Aseismic Deep Slab and Mantle Flow Beneath Alaska: Insight From Anisotropic Tomography

Abstract

We present high‐resolution 3‐D images of P wave velocity (Vp), azimuthal anisotropy (AAN), and radial anisotropy (RAN) down to 900‐km depth beneath Alaska obtained by inverting a large number of high‐quality arrival time data from local earthquakes and teleseismic events simultaneously. Our results show that the high‐Vp Pacific slab has subducted down to 450‐ to 500‐km depths. A prominent slab gap is revealed at depths of 65–120 km near the Wrangell volcanic field, which is likely a slab tear acting as a channel that provides ascending mantle materials to generate magmas feeding the surface volcanoes. In the back‐arc mantle wedge near the eastern slab edge, the AAN exhibits trench‐parallel fast‐velocity directions (FVDs), which may reflect along‐strike mantle flow. The FVDs in the subducting Pacific slab are nearly east‐west, which may indicate fossil anisotropy formed at the mid‐ocean ridge. A negative RAN is revealed within the subducting slab, which may be caused by the fast plate subduction with a steep dip angle. Trench‐normal FVDs of the AAN are revealed in the mantle below the Pacific slab, which may reflect mantle flow entrained by the subducting slab. A positive RAN is revealed in the mantle beneath the Yakutat slab, indicating that its shallow subduction flattens the mantle flow below the slab to be subhorizontal. Along‐strike FVDs of the AAN around the eastern slab edge may indicate the edge‐induced toroidal mantle flow.

Structural Heterogeneity in Source Zones of the 2018 Anchorage Intraslab Earthquake and the 1964 Alaska Megathrust Earthquake

“…Our results reveal strong crustal heterogeneities in the overriding NA plate. A prominent low‐velocity (low‐V) anomaly is visible in the upper crust below the Cook Inlet basin in both Vp and Vs images (Figures a and d), which reflects thick sediments as also revealed by previous tomographic studies (Eberhart‐Phillips et al, ; Gou, Zhao, et al, ; Martin‐Short et al, ; Qi et al, ; Tian & Zhao, ; Ward & Lin, ; Zhao et al, ) and geological investigations (Brims, ; Plafker et al, ). East of the Cook Inlet, the crustal materials exhibit predominately a high‐velocity (high‐V) feature beneath the Kenai and Chugach Mountains (Figures a and d), especially in the Vp tomographic result, which may reflect turbidites, granitic plutons, and oceanic volcanic rocks of the Chugach terrane (Eberhart‐Phillips et al, ; Nilsen & Zuffa, ).…”

“…Our tomographic results do not reveal any visible velocity anomaly representing the subducted Yakutat terrane (i.e., the Yakutat slab) overlying the Pacific slab within the flat subduction area, which is characterized by an apparent thick low‐V and high‐Vp/Vs layer revealed by the previous studies (Bauer et al, ; Eberhart‐Phillips et al, ; Ferris et al, ; Gou, Zhao, et al, ; Y. Kim et al, ; Qi et al, ; Rondenay et al, ; Rossi et al, ). However, when we arranged a low‐V layer atop the subducting slab in a synthetic resolution test, it can be well recovered in both the Vp and Vs tomography (Figure S27).…”

“…In addition to the Moho, we also consider two crustal discontinuities modified from the CRUST1.0 model (Laske et al, ) and finally constructed a three‐layer crustal model (Figures S3 and S4). To date, several models for the geometry of the subducting Pacific slab beneath Alaska have been proposed (e.g., Gou, Zhao, et al, ; Hayes et al, ; Jadamec & Billen, ; J. Li et al, ; Zhao et al, ), but these slab models are different from each other, especially in the fore‐arc area of south‐central Alaska. In this study, we adopt the slab model produced by J. Li et al () in the starting model for the tomographic inversion, which was inferred from the Wadati‐Benioff zone seismicity located precisely using the double‐difference earthquake location method (Waldhauser & Ellsworth, ).…”

“…The fault‐bounded composite oceanic‐continental Yakutat terrane has been colliding with the NA plate since about 30 Ma (Lease et al, ; Plafker et al, ) and subducted along with the Pacific plate at a similar convergence rate of ~45.6 mm/year (Fletcher & Freymueller, ). Previous tomographic studies (Eberhart‐Phillips et al, ; Gou et al, ; Qi et al, ; Tian & Zhao, ) and receiver function analyses (Bauer et al, ; Ferris et al, ; Y. Kim et al, ; Rondenay et al, ; Rossi et al, ) have depicted the subducted portion of the Yakutat terrane as a thick layer upon the subducting Pacific slab with a lower velocity and a higher Vp/Vs than the surrounding mantle. In south‐central Alaska, the interaction between the Yakutat terrane and the NA plate plays an important role in orogenesis and exhumation (Enkelmann et al, ; Koons et al, ), volcanism (Richter et al, ; Rondenay et al, ), and interplate coupling (Zweck et al, ).…”

“…The seafloor age is from Müller et al (), whose scale is shown atop (a). The thin blue dashed lines in (a) denote depth contours of the subducting Pacific slab (Gou, Zhao, et al, ). The red (Ohta et al, ), green (Fu & Freymueller, ), and purple (Wei et al, ) dashed lines in (b) denote three areas where long‐term slow slip events occurred.…”