Statistical study of seismic heterogeneities at the base of the mantle from PKP differential traveltimes

Abstract

S U M M A R YStochastic tomography can provide crucial information on the small-scale structures at the base of the mantle that are still beyond the reach of deterministic imaging. For this purpose, we relate differential traveltime variances of core phases to the correlation function of velocity heterogeneities at the base of the mantle. A global data set of PKP traveltimes is then used to invert for statistical properties of velocity perturbations in the D layer. We find that the average thickness of D is 350 ± 50 km with 1.2 ± 0.3 per cent rms velocity perturbations. The horizontal correlation length is 3 • ± 1 • and the vertical correlation length 150 ± 75 km. The statistical analysis of the traveltimes of core phases reveals a much stronger energy at short wavelengths compared to global tomographic models, which is in good agreement with the results of most studies of the seismic wavefield scattered at the base of the mantle, based upon the analysis of PKP precursors.

GlobalPwave tomography of Earth's lowermost mantle from partition modeling

“…The 300 km thickness scenario yielded the lowest noise estimates for the three data sets, indicating a better fit to the data. This thickness is in agreement with the statistical analysis of Garcia et al [] who estimated the average thickness of the lowermost layer to be 350 ±50 km. If we model the layer as being too thin, valuable information from P raypaths with bottoming depths above the 300 km level may be lost.…”

“…The scale length of the heterogeneity depends on the data set(s) used, but invariably structure ranging from wavelengths of hundreds to thousands of kilometers is revealed. Lowermost mantle heterogeneity at even smaller scales is almost certainly present, however, [e.g., Cormier , ; Helffrich , ; Margerin and Nolet , ; Garcia et al , ], but its resolution is not justified by the inferred noise levels of our data sets. It is important to note that no single data set, or even pair of data sets, produces as high resolution a map as does the combination of all three data sets (Figure ).…”

“…Garcia et al [], who performed a statistical analysis of P wave heterogeneity in the lowermost mantle also obtained a higher estimate (1.2±0.3%). This can be explained by the fact that like in our study, Garcia et al [] did not use the damping procedures typically employed in other inversion approaches. There is evidence, however, that P wave velocity variations in the lowermost mantle can reach even greater extremes than is revealed in our study.…”

Deep Earth Structure: Lower Mantle and D″

“…However, as is true of the lithosphere, key dynamic and chemical structures exist in D 00 on scales below the limit of deterministic resolution using seismic waves with wavelengths longer than 10 km or so. This has been demonstrated by the analysis of shortperiod precursors to PKP phases (e.g., Bataille and Flatté, 1988;Bataille et al, 1990;Brana and Helffrich, 2004;Cao and Romanowicz, 2007;Cleary and Haddon, 1972;Cormier, 1999;Doornbos, 1976;Garcia et al, 2009;Haddon and Cleary, 1974;Hedlin and Shearer, 2000;Hedlin et al, 1997;Miller and Niu, 2008;Thomas et al, 2009) and P 0 P 0 (Wu et al, 2012). FCC 90.7° 10 May1994 BB Disp.…”

Finite-frequency structural sensitivities of short-period compressional body waves

“…Since they are mostly sensitive to lowermost mantle structure, PKP ab‐ PKP df and PKP bc‐ PKP df differential traveltimes have been used to constrain the heterogeneities in the D″ layer (e.g. Garcia et al 2009). Kernels for these differential traveltimes are simply obtained by computing the difference of the kernels corresponding to the two PKP branches involved (Calvet & Chevrot 2005).…”