TY - JOUR
T1 - Constraining Mantle Viscosity Structure From a Statistical Analysis of Slab Stagnation Events
AU - Wang, Yongming
AU - Li, Mingming
N1 - Funding Information:
We thank three anonymous reviewers and Editor Thorsten Becker for their insightful and constructive reviews and comments that significantly improve the paper. The work is supported by NSF grant EAR‐1849949 and EAR‐1855624. The geodynamic models were performed on the Agave supercomputer at Arizona State University. Figures are drawn using the Generic Mapping Tools (GMT, www.soest.hawaii.edu/gmt/ ). The seismic tomography models are available at the websites http://d-earth.jamstec.go.jp/GAP_P4/ (GAP_P4), http://ds.iris.edu/spud/earthmodel (LLNL‐G3Dv3 and HMSL‐P06), http://seismo.berkeley.edu/wiki_br/Main_Page (SEMUCB‐WM1) and https://jritsema.earth.lsa.umich.edu (S40RTS and SP12RTS), respectively.
Funding Information:
We thank three anonymous reviewers and Editor Thorsten Becker for their insightful and constructive reviews and comments that significantly improve the paper. The work is supported by NSF grant EAR-1849949 and EAR-1855624. The geodynamic models were performed on the Agave supercomputer at Arizona State University. Figures are drawn using the Generic Mapping Tools (GMT, www.soest.hawaii.edu/gmt/). The seismic tomography models are available at the websites http://d-earth.jamstec.go.jp/GAP_P4/(GAP_P4), http://ds.iris.edu/spud/earthmodel (LLNL-G3Dv3 and HMSL-P06), http://seismo.berkeley.edu/wiki_br/Main_Page (SEMUCB-WM1) and https://jritsema.earth.lsa.umich.edu (S40RTS and SP12RTS), respectively.
Publisher Copyright:
© 2020. American Geophysical Union. All Rights Reserved.
PY - 2020/11
Y1 - 2020/11
N2 - The viscosity structure of Earth's mantle, even the 1-D radial viscosity profile, remains not well constrained. The dynamics of the subducting slabs is strongly affected by, and can be used to constrain, the viscosity structure of the mantle. Here, we perform fully dynamic, self-consistent mantle convection models to study the dynamics of subducted slabs in the deep mantle. We use a statistical analysis approach to quantify how the depth distribution of flat-lying slabs is affected by the depth-dependence of mantle viscosity. We find that, for cases in which the viscosity increases at 660 km depth, whether sharply or gradually, flat-lying slabs preferentially occur above this depth, and importantly, up to ∼30% of the subducted slabs previously flatted at this depth later sink to the deep lower mantle and maintain a flat-lying morphology. The frequency of (or the probability to have) flat-lying slabs at ∼1,000 km depth in these cases is similar to cases in which the viscosity jump occurs at 1,000 km depth. Therefore, to explain the presence of flat-lying slabs at ∼1,000 km depth for the Earth does not require a viscosity jump at this depth. In contrast, a viscosity jump merely occurring at ∼1,000 km depth causes a lack of flat-lying slabs in the uppermost lower mantle at ∼700–900 km depth and is inconsistent with seismic observations. The presence of flat-lying slab materials in the Earth's uppermost lower mantle requires a viscosity increase at 660 km depth.
AB - The viscosity structure of Earth's mantle, even the 1-D radial viscosity profile, remains not well constrained. The dynamics of the subducting slabs is strongly affected by, and can be used to constrain, the viscosity structure of the mantle. Here, we perform fully dynamic, self-consistent mantle convection models to study the dynamics of subducted slabs in the deep mantle. We use a statistical analysis approach to quantify how the depth distribution of flat-lying slabs is affected by the depth-dependence of mantle viscosity. We find that, for cases in which the viscosity increases at 660 km depth, whether sharply or gradually, flat-lying slabs preferentially occur above this depth, and importantly, up to ∼30% of the subducted slabs previously flatted at this depth later sink to the deep lower mantle and maintain a flat-lying morphology. The frequency of (or the probability to have) flat-lying slabs at ∼1,000 km depth in these cases is similar to cases in which the viscosity jump occurs at 1,000 km depth. Therefore, to explain the presence of flat-lying slabs at ∼1,000 km depth for the Earth does not require a viscosity jump at this depth. In contrast, a viscosity jump merely occurring at ∼1,000 km depth causes a lack of flat-lying slabs in the uppermost lower mantle at ∼700–900 km depth and is inconsistent with seismic observations. The presence of flat-lying slab materials in the Earth's uppermost lower mantle requires a viscosity increase at 660 km depth.
KW - mantle viscosity structure
KW - numerical modeling
KW - self-consistent model
KW - slab dynamics
KW - statistical analysis
UR - http://www.scopus.com/inward/record.url?scp=85096408825&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85096408825&partnerID=8YFLogxK
U2 - 10.1029/2020GC009286
DO - 10.1029/2020GC009286
M3 - Article
AN - SCOPUS:85096408825
SN - 1525-2027
VL - 21
JO - Geochemistry, Geophysics, Geosystems
JF - Geochemistry, Geophysics, Geosystems
IS - 11
M1 - e2020GC009286
ER -