Details

Autor(en) / Beteiligte

• Seydoux, Leonard
• Shapiro, Nikolai
• de Rosny, Julien

Titel

Estimation of the entropy of seismic ambient noise: Application to passive imaging

Ist Teil von

• The Journal of the Acoustical Society of America, 2017-05, Vol.141 (5), p.3526-3527

Erscheinungsjahr

2017

Link zum Volltext

Quelle

AIP Journals Complete

Beschreibungen/Notizen

• Recovering Green’s functions from diffuse ambient noise correlation is an efficient technique for passive seismic imaging. However, the field diffuseness is not completely fulfilled in practice because the noise is generated in preferential areas and contaminated by some highly coherent signals due to earthquakes, for instance. Here we show that the noise entropy is a robust estimator of the diffuseness. In the first part, we directly use the entropy as a metric of seismic noise activity to detect tremors around Piton de la Fournaise Volcano or highly coherent noise sources at continental scale using data collected by USArray seismic array. Then in the second part, we take the benefit of the entropy to propose an original equalization process based on the analysis of the covariance matrix to mitigate the effect a poorly diffuse noise field. The efficiency of the method is validated with several numerical tests. We apply the method to the data collected by the USArray, when a strong earthquake occurred. The method shows a clear improvement compared with the classical equalization to attenuate the highly energetic and coherent waves incoming from the earthquake, and allows to perform reliable travel-time measurement.