Constraints on lithosphere and mantle rheology from in-situ observations Shijie Zhong Department of Physics University of Colorado at Boulder Collaborators: Archie Paulson and John Wahr on post-glacial rebound Jereon van Hunen, Mike Ritzwoller & Nikolai Shapiro on small-scale convection Tony Watts on oceanic islands loading
Outline 1. Constraint of post-glacial rebound process on mantle viscosity (asthenosphere). 2. Constraint of Pacific seismic structure and sublithospheric small-scale convection on mantle rheology at the lithosphere-asthenosphere boundary (LAB). 3. Constraint of load-induced deformation (e.g., near Hawaii) at seafloor on lithosphere rheology.
Post-glacial rebound: The first example of ongoing isostatic compensation. Raised beaches, southeast coast of Hudson Bay
Constraints on mantle viscosity from Post-glacial rebound Haskell (1935), Cathles (1975), Peltier (1976; 1998), Mitrovica & Forte (2004), Tamisiea et al. (2007). Our goal is to combine time-varying gravity anomalies from GRACE and relative sea-level changes data. Raised beaches, southeast coast of Hudson Bay
RSL and GRACE data over Hudson Bay region From April 2002 to December 2006 (Tapley et al., 2004; Paulson et al., 2007) Dyke and Peltier (2000)
1-layer (i.e., whole mantle) viscosity inversion Predict RSL and time-varying gravity from viscoelastic loading models, using ICE 5G [Peltier, 2004] as time-dependent loads, compute misfit to the observations, and constrain mantle viscosity. Paulson et al. (2007)
2-layer viscosity inversion -- grid search Upper mantle viscosity log10 (Pas) Lower mantle viscosity log10 (Pas) Paulson et al. (2007) Paulson and Richards (2009) further explored the trade-off between layer thickness and viscosity magnitude.
3-layer viscosity inversion Paulson et al. (2007)
Remarks GRACE and RSL data from Hudson Bay region suggest: 1) mantle viscosity for a uniform mantle (i.e., a single layer) 1.4-2.3x10 21 Pa s; 2) for two-layer mantle model, minimum misfit occurs for the upper mantle and lower mantle viscosities that are 5.3x10 20 and 2.3x10 21 Pa s, respectively. However, other two-layer models work as well (i.e., trade-off issues). Long-wavelength geoid models prefer a weak upper mantle (e.g., Hager & Richards, 1989), but obtaining finer structure in the viscosity is also a challenge.
Asthenosphere-lithosphere transition thermo-mechanical aspect Activate energy E ~450 KJ/mol and n~3.5. [Karato & Jung, 2003; Hirth & Kohlstedt, 2003]. High-T flow law:
Apparent Thermal Age From Surface Wave Tomography for the Pacific Seafloor age Apparent thermal age The difference Ritzwoller et al., 2004
Thermal Age vs Lithospheric Age Ritzwoller et al., 2004
Sub-lithospheric small-scale Convection (SSC) (Richter, 1973; Buck & Parmentier, 1986; Davaille & Jaupart, 1994; Conrad & Molnar, 1999; Choblet & Sotin, 2000; Korenaga, 2003; Zaranek & Parmentier, 2004; Lee et al., 2005) T s T m T m } δ unstable Only the bottom sub-layer δ unstable with differential temperature ΔT r may go unstable and be removed. ΔT r ~ E -1 [Davaille & Jaupart, 1994].
3-D modeling of SSC and constraint on mantle rheology V plate Van Hunen et al., (2005; 2003) E~300 KJ/mol & n=3.5
Remarks The seismic structure of Pacific upper mantle, if caused by sub-lithospheric small-scale convection, suggests that the rheological activation energy be ~300 KJ/mol (n=3.5), which is noticeably smaller than experimentally determined values. The fine structure for North America upper mantle from USArray may help further constrain mantle rheological parameters. Eagar et al. (2010) Yang & Forsyth (2006)
Getting to the hard part - lithosphere rheology (large stress & low temperature) Brittle/Frictional at shallow depths Low-T rheology (Peirels stress): High-T flow law (power-law): Modified from Mei et al. (2010) Water s effect (Katayama & Karato, 2008)
Probe lithosphere rheology using seamountinduced deformation Watts & ten Brink (1989) Kauai Oahu Molokai Hawaii Watts (2001)
Controls on elastic thickness Te and stress relaxation? Hawaii Watts (2001); Watts & Zhong (2000)
The meaning of elastic thickness: high viscosity layer 30 km
Constraints of Te on lithospheric mantle rheology E=120 KJ/mol with n=1 (Watts & Zhong, 2000) Consistent with Courtnery & Beaumont (1983)
Remarks Modeling surface deflection in response to long-term (~1 Ma or longer) seamount and oceanic island loads may help constrain lithosphere rheology. Preliminary results suggest that the lithosphere appears significantly weaker than inferred from laboratory studies, thus raising the possible role of water (Katayama & Karato, 2008) or different (weaker) deformation mechanism (e.g., Demouchy et al., 2009) at lithospheric conditions.
Final remarks Asthenospheric viscosity structure (thickness and magnitude) remains a challenge. Need to resolve the trade-off issue (more, different type of data). High resolution seismic images for the upper mantle (e.g., from USArray) may provide constraints on the rheological transition at the LAB. With better seismic studies of crustal and sedimentary structures near oceanic islands (Hawaii), loading calculations can provide constraints on lithosphere rheology.