No abstract

We present a three‐dimensional density model of the lithosphere and upper mantle for the Middle East and surroundings based on seismic, gravity, and seismic tomography data and analyze the main factors responsible for the density variations. The gravity effect of the crust is calculated and removed from the observed field using the most recent crustal model. The residual gravity anomalies are jointly inverted with the residual topography to image the density distribution within the upper mantle. The inversion is constrained by an initial density model based on seismic tomography. The obtained density variations span in a large range (±60 kg/m3), revealing strong asymmetry in the density structure of the Arabian plate. The uppermost mantle layer in the Arabian Shield is relatively dense. However, below a depth of ~100 km we observe a strong low‐density anomaly. In contrast, the mantle density in the Arabian platform increases at the same depths. The most pronounced decrease of the mantle density occurs in the Gulf of Aden, Red Sea, and East African Rift. Underneath the northern Red Sea the low‐density anomaly is limited to the depth ~150 km, while in the southern part it is likely linked to a mantle plume. The densest mantle material is found under the South Caspian basin, which is likely associated with an eclogite body in the uppermost mantle. In the collision zones (the Zagros Belt and the Hellenic Arc), the high‐density lithosphere shows the location of the subducting plates.

We construct a new-generation 3D density model of the upper mantle of Asia and its surrounding areas based on a joint interpretation of several data sets. A recent model of the crust combining nearly all available seismic data is employed to calculate the impact of the crust on the gravity anomalies and observed topography and to estimate the residual mantle anomalies and residual topography. These fields are jointly inverted to calculate the density variations in the lithosphere and upper mantle down to 325 km.As an initial approximation, we estimate density variations using a seismic tomography model. Seismic velocity variations are converted into temperatures and then to density variations based on mineral physics constraints. In the Occam-type inversion, we fit both the residual mantle gravity anomalies and residual topography by finding deviations to the initial model. The obtained corrections improve the resolution of the initial model and reflect important features of the mantle structure that are not well resolved by the seismic tomography. The most significant negative corrections of the upper mantle density, found in the Siberian and East European cratons, can be associated with depleted mantle material. The most pronounced positive density anomalies are found beneath the Tarim and South Caspian basins, Barents Sea, and Bay of Bengal. We attribute these anomalies to eclogites in the uppermost mantle, which have substantially affected the evolution of the basins. Furthermore, the obtained results provide evidence for the presence of eclogites in the oceanic subducting mantle lithosphere.

In this study we address the question of whether regional gravity field modeling techniques of GRACE data can offer improved resolution over traditional global spherical harmonic solutions. Earlier studies into large, equatorial river basins such as the Amazon, Zambezi and others showed no obvious distinction between regional and global techniques, but this may have been limited by the fact that these equatorial regions are at the latitudes where GRACE errors are known to be largest (due to the sparse groundtrack coverage). This study will focus on regions of higher latitude, specifically Greenland and Antarctica, where the density of GRACE measurements is much higher. The regional modeling technique employed made use of spherical radial basis functions (SRBF), complete with an optimal filtering algorithm. Comparisons of these regional solutions were made to a range of other publicly available global spherical harmonic solutions, and validated using ICESat laser altimetry. The timeframe considered was a 3 year period spanning from