Gravity model for the North Atlantic ocean mantle: results, uncertainties and links to regional geodynamics

Abstract:

We present the results of gravity modeling for the North Atlantic region based on interpretation of GOCE gravity satellite data. First, to separate the gravity signal caused by density anomalies within the crust and the upper mantle, we subtract the lower harmonics in the gravity field, which are presumably caused by deep density structure of the Earth (the core and the lower mantle). Next, the gravity effect of the upper mantle is calculated by subtracting the gravity effect of the crustal model. Our “basic model” is constrained by a recent regional seismic model EUNAseis for the crustal structure (Artemieva and Thybo, 2013); for bathymetry and topography we use a global ETOPO1 model by NOAA. We test sensitivity of the results to different input parameters, such as bathymetry, crustal structure, and gravity field. For bathymetry, we additionally use GEBCO data; for crustal correction – a global model CRUST 1.0 (Laske, 2013); for gravity – EGM2008 (Pavlis, 2012). Sensitivity analysis shows that uncertainty in the crustal structure produces the largest deviation from “the basic model”. Use of different bathymetry data has little effect on the final results, comparable to the interpolation error. The difference in mantle residual gravity models based on GOCE and EMG2008 gravity data is 5-10 mGal. The results based on two crustal models have a similar pattern, but differ significantly in amplitude (ca. 250 mGal) for the Greenland-Faroe Ridge.

The results demonstrate the presence of a strong gravity and density heterogeneity in the upper mantle in the North Atlantic region. A number of mantle residual gravity anomalies are robust features, independent of the choice of model parameters. This include (i) a sharp contrast at the continent-ocean transition, (ii) positive mantle gravity anomalies associated with continental fragments (microcontinents) in the North Atlantic ocean; (iii) negative mantle gravity anomalies which mark regions with anomalous oceanic mantle and the Mid-Atlantic Ridge. To understand better a complex geodynamics mosaic in the region, we compare our results with regional geochemical data (Korenaga and Klemen, 2000), and find that residual mantle gravity anomalies are well correlated with anomalies in epsilon-Nd and iron-depletion.