Figure - available from: Frontiers in Earth Science
This content is subject to copyright.
Schematic of the reasoning about the physical processes associated with the triggering of lahars by means of rainfall utilized to construct our BBN model: Multihaz (see Figure 2 and Table 1). Together with Table 2, this reasoning defines the structure and parameterization of Multihaz (see text for more details).
Source publication
Volcanic water-sediment flows, commonly known as lahars, can often pose a higher threat to population and infrastructure than primary volcanic hazardous processes such as tephra fallout and Pyroclastic Density Currents (PDCs). Lahars are volcaniclastic flows of water, volcanic debris and entrained sediments that can travel long distances from their...
Citations
... It is based on the statistical correlation between the inundated area and the mass flow volume inferred from past events. However, in recent years, a few examples of probabilistic hazard assessment for lahars based on more robust statistical treatments, like statistical surrogates or emulation approaches, have been proposed for different volcanoes worldwide, such as Mead and Magill (2017) on Ruapehu (New Zealand), Tierz et al. (2017) on Vesuvius (Italy), and Gattuso et al. (2021) on Vulcano (Italy). ...
... Costa et al., 2009;Selva et al., 2010Selva et al., , 2018Sandri et al., 2016;Massaro et al., 2023), while systematic quantitative hazard assessments from lahars (see, for example, Jenkins et al., 2022) have been lacking. An exception is provided by Tierz et al. (2017), who applied a Bayesian belief network to assess the effect of different factors (linked to rainfall intensity and volcanoclastic volume) on the probability of different initial volumes of lahars. However, that study did not explore the variability in the hazard assessment related to the initial flow conditions (mostly linked to the flow volume, detachment area, and volumetric solid fraction). ...
... a large extension and the former by the availability of a large thickness of deposits from PDCs. The catchments on Somma-Vesuvius and those in the northeastern Apennine section (numbers 7 and 8 in our case) were also identified in Tierz et al. (2017) as those able to generate larger-initialvolume lahars in the case of a medium-sized eruption. We also notice that in some Apennine catchments (numbers 7 to 11) some simulations do not have significant deposits from tephra fallout to be remobilized (null initial volumes, probably because we simulated deposits from eruptions under wind fields not directed towards those catchments). ...
In this study we present a novel general methodology for probabilistic volcanic hazard assessment (PVHA) for lahars. We apply the methodology to perform a proba-bilistic assessment in the Campanian Plain (southern Italy), focusing on syn-eruptive lahars from a reference size eruption from Somma-Vesuvius. We take advantage of new field data relative to volcaniclastic flow deposits in the target region (Di Vito et al., 2024b) and recent improvements in modelling lahars (de' Michieli Vitturi et al., 2024). The former allowed defining proper probability density functions for the parameters related to the flow initial conditions, and the latter allowed computationally faster model runs. In this way, we are able to explore the effects of uncertainty in the initial flow conditions on the invasion of lahars in the target area by sampling coherent sets of values for the input model parameters and performing a large number of simulations. We also account for the uncertainty in the position of lahar generation by running the analysis on 11 different catchments threatening the Campanian Plain. The post-processing of the simulation outputs led to the production of hazard curves for the maximum flow thickness reached on a grid of points covering the Campanian Plain. By cutting the hazard curves at selected threshold values, we produce a portfolio of hazard maps and probability maps for the maximum flow thickness. We also produce hazard surface and probability maps for the simultaneous exceeding of pairs of thresholds in flow thickness and dynamic pressure. The latter hazard products represent , on one hand, a novel product in PVHA for lahars and, on the other hand, a useful means of impact assessment by assigning a probability to the occurrence of lahars that simultaneously have a relevant flow thickness and large dynamic pressure.
... Volcanic multi-hazard impact and risk assessment is in the early developmental stages, with recent advances in multi-hazard assessment (Tierz et al. 2017;Hayes et al. 2020), single-phase multi-hazard impact assessment (Zuccaro et al. 2008;Jenkins et al. 2014a), and co-development of impact-based decision-support tools (Hicks et al. 2014;Wild et al. 2019Wild et al. , 2021 providing valuable insight into best-practice approaches and research development required. ...
