3-Dimensional Inversion of Regional P and S Arrival Times in the East Aleutians and Sources of Subduction Zone Gravity Highs

Publication Type  Journal Article
Year of Publication  1994
Authors  Abers, G. A.
Journal Title  Journal of Geophysical Research-Solid Earth
Volume  99
Issue  B3
Pages  4395-4412
Journal Date  Mar 10
ISBN Number  0148-0227
Accession Number  ISI:A1994NB20700009
Key Words  arc-continent collision; earthquake location; velocity structure; seismic velocity; deep-structure; mantle; crust; relocation; anomalies; beneath
Abstract  

Free-air gravity highs over forearcs represent a large fraction of the power in the Earth's anomalous field, yet their origin remains uncertain. Seismic velocities, as indicators of density, are estimated here as a means to compare the relative importance of upper plate sources for the gravity high with sources in the downgoing plate. P and S arrival times for local earthquakes, recorded by a seismic network in the eastern Aleutians, are inverted for three-dimensional velocity structure between the volcanic arc and the downgoing plate. A three-dimensional ray tracing scheme is used to invert the 7974 P and 6764 S arrivals for seismic velocities and hypocenters of 635 events. One-dimensional inversions show that station P residuals are systematically 0.25-0.5 s positive at stations 0-30 km north of the Aleutian volcanic arc, indicating slow material, while residuals at stations 10-30 km south of the arc are 0.1-0.25 s negative. Both features are explained in three-dimensional inversions by velocity variations at depths less than 25-35 km. Tests using a one-dimensional or a two-dimensional slab starting model show that below 100 km depth, velocities are poorly determined and trade off almost completely with hypocenters for earthquakes at these depths. The locations of forearc velocity highs, in the crust of the upper plate, correspond to the location of the gravity high between the trench and volcanic arc. Free-air anomalies, calculated from the three-dimensional velocity inversion result, match observed gravity for a linear density-velocity relationship between 0.1 and 0.3 (Mg m-3)/(km s-1), when a 50-km-thick slab is included with a density of 0.055 +/- 0.005 Mg m-3. Values outside these ranges do not match the observed gravity. The slab alone contributes one third to one half of the total 75-150 mGal amplitude of the gravity high but predicts a high that is much broader than is observed. The inclusion of upper-plate velocity anomalies predicts the correct width of the anomaly, 100-150 km, where the anomaly is most positive. Because the forearc gravity high is continuous along the entire Aleutian arc and is found in most arcs globally, high upper-plate forearc velocities are suspected to be a common feature of the upper plate of most subduction zones. The forearc mass excesses appear to be sustained by upward regional flexure of the upper plate that partly balances the depression of the lower plate at the trench, thus elevating high-density and high-velocity material. Thus a part of the downward flexure of the subducting plate is regionally compensated by shallow positive mass anomalies in the upper plate, and the strength of the upper plate helps generate the forearc gravity highs.

Notes  

Nb207Times Cited:23Cited References Count:46

URL  <Go to ISI>://A1994NB20700009