The seismically active Andean and Central American margins: Can satellite gravity map lithospheric structures?

Abstract

The spatial resolution and quality of geopotential models (EGM2008, EIGEN-5C, ITG-GRACE03s, and GOCO-01s) have been assessed as applied to lithospheric structure of the Andean and Central American subduction zones. For the validation, we compared the geopotential models with existing terrestrial gravity data and density models as constrained by seismic and geological data. The quality and resolution of the downward continued geopotential models in the Andes and Central America decrease with increasing topography and depend on the availability of terrestrial gravity data. High resolution of downward continued gravity data has been obtained over the Southern Andes where elevations are lower than 3000 m and sufficient terrestrial gravity data are available. The resolution decreases with an increase in elevation over the north Chilean Andes and Central America. The low resolution in Central America is mainly attributed to limited surface gravity data coverage of the region. To determine the minimum spatial dimension of a causative body that could be resolved using gravity gradient data, a synthetic gravity gradient response of a spherical anomalous mass has been computed at GOCE orbit height (254.9 km). It is shown that the minimum diameter of such a structure with density contrast of 240 kg m−3 should be at least ∼45 km to generate signal detectable at orbit height. The batholithic structure in Northern Chile, which is assumed to be associated with plate coupling and asperity generation, is about 60–120 km wide and could be traceable in GOCE data. Short wavelength anomalous structures are more pronounced in the components of the gravity gradient tensor and invariants than in the gravity field. As the ultimate objective of this study is to understand the state of stress along plate interface, the geometry of the density model, as constrained by combined gravity models and seismic data, has been used to develop dynamic model of the Andean margin. The results show that the stress regime in the fore-arc (high and low) tends to follow the trend of the earthquake distributions.