Three-dimensional kernel-based coda attenuation imaging of caldera structures controlling the 1982-84 Campi Flegrei unrest

Waheed Gbenga Akande* (Corresponding Author), Luca De Siena, Quan Gan

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

16 Citations (Scopus)


Coda-wave attenuation imaging has risen as a state-of-the-art technique to depict volcanic structures using their dispersion effects. The 1982–84 seismic and deformation unrest at Campi Flegrei caldera (Italy) is a unique example of non-eruptive volcanic activity in a structured caldera. Here, we propose the first application of 3D coda-attenuation kernels to image caldera structures at multiple frequencies during unrest. Using sensitivity kernels is necessary to assess the effective resolution of coda imaging in highly heterogeneous volcanoes. The technique relies on the solution of Paasschens' equations in the framework of radiative transfer theory. The results map coda attenuation in the 3D space without need of pre-existent velocity models. The resolution and stability of the inversion solutions were examined by changing the damping parameters and outputting the corresponding images, inverting for different node spacings and performing checkerboard tests. These tests show that the resolution of the multiple-scattering model is much lower than that provided by a standard isotropic-scattering and/or single scattering technique. The best resolution in our model is obtained between depths of 1 km and 3 km in the centre of the model, between Pozzuoli town and Solfatara crater. The results are discussed at a frequency of 3 Hz, due to both longer coda durations and broader kernel illumination: in this frequency range, coda-based 3D imaging had so far failed to provide stable results. The interpretation is performed based on the extensive geological and geophysical knowledge of the caldera. High-attenuation anomalies below Solfatara and Monte Nuovo mark areas either saturated with water or enriched in molten rocks, feeding the respective fumarole fields. The flattening and horizontal elongation of these anomalies below 2 km depth is a manifestation of the blocking and spreading around of the rising hot magmatic fluids below a previously-inferred high-velocity, low-attenuation and highly-deforming caprock. This caprock is not uniform, likely due to the remnants of erupted structure. A SW-to-NE-trending low-attenuation and high-velocity anomaly deeps down to 3 km under Pozzuoli. We infer that the high seismicity in the region is a consequence of the stress sustained by the caprock from a 4-km-deep deformation source in 1983–84. On top of this source, high coda attenuation corresponds to high Vp/Vs ratios and high direct-wave attenuation. The region's characteristics are likely due to the accumulation of magmatic fluids above a magmatic sill.

Original languageEnglish
Pages (from-to)273-283
Number of pages11
JournalJournal of Volcanology and Geothermal Research
Early online date14 Jun 2019
Publication statusPublished - 1 Sept 2019

Bibliographical note

Acknowledgements W. G. Akande acknowledges the financial support of his sponsor, the Petroleum Technology Development Fund (PTDF), for his research. L. De Siena received the Public Engagement for Early Career Researchers funding from the Scottish Alliance for Geosciences Environment(SAGES) to collect data and discuss the results of the present work.We thank Edoardo Del Pezzo and Angel De La Torre for providing the codes necessary to compute the three-dimensional kernels using Paasschens' equations. The support of the editor,Diana Roman, and constructive remarks of two anonymous reviewer helped us constrain our interpretation, highlight the novelty of our approach, and avoid resolution biases.


  • Campi Flegrei
  • Coda waves
  • Geothermal
  • Seismic attenuation
  • Seismic tomography
  • Unrest


Dive into the research topics of 'Three-dimensional kernel-based coda attenuation imaging of caldera structures controlling the 1982-84 Campi Flegrei unrest'. Together they form a unique fingerprint.

Cite this