The Imprint of Water-Column Resonance on Secondary Microseismic Source Spectra

M Meschede1, É Stutzmann1, V Farra1, M Schimmel2, F Ardhuin3

1 Institut de Physique du Globe de Paris, Paris 75005, France
2 Institut of Earth Sciences Jaume Almera - CSIC, 08028 Barcelona, Spain
3 University of Brest, CNRS, IRD, Ifremer, Laboratoire d'Océanographie Physique et Spatiale, IUEM

The Secondary Microseismic Frequency Band

P-wave Generation from Interacting Ocean Waves

inspired by Gualtieri et al. (2014) and Ardhuin et al. (2013, 2015)

Secondary Microseism Pressure Models

[model by Ardhuin et al. (2011), distributed by IFREMER
for video search "secondary microseism" on youtube]

Secondary Microseism Pressure Models

Connect
ocean wave pressure $\Leftrightarrow\,$ seismic observations

[model by Ardhuin et al. (2011), distributed by IFREMER
for video search "secondary microseism" on youtube]

Building a Dataset

  • five arrays (North-America)
  • combined timespan of five years (2006, 2008, 2011, 2015, 2015)
  • daily synthetic and observed beams for 3 models
    (1809 beams as a function of slowness and frequency)
  • this presentation:
    spectral characteristics of the strongest-sources

From Pressure Model to Synthetic Beam

Comparison of backprojected beams

The Strongest-Source Distribution

Selecting Comparable Sources

  • On 26% (467/1809) days the strongest-sources match
  • non-matching sources are mostly microseisms as well

The Spectral Double Peak

  • Spectral double peak in observation and models with resonance
  • Amplitudes are predicted within factor of 0.5 - 6 (1 std)
  • peak frequencies are either low or high

Associating spectral shapes with geographical regions

  • Spectral Shape and Geographical Region are related
  • Low frequency peak is broader than the high frequency one
  • High frequency spectra are better modeled

Source Site Resonance

  • resonance when $2h/\lambda + 1/2 = n$
  • $h = 1/4 \lambda,~3/4 \lambda,~5/4 \lambda, ...$
  • $n=1$ (shallow ocean) and $n=2$ (deep ocean) most important for secondary microseisms

Source Site Resonance Map

  • n=1 in shallow ocean [cross]
  • no resonance from 0.10 - 0.24Hz [circle]
  • n=2 in deep ocean [triangle]

Conclusions

  • secondary microseism P-wave spectra are strongly affected by resonance effects
  • bathymetry favours resonance at 0.15 and 0.20Hz (first and second harmonic)
  • ocean wave model can reasonably well predict the spectra