The Campi Flegrei caldera, Italy: 3-D structural model from seismic reflection data, and lithology characterization N. Maercklin 1, M. Vassallo 1, G. Festa2, A. Zollo2, D. Dello Iacono 2 & J. Virieux3 1 AMRA S.c.a.r.l., Naples, Italy (RISSC Lab); 3 2 Università di Napoli Federico II, Naples, Italy LGIT, Université Joseph Fourier, Grenoble, France European Seismological Commission, 32nd General Assembly, Montpellier, France, September 7, 2010
Introduction The Campi Flegrei caldera, Italy: 3-D structural model from seismic reflection data, and lithology characterization Outline: Introduction to the study area and geological overview Marine active seismic experiment (SERAPIS 2001) and reflection data 1-D layered model from reflection data and lithological interpretation 3-D reflection tomography for three main reflectors, preliminary results Summary and outlook
Introduction Active volcanoes in Campanian Plain: Vesuvius, Campi Flegrei, and Ischia. Campi Flegrei: partly submerged caldera, located in a very densely populated area (Naples).
Campi Flegrei caldera Caldera formed by two areal collapses after large explosive eruptions 39 ka (CI) and 15 ka (NYT) ago. Intra-caldera eruptions, most recent in 1538 (Monte Nuovo, 130 m). Current activity: Uplift episodes 1970-72 and 1982-84, accompanied by seismicity, 3.5 m cumulative uplift (cross). Hydrothermal activity (e.g. Solfatara).
SERAPIS Experiment Marine seismic survey in the Bays of Naples and Pozzuoli (2001). 30 three-component ocean-bottom seismometers in the Bay of Pozzuoli. Previous tomography results: High P-velocity anomaly, buried caldera rim (Zollo et al., 2003, GRL). SW-NE regional normal fault possibly linked to a deep magma reservoir (Judenherc & Zollo, 2004, JGR). No evidence for magma in upper 6 km (e.g. Battaglia et al., 2008, GP)
SERAPIS Experiment Marine seismic survey in the Bays of Naples and Pozzuoli (2001). 30 three-component ocean-bottom seismometers in the Bay of Pozzuoli. Reflection analysis: CMP gathers (500x500 m cells). Identification of reflected arrivals by move-out analysis and stacking. Manual picking of PP and PS reflection travel times and amplitudes.
Seismic reflection data Seismogram example (deepest reflections highlighted) Picked travel times (PP in red)
1-D Model from reflection data Highlights: Thermo-metamorphic rock layer around 3 km depth with low Vp/Vs (water- and/or gas-bearing; see Vanorio et al., 2005, JGR) Low-velocity layer with high Vp/Vs below 7.5 km depth, interpreted as partial melt Layered model of the Campi Flegrei caldera from PP and PS travel times and PS-to-PP amplitude ratios (Zollo et al., 2008, GRL)
3-D Reflection tomography Data: 9049 PP travel time picks from three reflectors (PS not used here) Inversion method: Jive3D (Hobro et al., 2003, GJI) Model consists of layers with constant or variable velocity, separated by reflectors Forward computation uses ray perturbation and efficient two-point ray tracing Regularized least-squares inversion of refraction and reflection travel times for reflector depths and velocities; iterative linearized approach with increasing model roughness Parameter uncertainties at all grid nodes estimated using an a-posteriori covariance analysis Application in this study: Pre-determined, fixed P-velocity within each layer Three reflectors, spatially sampled at 200 x 200 m; layer-stripping procedure
Tomography: Initial 1-D model Data misfit for each reflector as a function of depth and P-velocity above the reflector Reflector 1: Z = 0.55 km, Vp = 1.65 km/s Reflector 2: Z = 3.20 km, Vp = 3.70 km/s Reflector 3: Z = 7.60 km, Vp = 5.80 km/s
Tomography: Reflector near 0.5 km Depth uncertainty (blue = smaller) Reflector depth (dark = deeper)
Tomography: Reflector near 3.0 km Depth uncertainty (blue = smaller) Reflector depth (dark = deeper)
Tomography: Reflector near 7.5 km Depth uncertainty (blue = smaller) Reflector depth (dark = deeper)
PS travel time residuals Reflector 2 near 3.0 km depth Reflector 3 near 7.5 km depth PS travel time residuals relative to a 1-D layered velocity model (Red indicates a deeper reflector depth)
Summary and outlook 3-D reflection tomography reveals the morphology of three main reflectors (preliminary), i.e. the base of marine sediments, the top of the fluid-/gas-bearing layer, and the top of an extended magma reservoir. The two upper reflectors are well resolved, and the reflector near 3 km depth features a morphological high coinciding with the buried caldera rim. Ongoing and future work: Improvement of reflector images and model resolution analyses. Joint inversion of PP and PS data; possibly also inclusion of first-arrival times. Rock physics modelling to characterize lithologies (fluids at 3 km depth and magma). Questions: Small magma patches above the 7.5 km reflector? Structural relation between the Campi Flegrei caldera and Vesuvius volcano?
Tomography: Data misfit and model roughness Reflector 1 near 0.5 km depth Reflector 3 near 7.5 km depth Reflector 2 near 3.0 km depth
First-arrival tomography (active & passive) Vp Crossplots: Vp/Vs Vp and Vs Vp Vp/Vs 3-D velocity model of Battaglia et al., 2008, GP)
Melt layer below 7.5 km depth Geothermal gradient (from Carlino and Somma, 2010, BV) Hashin-Shtrikman bounds: Range of Vp and Vs for isotropic mixture of melt and solid (Campi Flegrei: Zollo et al., 2008, GRL) Vp and Vs for a model considering a critical melt fraction (Yellowstone: Chu et al., 2010, GRL)