Seismic properties profiles of the alpine slab predicted by petrophysics versus ambient noise tomography lithospheric model

The objective of the present study is to use potential lithologic analogues sampled in the European crust units exhumed in the Alps to predict the seismic properties of the buried continental crust panel. To this end, from the chemical compositions of representative rock samples, we calculate seismic velocities (Vp, Vs or Vp/Vs) at any P and T, under the assumption that the rocks have completely re-equilibrated during burial.

The sample catalog comprehend (1) the mafic intercalations, present in the Variscan basement series of the External Crystalline Massif; (2) the rocks involved in the Grand Paradis - Schistes Lustrés contact (metabasites and garnet bearing micaschists of the upper unit, mylonite and gneiss of the lower unit); (3) those along the Lanzo-Canavese contact (serpentinites, blue schist facies mylonites and biotite bearing gneiss); (4) lithologies of the Ivrea domain (peridotites, garnet bearing gabbros, textured mafic rocks, amphibolitic and mylonitic paragneiss), (5) those from the Gruf massif (biotite bearing orthogneiss, deformed leucogranites and charnockites from the Gruf complex and amphibolites and serpentinites from the Chiavenna unit); (6) lithologies from Alpine Corsica (pelitic gneisses of the granulite facies and more or less foliated metagabbros, from the San Petrone and Farinole unit).

In these diagrams, the main seismic contrasts appear to correspond to the early stages of jadeite crystallization (mainly in the Vp/Vs diagram), as well as to the boundaries of the garnet and clinopyroxene stability fields. Considering the selected rocks as relevant analogues, we then compare the evolution of seismic properties along the top of the Alpine dipping panel with profiles inferred from recent Vp and Vs tomography models (CIFALPS 1 and AlpARRAY), varying the effective thermal profile of the Alpine panel, its reaction degree and overall chemistry. Preliminary results suggest that the lower crust of the plunging panel has a seismic velocity too low to be eclogitized. Its velocity rates are closer to those of an underreacted quartzo-felspathic gneiss. At first sight, observed velocities are too low compared to values predicted for any lithology fully reacted during subduction. The best-fitting scenario turns out to be that of a lower crust thermally relaxed in the variscan without significant mineralogical footprint of subduction. If detected, the velocity rise due to eclogitization might offset of several tenth along the slab, implying a sensible impact of reaction kinetics.