Determination of crustal structure in central Australia by inversion of traveltime residuals

Teleseismic traveltime anomalies for events recorded on portable seismograph arrays across the northern and southern margins of the Amadeus Basin in central Australia are inverted to obtain a picture of the velocity structure at a scale on the order of 5-10 km. A modified SIRT algorithm is used to iteratively redistribute traveltime anomalies along incoming ray paths, subject to regularization constraints, to obtain a direct estimate of the lithospheric velocity field beneath the arrays. Model structure is assumed to be 2-D, based on the strong east-west strike of surface geology and gravity structure in the area. Spurious structures commonly generated in SIRT inversions are suppressed using a filter based on the density of rays. A weighting towards near-surface structure is also applied to test the robustness of the inference of deep structures. The most prominent feature of the solutions for all three lines is a sharp interface between slow and fast regions, dipping away from the basin and beneath the adjacent basement blocks. The fastest region, on the upper side of the interface, corresponds to a belt of high-grade metamorphic rocks where they crop out at the surface. The interface between slow and fast regions extends to at least 50 km depth in all cases, dipping at about 50-60 degrees on the northern lines and 60-80 degrees on the southern line. On the northern lines the interface can be correlated with the Redbank Thrust identified in deep seismic reflection data, but the velocity interface seems to dip more steeply. Secondary features on the profiles include a possible change in dip of the interface on the southern line and a more complex structure on one of the northern lines, suggesting that one or more steeper faults cut through the hanging wall of the thrust there. Predicted gravity profiles derived from the velocity sections using empirical velocity-density relations resemble observed gravity surprisingly well in their style and approximate magnitude, providing support for the general features of the models. The results are consistent with a thick-skinned thrusting model for the deformation of the central segment of the basin margins and indicate that the style of deformation in the Late Proterozoic-Cambrian compressional event at the southern edge of the Amadeus Basin was similar to that in the later Alice Springs Orogeny in the north.