Seismic Heterogeneity in the Upper Crust near the 1991 ... - CiteSeerX
GEOPHYSICAL RESEARCH LETTERS, VOL. 27, NO. 16, PAGES 2369-2372, AUGUST 15, 2000
Seismic heterogeneity in the upper crust eruption site on the East Pacific Rise, 9ø50'N
near
the
1991
T. Tian,1,•William S. D. Wilcock,2DouglasR. Toomey,3andRobertS. Detrick4 Abstract. We report the resultsof a small-apertureuppercrustaltomographyexperimentconductedon the East Pacific Rise near9ø50'N,the siteof a volcaniceruptionin 1991. The experimentgeometrycomprised3 ocean bottom seismome-
mary objective of imaging the lower crust and the crustmantleboundary(the Moho) in order to constrainalong-axis variationsin melt supplybeneatha fast spreadingridge. As a minor componentof the UNDERSHOOT experiment, we tersand491airgun shots firedin a 16x 16km2 area.Two- conducteda tomographicexperimentnear 9ø50'N to image and three-dimensional inversionsindicate that the average uppercrustalstructure.The experiment(Figure 1) comprised uppercrustalvelocities,bothon andoff axis,increaseby -0.2 three ocean bottom seismometers(OBSs) and 491 airgun km/s north of 9ø52'N. This location coincides with the northshots. One OBS was deployedon the ridge axis at 9ø50.5'N ern limit of the 1991 eruptionand of vigoroushydrothermal- and the other two were located 6 km to either side of the rise. ism. Our data can be explainedby a 100 m southwardin- The seismicsourcewas the 20-gun array of the R/V Maurice creasein layer 2A thickness. On axis, previouswork indi- Ewing with a total volumeof 139 liters (8,500 cubicinches). catesthatlayer2A thicknessdoesnot vary in thisregion. We The shotswere spaced300 m apart along eight 15-km-long infer that the sheeteddikes abovethe axial magmachamber shot lines that were orientedperpendicularto the ridge axis are thermallysegmentedwith the dike layer southof 9ø52'N and separatedby about 2 km. This experimentgeometry beingon averageat least300øCwarmerthanto the north. Off yields goodray samplingin the uppercrustdown to about2 axis,layer2A may be thickersouthof 9ø52'Nif a higherpro- km depth. Our data set for this studycomprisedtravel time portionof eruptionsflow off axis. picksfor over 800 P-wavefirst arrivalsthatwereidentifiedby visual methodswith an estimateduncertaintyof less than 10 ms. We haveexcludedtravel timesfor pathsthat passbelow 1. Introduction the magmachamber[Wilcocket al., 1995]. The EPR between 9 ø and 10øN is a well-studied section of We have appliedthe tomographictechniqueof Toomeyet a fast-spreading ridge, with the regionnear 9ø50'N receiving al. [1994] which employsgraphtheory [Moser, 1991] to esparticularattentionsincethe detectionin 1991 of an eruption timateray pathsand travel times and includesthe effectsof along a fourth-order segment extending from 9ø45'N to bathymetry. For this study,the slownessmodelfor the for9ø51.5'N[Haymortet al., 1993]. The eruptionsitelies on the ward problem isdefined ona 200mgridina 16x 16x 2 km3 shallowestand most inflatedportionof this pan of EPR and volumewhich is rotatedto the trend of the ridge axis and enoverliesthe hypothesizedpositionof a mantle diapir [e.g., Barth and Mutter, 1996]. Sincethe 9ø45'-9ø51.5'Nsegmentis known to be magmaticallyand hydrothermallyactive, it can be inferredthat this regionhasa high-magmabudgeton both • • \ • % •1'Study shorttimescalesof- 10-100 yearsand longertimescalesof at ', [ \ mmArea 9 55'N
least10syears [e.g.,Scheirer andMacdonald, 1993].
In this paper, we presentthe resultsof a small aperture seismictomographyexperimentnear 9ø50'N to image upper crustal structure. The results show that the upper-crustal structurechangesnear the fourth-ordersegmentboundaryat
!
9ø52'N bothon-axis andoutto agesof 10s years, themaximum agesampledby the experiment.
