GNGTS 2017 - 36° Convegno Nazionale

582 GNGTS 2017 S essione 3.1 characterized by fast wave speeds in the western, northern and southern part and by slower wave speeds beneath the eastern Phanerozoic margin. Previously, this has been mapped as a stepped lithosphere structure (Fishwick et al. , 2005) with relatively thin lithosphere around coastal eastern Australia, thicker lithosphere west of 145�� �� ������ ��� ���� ������� ����������� °E to 140°��E, and even thicker lithosphere west of 125�� �� ������ ����� ��� ����������� �� ������ �� ��� ������� ������� ���������� ���� ° ����� ��� ����������� �� ������ �� ��� ������� ������� ���������� ���� E to 120°E. These are interpreted to relate to the thermal effects associated with Phanerozoic tectonic activity. Furthermore, it has been identified an unusual area of low seismic wavespeeds in the uppermost mantle beneath central Australia, also with high radial anisotropy (Fichtner et al. , 2010). This low velocity anomaly is underlain by a region of fast wavespeeds more typical of cratonic lithosphere. This layered velocity structure may potentially have a thermal origin, due to the redistribution of high heat producing elements within the crust, or perhaps reflects the presence of amphibole in the uppermost mantle. Method. To discern temperature and compositional variations of theAustralian upper mantle, we apply an iterative technique, which jointly interprets seismic tomography and gravity data (Kaban et al. , 2014; Tesauro et al. , 2014). This technique consists in removing the effect of the crust from the observed gravity field and topography. In the second step, the residual mantle gravity field and residual topography are inverted to obtain a 3-D density model of the upper mantle. The inversion technique accounts for the notion that these fields are controlled by the same factors but in a different way (e.g., depending on depth and horizontal dimension of the heterogeneity.) This enables us to locate the position of principal density anomalies in the upper mantle. Afterwards, the thermal contribution to the density structure is estimated by inverting the seismic tomography model AusREM (http://rses.anu.edu.au/seismology/AuSREM/index.php ), assuming a laterally and vertically uniform ‘‘fertile’’ mantle composition. After removing this effect from the total mantle anomalies, the residual ‘‘compositional’’ fields are obtained. Some features of the composition density distribution, which are invisible in the seismic tomography data, are detected for the first time in the upper mantle. In this way, we could improve the initial thermal and compositional models by applying an iterative approach to account for the effect of composition on the thermal model. Results. The final thermal model with respect to the initial one shows temperatures higher than 100-150 °C in the Archean and Proterozoic upper mantle. The temperature model allows the mapping of lithospheric thickness (defined at 1300°C) for lithospheres as thick as ca. 250 km. These results show a clear variation in lithospheric thickness from east to west with steps that match closely the age-related steps previously suggested by Fishwick et al. (2005). However, the Archean to Proterozoic South Australian Craton shows relatively warm and thin lithosphere that may be compositionally fertile. This indicates that the lithosphere may be Proterozoic in age. In the North and West Australian cratons, the lithosphere is thick and relatively cool, with compositions that are depleted. This suggests a likely Archean lithosphere, relatively undisturbed through the Proterozoic. Central Australia shows an interesting structure with thick, low temperature lithosphere dominant but with a more fertile composition, with above a thin layer of low-velocity mantle. Our results indicate a high temperature for this feature that is not incompatible with the gravity field. However, this high temperature anomaly is hard to reconcile with the tectonic history of the region. Instead, we suggest a low-density mineral phase, such as amphibole, that may reduce the density relative to our assumed composition. These results are compatible with previous studies of the Australian continent including previous gravity studies (Aitken et al. , 2015). In contrast with previous studies, this analysis clearly separates compositional domains withinAustralia, in particular the separation ofArchean lithosphere from Proterozoic. We observe larger iron depletion in the West Australian Craton than in the Proterozoic terranes (Fig. 1). At depths larger than 150 km, the depletion becomes negligible beneath the Proterozoic regions, while persists in the West Australian Craton also below the depth of the lithosphere. We interpret this feature as the result of the leakage of