansion nd uplift of older crust due o lateral conductive heat flow from the younger crust across he FZ [Louden and Forsyth, 1976; Forsyth and Wilson, 1984]; 2) magma traveling along he ridge and damming at the young side of the RTI [Vogt and Johnson, 1975] and/or erupting hrough the transform fault [Menard and Atwater, 1968, 1969; Gallo et al., 1986]; (3) a nearly continuous eakage of magma from the ridge axis into the older crust [Gallo et al., 1986; Kastens et al., 1986]; and (4) episodic and tem- porary propagation of the ridge tip across he FZ [Fox and Gallo, 1984; Hey et al., 1980, 1986]. Forward modeling of the velocity structure n and around he RTI high can determine where areas of lower velocities occur near the RTI, suggesting ones of higher emperatures.  7806 BEGNAUD ET AL.: VELOCITY STRUCTURE, CLIPPERTON FRACTURE ZONE Experiment The CLASSIC experiment as a cooperative ffort between he University of California, Davis, and the Scripps nstitution f Oceanography SIO). Using he R/V Maurice Ewing cruise W9405), he experiment n- compassed roughly 150x110 km 2 area centered n the CFZ (Plate 1). This paper resents esults rom a deployment entered near he eastern TI high where ix OBSs out of seven) and one OBH (out of five) retrieved sable ata. The OBSs were microprocessor-based cean bottom seismometers from SIO [Moore et al., 1981; Willoughby t al., 1993]. These nstruments ontinuously ecorded ne vertical, wo horizontal, nd one hydrophone omponent t 128 samples per second. The OBHs were seafloor ata oggers evel- oped at SIO to continuously ecord hydrophone ata van Avendonk et al., 1995]. The seismic ource as a tuned rray of 10 air guns with a total gun volume of ~65.5 L (4000 cubic nches). The guns were owed behind he ship at ~ 10 m depth and ired at 100-s ntervals uring he second eployment iving a shot spacing of roughly 230-240 m. Shot ntervals of less than 100 s were ound o be too short; he resulting ever- beration ersisted ithin he water column, ncreasing ig- nificantly the noise levels at the seafloor instruments. Shot and receiver locations were determined from Global Positioning ystem GPS) measurements board hip. We shot a total of 15 separate ines during he second deployment, ive of which have been modeled and will be discussed n this paper. We chose hese ive shot ines (Plate 1) because nstruments ere ocated irectly on the lines, allowing wo-dimensional 2-D) modeling or lines with two or more instruments. Lines 27 and 28 were east- west lines and crossed he ridge axis north of the RTI. Only one instrument was located under each of these profiles; hus, he ines were nsufficient or 2-D analysis. Instead, we undertook ne-dimensional l-D) analysis o examine the structure east and west of the ridge axis. Lines 30 and 32, running parallel o the trend of the frac- ture zone, had three and two instruments, espectively. Line 22 ran north-south irectly on the northern idge seg- ment and extended 50 km south of the RTI. Four instru- ments were ocated along his profile. Record Sections and Arrival Time Data Data display and arrival ime picking were done using Seismic Unix [Cohen and Stockwell, 1996]. Example record sections for the five modeled lines are shown in Figures 1 and 2. P wave arrivals, as well as any observable n and Pm? arrivals, were visually icked rom 4277 shot-receiver airs along he five lines. We estimated picking rror or each trace by visually determining he extreme pper and ower bounds on the arrival time. Although hey could only constrain -D models, we in- clude ata or ines 7 and 8 to compare rustal tructure north ith hat within nd outh f he CFZ. Record ec- tions or the lines (Figures a and lb) reveal hat arrivals on line 27, %25.5 km north of the RTI, are earlier and clearer han or line 28, -9.5 km north of the RTI. The smaller amplitudes esult n larger picking errors or line 28 (average ~+60 ms) than or line 27 (average +50 ms). The seismic traces from west of OBS-Janice on line 28 display smaller amplitudes han races rom the east, and no Pn arrivals can be seen west of the axis. Overall, the record sections or line 32 (over older crust ~7 km south of the FZ) display some of the clearest arrivals with picking errors averaging ~+40 ms for OBS- Judy and OBH-NightTrain (Figure lc). Data for line 30 along he FZ also have similarly distinct