University:
Widener University
Course:
PHYS-313 | Classical Mechanics
Academic year:

2023
Views:

177

Pages:

8
Author:

Lauryn Herring

Seismic Evidence for Olivine Phase Changes at the 410- and 660Kilometer Discontinuities Overview • Some disagreement about what phase change(s) cause the 660 – Irifune et al (1998), e.g.: suggest post-spinel occurs at 22 GPa, as opposed to 24 GPa – Seismic results ambiguous – Gt-Pv and Gt-Il-Pv suggested • This paper uses estimates of Vs from tomography studies and differential travel times from incoming Pds waves to calculate the seismic Clapeyron slope and thus constrain the phase changes that could be responsible for the 660. Seismically Inferred Clapeyron Slope • It is not meaningful to define thermodynamic Clapeyron slopes for the 410 and 660, since there are multiple phases • Bina & Helffrich (1994) define the seismic Clapeyron slope, which can be determined experimentally and is dependent on thermodynamic properties. • Seismic Clapeyron slope can be affected by isostructural phases with variable chemistry at the depth of the discontinuity, in particular because of the exchange of Mg and Fe between olivine and other minerals. Correlation between TZ thickness and TZ temperature. • If the 660 is caused by the post-spinel phase change, then the 410 and 660 should be deflected away (towards) each other for a cold (hot) anomaly. • The height of the TZ (HTZ) would then correlate with TZ temperature, as well as with Vs within the TZ (since Vs is temperature dependent). • If the 660 is caused by the gt-pv transition, which has a positive Clapeyron slope, the correlation between HTZ and temperature (and HTZ and Vs) would be weak or absent. Data selection • Pds waves used to calculate tdiff: t(diff) = t(P660s) – t(P410s) • Reject traces with low signal-to-noise ratios, stacks that didn’t show both P660s and P410s, and stations which did not have enough acceptable traces. • Used previously determined tomographic model values of Vs at their stations. • Reliable estimates for the upper (δVs410) and lower (δVs660) parts of the TZ obtained for eight stations. Results I • tdiff correlates with δVsTZ = δVs410 + δVs660 (r = .94) • For a TZ of constant thickness, tdiff would have negative correlation with velocity. • Convert tdiff to HTZ anomaly (δHTZ) using P and S velocities from iasp91, δVsTZ, and R = δlnVs/δlnVp = 1.7 +/- 0.7 • Convert δVsTZ to δTTZ (temperature anomaly) using dlnVs/dT = -1.35 x 10-4 K-1 and a 0.4 x 10-4 K-1 uncertainty. • HTZ correlates with δTTZ with r = 0.98; slope of best fitting line = -0.13 =/- 0.07 km/K; consistent with slope inferred from mineralogical Clapeyron slopes of alphabeta and post-spinel phase changes. Results II: Clapeyron slope • Use δVs estimates to compute seismic Clapeyron slopes (γ410 and γ660) δHTZ = (dd/dP)660 x γ660 x (dT/dlnVs)δlnVs660 – (dd/dP)410 x γ410 x (dT/dlnVs)δlnVs410 (dd/dP)660(410) describes the depth-pressure relationship at the 660(410) • Find γ410 and γ660 by minimizing chi-square function δHTZi – δHTZ, δHTZi the data at the ith station and δHTZ calculated Conclusions • The seismic Clapeyron slopes found for the 410 and 660 are consistent with the mineralogical Clapeyron slopes of the olivine transformations • Inconsistencies with other studies, where weak correlations between tdiff and VsTZ have been found, is due to differences in spatial resolution between tdiff and VpTZ and VsTZ from global tomography.