9 50)V
ß
2. Experiment and Methods
•etee•ß i•
eeee
ß
The 1997 UNDERSHOOT experiment[Toomey et al., 1998] was a segment-scale seismicexperimentwith the pri9 45'N
',
..,..'"
•• "•..:••2*• •High_ t' ,[ • • .
1Geophysics Program, University ofWashington, Seattle 2School ofOceanography, University ofWashington, Seattle 3Department ofGeological Sciences, University ofOregon, Eugene 4Department of Geology andGeophysics, WoodsHoleOceano-
'"
.
104 20'W
• .
• I •-
104 15 'W
Cross-Section. 7. (•gure*. . 104 10'W
Figure 1. Bathymetrymap contouredat 100 m showingthe
graphicInstitution,Woods Hole, Massachusetts
5NowatDepartment of Electrical Engineering, Arizona StateUni- configurationof the seismictomographyexperimentand the
versity,Tempe
locationof high-temperature vent sites[Haymonet al., 1991]. Dashed
Copyright2000 by the AmericanGeophysicalUnion. Papernumber 1999GL011191. 0094-8276/00/1999GL011191
lines show the locations
of the four vertical
cross-
sectionsshownin Figure 4. The insetfigure showsthe location of the experimentarea on the East PacificRise between the Clipperton(CTF) and Siqueiros(STF) transformfaults.
$05.00
2369
2370
TIAN ET AL.'
SEISMIC
HE•ROGE•I•
W
•
E
2 (a)
.-
i .....
ø.1 ........
::.
_8
_4
_2
0
2
4
6
8
,0
et al., 1994]. The travel time residuals for the two-dimensional
inversion
vary systematically along-axis(Figure 3). The 'mostprominent change occurs in the northernpart of the experimentnear Velocity, km/s 9ø52'N (y = 3 km). Here, the mean residualscalculatedfor eachacross-axisshotline decreaseby about 30-40 ms to the north, a result that suggestsslower seismicvelocitiesto the ...................................................................... • -•'*:?-:•"::-'"'"• .......... *"'• ................. south. This changecoincidesclosely with a fourth-order -6 -4 -2 0 2 4 6 8 10 segmentboundaryat 9ø51.5'N [Haymortet al., 1991] that v
Seismic heterogeneity in the upper crust eruption site on the East Pacific Rise, 9ø50'N
near
the
1991
T. Tian,1,•William S. D. Wilcock,2DouglasR. Toomey,3andRobertS. Detrick4 Abstract. We report the resultsof a small-apertureuppercrustaltomographyexperimentconductedon the East Pacific Rise near9ø50'N,the siteof a volcaniceruptionin 1991. The experimentgeometrycomprised3 ocean bottom seismome-
mary objective of imaging the lower crust and the crustmantleboundary(the Moho) in order to constrainalong-axis variationsin melt supplybeneatha fast spreadingridge. As a minor componentof the UNDERSHOOT experiment, we tersand491airgun shots firedin a 16x 16km2 area.Two- conducteda tomographicexperimentnear 9ø50'N to image and three-dimensional inversionsindicate that the average uppercrustalstructure.The experiment(Figure 1) comprised uppercrustalvelocities,bothon andoff axis,increaseby -0.2 three ocean bottom seismometers(OBSs) and 491 airgun km/s north of 9ø52'N. This location coincides with the northshots. One OBS was deployedon the ridge axis at 9ø50.5'N ern limit of the 1991 eruptionand of vigoroushydrothermal- and the other two were located 6 km to either side of the rise. ism. Our data can be explainedby a 100 m southwardin- The seismicsourcewas the 20-gun array of the R/V Maurice creasein layer 2A thickness. On axis, previouswork indi- Ewing with a total volumeof 139 liters (8,500 cubicinches). catesthatlayer2A thicknessdoesnot vary in thisregion. We The shotswere spaced300 m apart along eight 15-km-long infer that the sheeteddikes abovethe axial magmachamber shot lines that were orientedperpendicularto the ridge axis are thermallysegmentedwith the dike layer southof 9ø52'N and separatedby about 2 km. This experimentgeometry beingon averageat least300øCwarmerthanto the north. Off