arrivals Table 1). Line 22 included segments ver young (0 Ma) crust o the north of the RTI and older crust (1.5 Ma) to the south; the data appear o reflect this change n age. Record sec- tions rom the four nstruments n this ine indicate higher attenuation of the seismic signals along the ridge axis compared o signals ropagating hrough older crust south of the RTI high (Figure 2). There are a number of possible easons or the enhanced attenuation ver the ridge axis. One is scattering rom the seafloor along the axis. However, we have no knowledge that he axis seafloor s any rougher han other nearby oca- tions. Another, more likely, explanation or the attenua- tion is higher than normal temperatures nd perhaps he presence of melt under the ridge axis [Wilcock et al., 1992]. The errors n picking he arrivals were affected by the attenuation of the source signals or some of the instru- ments. We attempted o pick to as great a distance s possible, djusting he picking errors o account or uncer- tainties aused y the weak arrivals Table 1). Forward Modeling The arrival ime data were modeled sing MacRay, an interactive 2-D seismic ray tracing program for the Macintosh M [Luetgert, 1988, 1992] based on the ray method of Cerveny t al. [1977]. The OBS and OBH loca- tions were used as source points with rays exiting he model at the ocean urface depth of 0). Both efracted nd reflected ays were allowed n the models. The 2-D veloc- ity models were determined y interfaces hat extended across he model (Figure 3). Each pair of interfaces bounded layer, within which a grid of vertical ines was placed. Throughout he model, velocities were continu- ously varied based n values pecified herever he vertical grid ntersected n nterface. The starting elocity models or the Clipperton ata sets included nine nterfaces Figure 3). The first, second, nd third were associated with the velocity structure n the water column. Water column velocities were calculated from expendable athythermograph oundings ade during the CLASSIC experiment. The topography f the fourth interface, he seafloor, was acquired rom decimated to a ~600-m nterval) Hydrosweep M data also collected uring the experiment. The remaining ive interfaces nd ayers were associated with the crust and upper mantle. In the initial models, he long-wavelength eatures f the crustal nterfaces mimicked the seafloor opography e.g., Figure 3). We constructed simple homogeneous ayer solution o the observed ata o provide an initial crustal velocity model or each ine. The initial models were further constrained to include continu- ously varying velocities using upper crustal values consis-  BEGNAUD ET AL.' VELOCITY STRUCTURE, CLIPPERTON FRACTURE ZONE 7807 a) W Line 27 E 4.0 3.5 3.0 2.5 20 0 -20 w Line 28 , , 1 0 E 4.0 3.5 3.0 2.5 c) w Line 32 -20 E BH,Night•rain l 0 - 3.5 w Line 30 , , E 3.0 •, 2.5 -40 -20 0 20 40 Distance From Instrument (km) Figure 1. Record sections nd observed icks for east-west rofiles (lines 27, 28, 32, and 30) north of the RTI. Observed arrivals are plotted with crosses. For clarity, every other seismic race and observed rrival re plotted. ach race s multiplied y Offsetl /2. nstruments re ocated t zero off- set, with negative offsets when shooting while traveling away from the instrument. (a) Record section for vertical channel of OBS-Sharyn 34.88 km from beginning f line) on line 27. The data gap was a result of the instrument ailing to record between an offset nterval of 3.29-12.91 km. (b) Record sec- tion for vertical channel of OBS-Janice. Data were ow-pass iltered at 1-21 Hz. (c) Record section rom one of two instruments OBH-NightTrain) on line 32. Data were band-pass iltered at 3-50 Hz to remove ow-frequency noise associated ith the OBH instruments. d) Record section or vertical channel rom one of three nstruments OBS-Karen) on line 30. tent with previous studies e.g., Vera et al., 1990; Detrick et al., 1993b; Kappus et al., 1995]. The velocity model was then teratively modified until the observed nd calcu- lated data sets displayed qualitative match. Since we were creating 2-D models from three- dimensional (3-D) data, off-profile bathymetric effects needed o be considered. The only area n our study egion where off-profile ffects were a concern as near ine 30 in
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks