Investigation of the Oceanic Crust and Mantle in the Eastern Mid Atlantic Next to a Major Transform Fault (Gloria Fault) By Receiver Function Analysis

Thursday, 18 December 2014
Katrin Hannemann1, Frank Krueger1 and Torsten Dahm1,2, (1)University of Potsdam, Potsdam, Germany, (2)GFZ Potsdam, Helmholtzstrasse 7, Germany
Within the project Deep OCean Test ARray (DOCTAR), we want to test how much we can enhance the signal to noise ratio (SNR) of teleseismic and regional events recorded at the ocean bottom by using broad band array methods. Furthermore, we want to learn more about the structure of the oceanic crust and mantle 100 km North of the Gloria Fault (major transform fault at the plate boundary between Eurasian and African plate in the Atlantic ocean). For the latter, we employ receiver functions and apparent P-wave incidence angles.

We deployed 12 ocean bottom stations (OBS) as a mid aperture array (75 km) in the deep Eastern Mid Atlantic (4-6 km) in 2011. Each free fall station consists of a broad band seismometer and a hydrophone. After 10 month of recording, the stations were recovered. We use P phase and Rayleigh phase polarization to estimate the orientation of the stations.

Different data quality and site effects at the stations need a careful review of the processing parameters (filter, deconvolution length) used for the calculation of the receiver functions. We defined different criteria as relative spike position within the deconvolution time window, and energy ratios of several time windows of the deconvolved traces to assess an evaluation of the receiver function quality in dependence on the used processing parameters. Additionally, we had a look at the relationship between the apparent incidence angle and the S-velocity and find that it differs for the ocean bottom in comparison to the free surface. Surprisingly, the densities of the oceanic crust and the water column, as well as the P-velocity of the water column have also an influence on the apparent incidence angle. We measured incidence angles for several events and find that the angles show a dependence on the dominant frequency of the event.

By comparison with synthetic receiver functions, we find that water multiples have a small or no influence at all on the real data receiver functions. We identify the MOHO conversion which indicates a crustal thickness of about 6 to 9 km which could be expected for 70 Ma old oceanic crust. Furthermore, there exist several features which are related to mantle structures. Finally, we use the Zhu and Kanamori (2000) algorithm, the apparent incidence angle, and the comparison with synthetic receiver functions to come up with a preliminary 1D velocity model.