Imaging the deep petrophysical architecture of Toba caldera from seismic data

Autor: Magrini, F., De Siena, L., Kaus, B., Riel, N., Diaferia, G., Forni, F.
Rok vydání: 2023
Zdroj: XXVIII General Assembly of the International Union of Geodesy and Geophysics (IUGG)
DOI: 10.57757/iugg23-0134
Popis: The crustal feeding systems of Toba caldera (Indonesia) remain largely unknown due to (1) the lack of seismic images resolving magmatic sills and fluid reservoirs and (2) our inability to translate seismic parameters into meaningful petrophysical quantities. Using data from available seismic arrays, we obtained thousands of dispersion curves sensitive to both the shallow crust and the upper mantle. The usual seismological approach is to invert them for phase-velocity maps at different periods and transform them into a shear-wave model. This model shows low-velocity sill-like structures at different depths under the northern and central Toba caldera; however, the uncertainty in their interpretation cannot be quantified just by looking at their shape, depths and correlation with surface features.Therefore, we used a Gibbs energy-minimization solver thatcomputes seismic velocities from phase petrology for the compositions relevant to Toba to forward phase velocities directly. A Bayesian Monte Carlo Markov chain inversion then inverts for temperature, composition, background host rock damageand melt content. As a result, every velocity value from our calculations is petrologically feasible. The results identify the portions of the crust where mafic sills are located, quantitatively defining their melt content, chemical composition, rock damage, temperature and associated uncertainty. Due to the low sensitivity of shear waves to temperature, there are still fundamental challenges in defining the thermal structure of the volcano; yet, seismo-petrological inversions are feasible today within one of the most complex crustal systems, a result few expected to achieve a decade ago. 
The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)
Databáze: OpenAIRE