... This study integrates the growing knowledge base surrounding the co-production of disaster knowledge (Hicks et al. 2014;Davies et al. 2015;Whybark 2015;Fearnley and Beaven 2018;Hayes et al. 2020;Mach et al. 2020), spatio-temporal dependencies in volcanic systems (Jenkins et al. 2007;Zuccaro and De Gregorio 2013;Elissondo et al. 2016;Tierz et al. 2017;Bebbington and Jenkins 2019;Ang et al. 2020), and infrastructure vulnerability to volcanic hazards (Wilson et al. , 2014Dominguez et al. 2021), to conduct a longitudinal volcanic multi-hazard impact assessment for distributed infrastructure sectors surrounding Taranaki Mounga. This approach for volcanic impact assessment allows the investigation of multi-phase, multi-hazard volcanic activity, and the relative contribution of these components to impact or risk. ...
Effective volcanic impact and risk assessment underpins effective volcanic disaster risk management. Yet contemporary volcanic risk assessments face a number of challenges, including delineating hazard and impact sequences, and identifying and quantifying systemic risks. A more holistic approach to impact assessment is required, which incorporates the complex, multi-hazard nature of volcanic eruptions and the dynamic nature of vulnerability before, during and after a volcanic event. Addressing this need requires a multidisciplinary, integrated approach, involving scientists and stakeholders to co-develop decision-support tools that are scientifically credible and operationally relevant to provide a foundation for robust, evidence-based risk reduction decisions. This study presents a dynamic, longitudinal impact assessment framework for multi-phase, multi-hazard volcanic events and applies the framework to interdependent critical infrastructure networks in the Taranaki region of Aotearoa New Zealand, where Taranaki Mounga volcano has a high likelihood of producing a multi-phase explosive eruption within the next 50 years. In the framework, multi-phase scenarios temporally alternate multi-hazard footprints with risk reduction opportunities. Thus, direct and cascading impacts and any risk management actions carry through to the next phase of activity. The framework forms a testbed for more targeted mitigation and response planning and allows the investigation of optimal intervention timing for mitigation strategies during an evolving eruption. Using ‘risk management’ scenarios, we find the timing of mitigation intervention to be crucial in reducing disaster losses associated with volcanic activity. This is particularly apparent in indirect, systemic losses that cascade from direct damage to infrastructure assets. This novel, dynamic impact assessment approach addresses the increasing end-user need for impact-based decision-support tools that inform robust response and resilience planning.
... Regardless of the nature of the model used, modelled footprints better reproduce the directionality of hazards but are always more demanding in terms of computing power and parametrisation (number of eruption source parameters and other input conditions). Global hazard modelling is now becoming viable thanks to the increasing available computing power and dedicated open-source software (e.g., Bertin et al. 2019;Biass et al. 2016;Mahmood et al. 2015;Palma et al. 2014;Tierz et al. 2017), potentially opening the door to global Probabilistic Volcanic Hazard Assessments (PVHA). However, such regional to global studies require a balance between model sophistication and the computing power and input data available. ...
Effective risk management requires accurate assessment of population exposure to volcanic hazards. Assessment of this exposure at the large-scale has often relied on circular footprints of various sizes around a volcano to simplify challenges associated with estimating the directionality and distribution of the intensity of volcanic hazards. However, to date, exposure values obtained from circular footprints have never been compared with modelled hazard footprints. Here, we compare hazard and population exposure estimates calculated from concentric radii of 10, 30 and 100 km with those calculated from the simulation of dome- and column-collapse pyroclastic density currents (PDCs), large clasts, and tephra fall across Volcanic Explosivity Index (VEI) 3, 4 and 5 scenarios for 40 volcanoes in Indonesia and the Philippines. We found that a 10 km radius—considered by previous studies to capture hazard footprints and populations exposed for VEI ≤ 3 eruptions—generally overestimates the extent for most simulated hazards, except for column collapse PDCs. A 30 km radius – considered representative of life-threatening VEI ≤ 4 hazards—overestimates the extent of PDCs and large clasts but underestimates the extent of tephra fall. A 100 km radius encapsulates most simulated life-threatening hazards, although there are exceptions for certain combinations of scenario, source parameters, and volcano. In general, we observed a positive correlation between radii- and model-derived population exposure estimates in southeast Asia for all hazards except dome collapse PDC, which is very dependent upon topography. This study shows, for the first time, how and why concentric radii under- or over-estimate hazard extent and population exposure, providing a benchmark for interpreting radii-derived hazard and exposure estimates.
Supplementary information
The online version contains supplementary material available at 10.1007/s00445-023-01686-5.
... Statistical and physical models are used to recognise geologic and monitored data patterns to inform an eruption forecast (Poland and Anderson, 2020). Event trees (e.g., Newhall and Hoblitt, 2002;Connor et al., 2001) are one of the main probabilistic tools used in volcanology today (Poland and Anderson, 2020), and have been applied in real eruption crises by the USGS VDAP (Volcano Disaster Assistance Program, (e.g., Newhall and Pallister, 2015), and in hindcasting (e.g., Sandri et al., 2009;Tierz et al., 2017;Tierz et al., 2020), and simulation exercises (e. g., Lindsay et al., 2010;Marzocchi et al., 2008). The structure of event trees emphasises the inherently probabilistic nature of volcanic systems, containing multiple possible outcomes (Newhall and Hoblitt, 2002). ...
Taupō volcano, located within the Taupō Volcanic Zone (TVZ) in the central North Island of Aotearoa-New Zealand, is one of the world’s most active silicic caldera systems. Silicic calderas such as Taupō are capable of a broad and complex range of volcanological activity, ranging from minor unrest episodes to large destructive supereruptions. A critical tool for volcanic risk management is eruption forecasting. The Bayesian Event Tree for Eruption Forecasting (BET_EF) is one probabilistic eruption forecasting tool that can be used to produce short-term eruption forecasts for any volcano worldwide. A BET_EF model is developed for Taupō volcano, informed by geologic and historic data. Monitoring parameters for the model were obtained through a structured expert elicitation workshop with 30 of Aotearoa-New Zealand’s volcanologists and volcano monitoring scientists. The eruption probabilities output by the BET_EF model for Taupō volcano’s 17 recorded unrest episodes (between 1877 and 2019) were examined. We found time-inhomogeneity in the probabilities stemming from both the changes over time in the monitoring network around Taupō volcano and increasing level of past data (number of non-eruptive unrest episodes). We examine the former issue through the lens of the latest episodes, and the latter by re-running the episodes assuming knowledge of all 16 other episodes (calibration to 2021 data). The time variable monitoring network around Taupō volcano and parameter weights had a substantial impact on the estimated probabilities of magmatic unrest and eruption. We also note the need for improved monitoring and data processing at Taupō volcano, the existence of which would prompt updates and therefore refinements in the BET_EF model.
... However, our application to the Neapolitan volcanoes highlights the need of homogeneous model definitions for an effective comparison among volcanoes and for producing coherent multivolcano long-term hazard and risk quantifications. The application of this concept, along with the inclusion of potential interaction between different hazardous phenomena (45,76), will represent an improtant step toward an effective inclusion of volcanic hazards in multihazard risk evaluations, toward the better understanding and the strengthening of disaster risk governance, as it is pursued through the Sendai Framework for Disaster Risk Reduction 2015-2030 of the United Nations Office for Disaster Risk Reduction (www.undrr.org/). ...
Volcanic activity typically switches between high-activity states with many eruptions and low-activity states with few or no eruptions. We present a simple two-regime physics-informed statistical model that allows interpreting temporal modulations in eruptive activity. The model enhances comprehension and comparison of different volcanic systems and enables homogeneous integration into multivolcano hazard assessments that account for potential changes in volcanic regimes. The model satisfactorily fits the eruptive history of the three active volcanoes in the Neapolitan area, Italy (Mt. Vesuvius, Campi Flegrei, and Ischia) which encompass a wide range of volcanic behaviors. We find that these volcanoes have appreciably different processes for triggering and ending high-activity periods connected to different dominant volcanic processes controlling their eruptive activity, with different characteristic times and activity rates (expressed as number of eruptions per time interval). Presently, all three volcanoes are judged to be in a low-activity state, with decreasing probability of eruptions for Mt. Vesuvius, Ischia, and Campi Flegrei, respectively.
... However, our application to the Neapolitan volcanoes highlights the need of homogeneous model definitions for an effective comparison among volcanoes and for producing coherent multivolcano long-term hazard and risk quantifications. The application of this concept, along with the inclusion of potential interaction between different hazardous phenomena (45,76), will represent an improtant step toward an effective inclusion of volcanic hazards in multihazard risk evaluations, toward the better understanding and the strengthening of disaster risk governance, as it is pursued through the Sendai Framework for Disaster Risk Reduction 2015-2030 of the United Nations Office for Disaster Risk Reduction (www.undrr.org/). ...
Volcanic activity typically switches between high-activity states with many eruptions and low-activity states with few or no eruptions. We present a simple two-regime physics-informed statistical model that allows interpreting temporal modulations in eruptive activity. The model enhances comprehension and comparison of different volcanic systems and enables homogeneous integration into multivolcano hazard assessments that account for potential changes in volcanic regimes. The model satisfactorily fits the eruptive history of the three active volcanoes in the Neapolitan area, Italy (Mt. Vesuvius, Campi Flegrei, and Ischia) which encompass a wide range of volcanic behaviors. We find that these volcanoes have appreciably different processes for triggering and ending high-activity periods connected to different dominant volcanic processes controlling their eruptive activity, with different characteristic times and activity rates (expressed as number of eruptions per time interval). Presently, all three volcanoes are judged to be in a low-activity state, with decreasing probability of eruptions for Mt. Vesuvius, Ischia, and Campi Flegrei, respectively.
... We simulate rainfall-driven debris flows using the LaharFlow dynamic hazard model [65,66]. Previously LaharFlow has been used to simulate two-phase, overland flows that feature strong morphodynamics, such as lahars, flash floods, and debris flows in arid-urban environments [65,67,68]. LaharFlow, as with many other models that simulate surface flows, utilises a shallow-layer formulation which assumes that the pressure distribution is hydrostatic and flow properties are vertically averaged. ...
Rapid urban expansion in many parts of the world is leading to increased exposure to natural hazards, exacerbated by climate change. The use of physics-based models of natural hazards in risk-informed planning and decision-making frameworks may provide an improved understanding of site-specific hazard scenarios, allowing various decision makers to more accurately consider the consequences of their decisions on risk in future development. We present results of physics-based simulations of flood, earthquake, and debris flow scenarios in a virtual urban testbed. The effect of climate change, in terms of increasing rainfall intensity, is also incorporated into some of the considered hazard scenarios. We use our results to highlight the importance of using physics-based models applied to high-resolution urban plans to provide dynamic hazard information at the building level for different development options. Furthermore, our results demonstrate that including building elevations into digital elevation models is crucial for predicting the routing of hazardous flows through future urban landscapes. We show that simulations of multiple, independent hazards can assist with the identification of developing urban regions that are vulnerable to potential multi-hazard events. This information can direct further modelling to provide decision-makers with insights into potential multi-hazard events. Finally, we reflect on how information derived from physics-based hazard models can be effectively used in risk-sensitive planning and decision-making.
... 3.5.0) 4 for concentrated PDCs (Kelfoun et al., 2009). 3) The LaharFlow 5 model for syn-eruptive or 'hot' lahars generated by snow cap melting (Woodhouse et al., 2016;Tierz et al., 2017;Espín Bedón et al., 2019). 4) The Decreasing Probability model within the Q-LavHA framework (v. ...
