Sistema de Información Científica Redalyc
Red de Revistas Científicas de América Latina y el Caribe, España y Portugal
Inglés Español
Observed seismic motion in the lakebed zone of the Valley of Mexico is complex because of the large epicentral distance, an irregular 3D crustal structure, large noise levels, strong site effects, and the possible presence of mixed modes and multipathing. These conditions tax the possibilities of the array-processing techniques. In this paper, we test three methods to determine the most appropriate to analyze the data from a new dense accelerograph array, recently installed in Colonia Roma of Mexico City: the conventional frequency-wavenumber (f-k) method, the maximum-likehood f-k method and the Multiple Signal Characterization technique (MUSIC). Tests with synthetic data show that the f-k methods are more robust in presence of mixed signals but that MUSIC has better resolution and is less sensitive to noise. Results from data analysis show the superiority of the conventional f-k method in case of data with small coherence and large spatial aliasing, while MUSIC is more useful in case of data with high coherence and small time delays. Our results show that the wavefield includes slow diffracted waves coming mainly from south to west at short periods and faster waves from the epicenter at long periods. These observations support the idea that the observed large duration of ground motion is explained by the interaction of different modes of surface waves, arriving at different times at Mexico City, with the resonance of the clay layer.

Palabras clave: >>>
Ver Resumen
Universidad Autónoma del Estado de México
Sistema de Información Científica Redalyc ®
Versión 3.0 | 2018