Seismic Attenuation Analysis using Lg waves and Ambient Noise Recordings: Application to the Iberian Peninsula and Morocco

Abstract

[eng] In this thesis I have carried out a comprehensive study of the attenuation properties of the Earth´s crust of the Iberia-Morocco region (IMR). I have investigated the crustal attenuation by means of the quality factor Q, which is inversely proportional to attenuation, using both earthquakes and noise-derived measurements. In order to fulfill the thesis objectives a large dataset including earthquake waveforms and seismic noise records has been used. Three different traditional earthquake methods have been implemented to estimate Q in the IMR: the two-station (TS) method, the coda normalization (CN) method and the spectral amplitude decay (SAD) method. For the estimation of Q, these approaches measure the spectral amplitude of the Lg wave (direct and coda) of regional events. Among all the methods evaluated, the TS method allows imaging the spatial variation of the Lg wave attenuation in the Iberian Peninsula whereas the CN and the SAD methods only estimate average attenuation values as well as its frequency dependence. For the Iberian Peninsula, high Lg Q values are observed in the stable Iberian Massif in western Iberia, while lower values are mainly found in the Pyrenean Range and in eastern and southern Iberia. For Morocco, the CN and the SAD methods produce similar results, indicating that the Lg Q models are robust to differences in the methodologies. The frequency-dependent Q estimates represent an average attenuation across a broad region of different structural domains and correlate well with areas of moderate seismicity. Additionally, I have studied the Lg propagation efficiency across the IMR. Results reflect an inefficient or even blocked propagation across the Gulf of Cádiz and for most paths crossing the western Alboran basin. The continental crust of the Iberian Peninsula and Morocco shows efficient Lg propagation. I have also investigated the potential of using ambient noise measurements to retrieve information about the anelastic structure of the Earth´s crust. Since noise preprocessing techniques modify the amplitude of the recovered empirical Green function of the medium, additional tests have to be done in order to verify the reliability of the attenuation results obtained. In this regard, I have carefully examined the influence of the distribution of noise sources and receivers on Q estimates. Azimuthally and spatially averaged Q values derived from noise recordings were further compared with earthquake attenuation measurements. Results reveal that the average Q estimates are in concordance with previous long-period surface-wave measurements from earthquakes in the central part of the Iberian Peninsula. Accurate Q estimates are also found in Morocco. I would like to emphasize that this thesis presents new contributions and improvements to the knowledge of the attenuation structure of the IMR. The first regional map that images the lateral variation of the Lg Q has been estimated for the Iberian Peninsula improving the spatial resolution of earlier studies. The frequency dependence of Lg Q has been also calculated for the first time in Morocco. Furthermore, this work is the first attempt to recover attenuation information from ambient seismic noise measurements in the study area. This novel technique allows us to investigate the attenuation structure of the Earth without the occurrence of earthquakes. Exploiting ambient seismic wavefields for attenuation studies will be a powerful tool to extract information about the anelastic structure and the geodynamics in areas of very low seismicity in the near future. It should be also noticed that recovering crustal attenuation values is important for many reasons. Attenuation estimates can be used to better quantify the hazard associated with earthquake ground shaking. Attenuation is also a valuable property in exploration seismology. For example, the presence of fluids can significantly attenuate the amplitude of the seismic waves.