Tomographic study of the East African Rift in Mozambique - Initial results
Abstract
Extensively investigated from the Red Sea to southern Tanzania, the East African Rift (EAR) structure is still unknown on its southern tip, Mozambique. The M7 Machaze earthquake of 2006, in central Mozambique, shed new light on the location of the rifting activity, motivating the current initiative. During 2011, project MOZART (funded by FCT, Lisbon, PI J. Fonseca) deployed a network of 30 broadband (120s) seismic stations from the SEIS-UK Pool in central Mozambique and NE South Africa, to investigate the structure of the southernmost section of the EAR through a number of techniques. We present preliminary results of data quality control based on full-waveform modeling, hypocentral locations, and regional structure based on the analysis of ambient noise. Data quality control based on full-waveform modeling relies on comparisons between MOZART waveform data and synthetic seismograms computed with a spectral element method (Komatitsch and Tromp, 2002). These comparisons allow an understanding of how well existing global 3D and 1D Earth models, which were built from independent data, explain MOZART data, and how suitable MOZART data are for waveform tomography. The 3D Earth model synthetics explain the surface wave phases well, but are more limited at matching amplitudes, showing that there is still scope for improvement of the Earth model. The 1D Earth PREM synthetics explain surface wave amplitudes broadly as well as the 3D Earth model, but lead to much poorer phase fits. We will also present preliminary results of the analysis of local seismicity. Local hypocenters are determined using the SEISAN software. The hypocenter inversion is a modified version of HYPOCENTER (Lienert et al., 1986, Lienert,1991, Lienert and Havskov, 1995). The current version uses different seismic phases for earthquake location. The hypocentral locations of small to moderate earthquakes help to identify the most seismically active regions. Finally, we present the first results of the analysis of seismic ambient noise, which will be used in the future as start models for tomography. We start by computing cross-correlations between all station pairs to produce empirical Green functions. We use approximately one year of ambient noise data, from April 2011 to July 2012. We will present maps of group velocity as a function of period by performing frequency-time analysis (FTAN). The ambient noise study will generate the first regional velocity models for this region.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2012
- Bibcode:
- 2012AGUFM.T53C2730D
- Keywords:
-
- 7270 SEISMOLOGY / Tomography;
- 8109 TECTONOPHYSICS / Continental tectonics: extensional