convection modeling of the postperovskite transition and constraints on the thermal conditions near...

Convection modeling of the postperovskite transition and constraints on the thermal

conditions near the core mantle boundary

Arie van den Berg, Utrecht University

David Yuen, University of Minnesota

Outline

• Seismic observations of irregular D” old and new

• Interpretation related to Pv/PPv transition

> estimate thermal conditions near CMB

• Modelling results of slab interaction with Pv/PPv

feasibility of heat flow estimates and thermal conductivity

• Conclusions

Cleary and Haddon, 1972

Van den Berg, Cloetingh and Doornbos, 1978

Hirose, 2007

Pv/PPv Slope 11 MPa/K,CMB intercept ~3550 K

Boehler, Annu. Rev. Earth Planet. Sci. 1996

Van den Berg, Rainey and Yuen, 2005

Mantel profiles based on a modified Hofmeister (1999) model

model of subducting slab interacting with the PV/PPv zone of D”

• Extended Boussinesq – ppv phase boundary• T,P dep.visc. dv_T=1000, dv_P=100• Layered piecewise uniform thermal conductivity • 2D domain 3000X3000km

- element resolution down to 5 km near CMB• Crustal (100) marker chain

Crustal markers

Spinel/Pv

Pv

PPv

K_cmb=5 W/(mK)

K_cmb=5 W/(mK)

K_cmb=5 W/(mK)

K_cmb=15 W/(mK)

K_cmb=15 W/(mK)

K_cmb=15 W/(mK)

Single tracer T-depth paths

Crustal markers in bottom PV(high k case)

Some conclusions

- Precursors to PKP from small scale scatterers near CMB can be

related to heterogeneity in the Pv/PPv region due to complex

phase distribution and/or crustal remnants

- PPv linked to D” allows temperature estimates near CMB from

seismic mapping of PPv > Pv in the thermal boundary layer

- Heatflow estimates based on simple linear T profiles are feasible

but require better constraints of k_cmb