... LaharFlow is a deterministic model that can produce realtime simulations of the distribution of lahars over a DEM (Woodhouse et al., 2016). It includes an empirical model for erosion and depositation and a phenomenological model for the variation of the basal stress as a function of the flow composition (Tierz et al., 2017). The parameters of LaharFlow were calibrated by using data from the 1985 Nevado del Ruiz lahar (Pierson et al., 1990). ...
... The parameters of LaharFlow were calibrated by using data from the 1985 Nevado del Ruiz lahar (Pierson et al., 1990). Even though LaharFlow is a relatively recent model, it has already been tested to model lahars in the literature (Tierz et al., 2017;Espín Bedón et al., 2019). The first-order parameters in LaharFlow are the DEM resolution and the volume. ...
Recent advances in the modeling of volcanic phenomena have allowed scientists to better understand the stochastic behavior of volcanic systems. Eruptions can produce various types of volcanic phenomena of different sizes. The size of a given volcanic phenomenon dominates its spatial distribution and is commonly represented by volume/mass parameters in the models that reproduce their behavior. Multi-hazard assessments depend on first-order parameters to forecast hazards at a given geographic location. However, few multi-hazard assessments consider the size of the eruption (e.g., tephra fallout) to co-parameterize the size of the accompanying phenomena (e.g., mass flows) in a given eruptive scenario. Furthermore, few studies simulate multi-phenomenon eruptive scenarios with semi-continuous variations in their size, something that allows a better quantification of the aleatoric variability of the system. Here, we present a multi-hazard assessment of the San Pedro volcano, a high-threat volcano from northern Chile, that produced two large-size Plinian eruptions (VEI 5 and 6) in the last 16 ka, and ten Strombolian eruptions (VEI 2) between 1870 and 2021 CE, with the latest occurring on 2 December 1960 CE. We use intra-scenarios (i.e., subdivisions of eruptive scenarios) to explore the size variability of explosive volcanic phenomena. The size of intra-scenarios is extrapolated from the largest-size deposits of each type of phenomenon from the geologic record of the San Pedro volcano. We simulate explosive intra-scenarios for tephra fallout, concentrated PDCs, and lahars, and effusive scenarios for blocky lava flows. On the local scale, mass flows are likely (66–100%) to affect transport and energy infrastructure within a 14 km radius of the volcano. On the regional scale, large-size eruptions (VEI 5) in the rainy season are about as likely as not (33–66%) to accumulate 1 cm of tephra on energy, transport, and mining infrastructure over a 50 km radius, and these same eruptions are unlikely (10–33%) to accumulate 1 cm of tephra on the city of Calama. This work shows how multi-phenomenon intra-scenarios can be applied to better quantify the aleatoric variability of the type and size of volcanic phenomena in hazard assessments.
... Similar to BETs, Bayesian Belief Networks (BBN) are graphical structures representing different events related to volcanic activity. Unlike BETs, BBNs describe the complexity of this activity as variable nodes interlinked by branches representing the causality between them (Christophersen et al. 2018;Tierz et al. 2017;Hincks et al. 2014;Aspinall and Woo 2014;Aspinall et al. 2003). The estimation of probabilities at each node in the BET or BBN schemes is done by implementing a computer algorithm based on Bayes' rule, which allows update of the output as new information becomes available (Christophersen et al. 2018;Marzocchi et al. 2008;Aspinall et al. 2003). ...
... The estimation of probabilities at each node in the BET or BBN schemes is done by implementing a computer algorithm based on Bayes' rule, which allows update of the output as new information becomes available (Christophersen et al. 2018;Marzocchi et al. 2008;Aspinall et al. 2003). Both ETs and BBNs have been used to forecast or hindcast eruptions for several volcanoes: Aluto, Ethiopia ; White Island, New Zealand (Christophersen et al. 2018); La Soufrière, Guadeloupe (Hincks et al. 2014); Galeras, Colombia (Aspinall et al. 2003); Santorini, Greece (Aspinall and Woo 2014); Etna and Somma-Vesuvius, Italy (Tierz et al. 2017;Cannavò et al. 2017); St. Helens, U.S.A (Newhall 1982); Soufrière Hills, Montserrat (Aspinall and Cooke 1998), Pinatubo, Philippines (Punongbayan et al. 1996) and others. ...
A probabilistic volcanic hazard assessment (PVHA) for Ceboruco volcano (Mexico) is reported using PyBetVH, an e-tool based on the Bayesian Event Tree (BET) methodology. Like many volcanoes, Ceboruco is under-monitored. Despite several eruptions in the late Holocene and efforts by several university and government groups to create and sustain a monitoring network, this active volcano is monitored intermittently rather than continuously by dedicated groups. With no consistent monitoring data available, we look at the geology and the eruptive history to inform prior models used in the PVHA. We estimate the probability of a magmatic eruption within the next time window (1 year) of ~ 0.002. We show how the BET creates higher probabilities in the absence of monitoring data, which if available would better inform the prior distribution. That is, there is a cost in terms of higher probabilities and higher uncertainties for having not yet developed a sustained volcano monitoring network. Next, three scenarios are developed for magmatic eruptions: i) small magnitude (effusive/explosive), ii) medium magnitude (Vulcanian/sub-Plinian) and iii) large magnitude (Plinian). These scenarios are inferred from the Holocene history of the volcano, with their related hazardous phenomena: ballistics, tephra fallout, pyroclastic density currents, lahars and lava flows. We present absolute probability maps (unconditional in terms of eruption size and vent location) for a magmatic eruption at Ceboruco volcano. With PyBetVH we estimate and visualize the uncertainties associated with each probability map. Our intent is that probability maps and uncertainties will be useful to local authorities who need to understand the hazard when considering the development of long-term urban and land-use planning and short-term crisis management strategies, and to the scientific community in their efforts to sustain monitoring of this active volcano.
... Cannavò et al. (2017) introduced BNs to real-time monitoring on Mount Etna, Italy. Tierz et al. (2017) used the flexible framework that BNs offer to assess rain-triggered lahars on Mount Somma-Vesuvius, Italy. Christophersen et al. (2018) undertook a pilot study to forecast eruptions for Whakaari/White Island, Aotearoa New Zealand in collaboration with the volcano monitoring scientists and volcanologists from two universities. ...
Volcano observatory best practice recommends using probabilistic methods to forecast eruptions to account for the complex natural processes leading up to an eruption and communicating the inherent uncertainties in appropriate ways. Bayesian networks (BNs) are an artificial intelligence technology to model complex systems with uncertainties. BNs consist of a graphical presentation of the system that is being modelled and robust statistics to describe the joint probability distribution of all variables. They have been applied successfully in many domains including risk assessment to support decision-making and modelling multiple data streams for eruption forecasting and volcanic hazard and risk assessment. However, they are not routinely or widely employed in volcano observatories yet. BNs provide a flexible framework to incorporate conceptual understanding of a volcano, learn from data when available and incorporate expert elicitation in the absence of data. Here we describe a method to build a BN model to support decision-making. The method is built on the process flow of risk management by the International Organization for Standardization. We have applied the method to develop a BN model to forecast the probability of eruption for Mt Ruapehu, Aotearoa New Zealand in collaboration with the New Zealand volcano monitoring group (VMG). Since 2014, the VMG has regularly estimated the probability of volcanic eruptions at Mt Ruapehu that impact beyond the crater rim. The BN model structure was built with expert elicitation based on the conceptual understanding of Mt Ruapehu and with a focus on making use of the long eruption catalogue and the long-term monitoring data. The model parameterisation was partly done by data learning, complemented by expert elicitation. The retrospective BN model forecasts agree well with the VMG elicitations. The BN model is now implemented as a software tool to automatically calculate daily forecast updates.
