Graduation Year
2022
Document Type
Dissertation
Degree
Ph.D.
Degree Name
Doctor of Philosophy (Ph.D.)
Degree Granting Department
Geology
Major Professor
Charles B. Connor, Ph.D.
Co-Major Professor
Jan M. Lindsay, Ph.D.
Committee Member
Sylvain Charbonnier, Ph.D.
Committee Member
Aurelie Germa, Ph.D.
Committee Member
Shanaka de Silva, Ph.D.
Keywords
eruption magnitude, eruption source parameters, numerical modeling, tephra sedimentation
Abstract
This dissertation addresses the estimation of the magnitudes and intensities of explosive volcanic eruptions using tephra fall deposits and numerical modeling. Tephra fall deposits retain information about eruption dynamics. Volcanologists sample tephra fall deposits and measure the thickness, extent and grain size distribution to estimate the erupted volume and plume height of explosive eruptions. These terms are known as eruption source parameters. These parameters are used to classify the intensity and magnitude of the eruption on a scale such as the Volcanic Explosivity Index (VEI). Accurate estimation of the intensity and magnitude of explosive eruptions is a crucial factor in estimating volcanic hazards and compiling future hazard scenarios.
Syn- and post-eruptive processes affect tephra fall deposits. These processes include burial, remobilization and erosion. Sampling the deposit impacts estimation of intensity and magnitude in two ways: sample distribution, often governed by access to the deposit, and uncertainty in the deposit thickness where it is sampled. Another source of uncertainty stems from how statistical or numerical models are used by volcanologists to estimate eruption source parameters, such as eruption mass, column height and grainsize distribution, from deposit data. One example is given by the simplified models that do not account for the umbrella clouds observed during large explosive eruptions. These aleatoric and epistemic uncertainties are always present and lead to biased estimates of eruption source parameters.
In the first part of this dissertation, a forward numerical model for tephra sedimentation from umbrella clouds is introduced. Using the 2450 BP eruption of Pululagua volcano (Ecuador) as a case study, this model shows that the umbrella cloud is the major controlling factor of tephra sedimentation associated with very large explosive eruptions. A direct outcome of this research is the update of the commonly used VEI scale to include the umbrella cloud radius as an eruption source parameter to classify large explosive eruptions. With this model, the erupted volume, plume height and umbrella cloud radius can be successfully estimated from tephra deposit thickness measurements with reasonable uncertainty.
In the second part of this work, the forward model introduced earlier is paired with two inversion and uncertainty quantification algorithms. The scope of this second part of the dissertation is to estimate eruption source parameters from deposit thickness data and to quantify the associated uncertainty. The justification for a systematic approach to uncertainty quantification stems from the fact that numerous combinations of eruption source parameters can characterize a tephra fall deposit equally well; therefore, a range of eruption source parameters with uncertainty quantification is preferred to a point estimate of a given eruption source parameter, such as single value reported for deposit volume. This work shows that uncertainty quantification is as important as producing point estimates for the purposes of assessing eruption hazards and classifying eruptions based on geologic data.
The third part of this dissertation addresses the challenges of estimating eruption source parameters when the bulk of the tephra deposit cannot be sampled. During the 2021 eruption of La Soufrière volcano (St. Vincent and the Grenadines) tephra sedimentation occurred largely over the Atlantic Ocean. Following this eruption, I went to St. Vincent and sampled the tephra deposit. Access to the tephra deposit on land was possible only along the coastlines, in the proximal deposit, within 4 km to 6 km from the volcano. The two lowermost units of this tephra sequence have been sampled (P1 and P2). Unlike subsequent units, these two are the only units that are continuously traceable in outcrops on both coasts of the island of St. Vincent. Using a tephra sedimentation model and inversion analysis, the erupted mass was estimated at ~1.1 x 1010 ± 0.2 x 1010 kg and ~ 2.8 x 1011 ± 0.6 x 1011 kg, for P1 and P2 respectively, with the uncertainty reported at one standard deviation. The average particle release height was estimated at 12 km a.s.l. for both units. This work shows that the erupted mass can be estimated with reasonable uncertainty following explosive eruptions on islands.
Scholar Commons Citation
Constantinescu, Robert, "Characterizing Explosive Volcanic Eruptions Using Tephra Fall Deposit Data" (2022). USF Tampa Graduate Theses and Dissertations.
https://digitalcommons.usf.edu/etd/10284
