Aller au compte twitter

  Thèse de Martha Deen

Seismic noise modelling and noise source analysis
Encadrant (et co-encadrant) : 
Yann Capdeville

Seismic noise is the continuous oscillations of the earth. In the frequency band of 2 to 300 seconds seismic noise is caused by ocean wave activities. This noise thus holds information on both the earth and the ocean. Even though some mechanisms have been known for decades, an integrated understanding of the noise in this period band is still missing. The objective of this research is achieving a unified understanding of these noise sources.

Up to now, the most energetic signature - the secondary microseism with periods from 2 to 10 seconds - is best understood. Non-linear interaction of ocean waves from almost opposing direction and almost equal period are its source mechanism. A peak signal with half the ocean wave period is measured on the seismogram. The other mechanisms, primary microseism (10-20 seconds) and seismic hum (20-300 seconds) are still under debate.
Both source mechanism are considered to be based on the linear interaction of an ocean wave with a sinusoidal bathymetry of the same wavelength. As a consequence, the period of the seismic wave is the same as that of the ocean waves and signals of periods up to 300 seconds can be measured.

As a method of gaining understanding on the noise content we analyse spectra from Ocean Bottom Stations (OBS) from the RHUM-RUM project. In this project 57 stations were deployed around la Réunion close to Madagaskar for a period of 13 months. Using a numerical ocean wave model we compute spectra with varying seismic attenuation and coastal contribution to reproduce the spectra of the data.
As a method of investigating the influence of bathymetry on the seismic noise generation, we model seismic data. To do this, we propagate seismic waves through an earth including ocean layer in a numerical model by spectral elements, taking into account 3D earth structure. The seismic sources are calculated from the numerical ocean wave model.

The PhD will lead to the identification and characterisation of noise source events, which applies to improving noise tomography; and seismic as well as ocean monitoring.

Date de soutenance: 
Friday 01 December 2017 - 14:00