yields goodray samplingin the uppercrustdown to about2 axis,layer2A may be thickersouthof 9ø52'Nif a higherpro- km depth. Our data set for this studycomprisedtravel time portionof eruptionsflow off axis. picksfor over 800 P-wavefirst arrivalsthatwereidentifiedby visual methodswith an estimateduncertaintyof less than 10 ms. We haveexcludedtravel timesfor pathsthat passbelow 1. Introduction the magmachamber[Wilcocket al., 1995]. The EPR between 9 ø and 10øN is a well-studied section of We have appliedthe tomographictechniqueof Toomeyet a fast-spreading ridge, with the regionnear 9ø50'N receiving al. [1994] which employsgraphtheory [Moser, 1991] to esparticularattentionsincethe detectionin 1991 of an eruption timateray pathsand travel times and includesthe effectsof along a fourth-order segment extending from 9ø45'N to bathymetry. For this study,the slownessmodelfor the for9ø51.5'N[Haymortet al., 1993]. The eruptionsitelies on the ward problem isdefined ona 200mgridina 16x 16x 2 km3 shallowestand most inflatedportionof this pan of EPR and volumewhich is rotatedto the trend of the ridge axis and enoverliesthe hypothesizedpositionof a mantle diapir [e.g., Barth and Mutter, 1996]. Sincethe 9ø45'-9ø51.5'Nsegmentis known to be magmaticallyand hydrothermallyactive, it can be inferredthat this regionhasa high-magmabudgeton both • • \ • % •1'Study shorttimescalesof- 10-100 yearsand longertimescalesof at ', [ \ mmArea 9 55'N
least10syears [e.g.,Scheirer andMacdonald, 1993].
In this paper, we presentthe resultsof a small aperture seismictomographyexperimentnear 9ø50'N to image upper crustal structure. The results show that the upper-crustal structurechangesnear the fourth-ordersegmentboundaryat
!
9ø52'N bothon-axis andoutto agesof 10s years, themaximum agesampledby the experiment.
9 50)V
ß
2. Experiment and Methods
•etee•ß i•
eeee
ß
The 1997 UNDERSHOOT experiment[Toomey et al., 1998] was a segment-scale seismicexperimentwith the pri9 45'N
',
..,..'"
•• "•..:••2*• •High_ t' ,[ • • .
1Geophysics Program, University ofWashington, Seattle 2School ofOceanography, University ofWashington, Seattle 3Department ofGeological Sciences, University ofOregon, Eugene 4Department of Geology andGeophysics, WoodsHoleOceano-
'"
.
104 20'W
• .
• I •-
104 15 'W
Cross-Section. 7. (•gure*. . 104 10'W
Figure 1. Bathymetrymap contouredat 100 m showingthe
graphicInstitution,Woods Hole, Massachusetts
5NowatDepartment of Electrical Engineering, Arizona StateUni- configurationof the seismictomographyexperimentand the
versity,Tempe
locationof high-temperature vent sites[Haymonet al., 1991]. Dashed
Copyright2000 by the AmericanGeophysicalUnion. Papernumber 1999GL011191. 0094-8276/00/1999GL011191
lines show the locations
of the four vertical
cross-
sectionsshownin Figure 4. The insetfigure showsthe location of the experimentarea on the East PacificRise between the Clipperton(CTF) and Siqueiros(STF) transformfaults.
$05.00
2369
2370
TIAN ET AL.'
SEISMIC
HE•ROGE•I•
W
•
E
2 (a)
.-
i .....
ø.1 ........
::.
_8
_4
_2
0
2
4
6
8
,0
et al., 1994]. The travel time residuals for the two-dimensional
inversion
vary systematically along-axis(Figure 3). The 'mostprominent change occurs in the northernpart of the experimentnear Velocity, km/s 9ø52'N (y = 3 km). Here, the mean residualscalculatedfor eachacross-axisshotline decreaseby about 30-40 ms to the north, a result that suggestsslower seismicvelocitiesto the ...................................................................... • -•'*:?-:•"::-'"'"• .......... *"'• ................. south. This changecoincidesclosely with a fourth-order -6 -4 -2 0 2 4 6 8 10 segmentboundaryat 9ø51.5'N [Haymortet al., 1991] that v
