A preliminary probabilistic multi-hazard assessment for Ceboruco volcano
Ceboruco (2280 m.a.s.l.), in the western Trans-Mexican Volcanic Belt, is
considered among the most hazardous volcanoes in Mexico. Some 55,000
people and important infrastructure (e.g. hydroelectric dams; highways;
railways) lie within the area covered by deposits of Holocene eruptions.
A diverse activity over the past 1000 years spans from effusive (e.g.
andesite lava flows, dacitic domes) to explosive (e.g. Strombolian,
Vulcanian and Plinian eruptions). With a poor monitoring network, a
first hazard map was published in 2019. Here we present the first
probabilistic hazard maps for Ceboruco, which constitute a progression
towards a more quantitative hazard assessment. We conduct a
probabilistic hazard assessment using the pyBET_VH (Bayesian Event Tree
for Volcanic Hazard) tool (i.e. software implementation of the
event-tree scheme), which allows the user to estimate and visualize the
probabilities and uncertainties associated with volcanic phenomena.
pyBET_VH merges information from eruptive history, expert elicitation,
and the output of other computer models to produce probabilistic hazard
maps (i.e. absolute and conditional probabilities and associated
uncertainties). We present the probability hazard maps for each eruptive
scenario (i.e. Scenario 1 – small magnitude effusive eruption; Scenario
2 – medium magnitude effusive and/or explosive eruption
(VEI<3), and Scenario 3 – large magnitude Plinian eruption)
and the associated uncertainties. Such maps can be used by civil
authorities and stakeholders for the purpose of crisis management as
well as for long-term development strategies by visualizing the
probabilities of areas around the volcano likely to be impacted by
volcanic phenomena. Using pyBET_VH has advantages and disadvantages:
the reliability of the output maps is directly related to the quality of
the input data, but the tool allows easy estimation and visualization of
the uncertainties; being an interactive tool, the user can continuously
update the probability maps as new information becomes available.