Vol.:(0123456789) Natural Hazards (2024) 120:9327–9360 https://doi.org/10.1007/s11069-023-06386-z 1 3 ORIGINAL PAPER Approaching the challenge of multi‑phase, multi‑hazard volcanic impact assessment through the lens of systemic risk: application to Taranaki Mounga Alana M. Weir1,2  · Thomas M. Wilson1  · Mark S. Bebbington3,4  · Sarah Beaven1  · Teresa Gordon5,6 · Craig Campbell‑Smart6 · Stuart Mead3  · James H. Williams1  · Roger Fairclough7 Received: 9 November 2022 / Accepted: 19 December 2023 / Published online: 16 January 2024 © The Author(s) 2024 Abstract 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 sys- temic risks. A more holistic approach to impact assessment is required, which incorpo- rates 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 pro- vide a foundation for robust, evidence-based risk reduction decisions. This study presents a dynamic, longitudinal impact assessment framework for multi-phase, multi-hazard vol- canic 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 interven- tion 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. Keywords Volcanic risk assessment · Disaster impact assessment · Co-creation · Dynamic risk · Disaster risk reduction · Aotearoa New Zealand Extended author information available on the last page of the article http://crossmark.crossref.org/dialog/?doi=10.1007/s11069-023-06386-z&domain=pdf http://orcid.org/0000-0001-7449-016X http://orcid.org/0000-0002-8816-0708 http://orcid.org/0000-0003-3504-7418 https://orcid.org/0000-0001-7325-6013 https://orcid.org/0000-0003-3135-439X https://orcid.org/0000-0002-7564-0032 9328 Natural Hazards (2024) 120:9327–9360 1 3 1 Introduction Effective volcanic impact and risk assessment underpins effective volcanic disaster risk management (DRM) (UNISDR 2015; Ministry of Civil Defence & Emergency Manage- ment (MCDEM) 2019; UNDRR 2019). The drivers of disaster risk can be influenced by practitioners, policy makers, society and individuals, but accurate assessment and reevalu- ation of risk is required to enable effective disaster risk management decisions (Spence 2004; Coppola 2011; World Bank 2014). Scientific advances in volcanic hazard assessment have increased the capacity for scientists and science agencies to provide more reliable advice to inform decision-making about future volcanic eruptions (Marzocchi et al. 2008, 2010; Lindsay et al. 2010; Sandri et al. 2014; Hayes et al. 2020; Wild et al. 2020). How- ever, when developing appropriate risk management actions for those exposed to volcanic risk, evidence strongly suggests that impact and risk information is substantially more use- ful, useable and used than hazard information alone (Hicks et al. 2014; Hudson-Doyle et al. 2014b; Armijos et al. 2017; Thompson et al. 2017b; Hicks and Barclay 2018). Impact assessment requires detailed information regarding hazard potential, exposure and the vulnerability characteristics of the exposed assets and communities (UNISDR 2015). Volcanic eruptions are recognised to be multi-phase events that have the potential to produce multiple hazards throughout an eruptive episode (Jenkins et al. 2007; Marzoc- chi and Bebbington 2012; Bebbington and Jenkins 2019, 2022). These hazards can have differing physical and chemical characteristics and are subject to different constraints in their spatial and temporal extent. Society is also dynamic, with various drivers influencing exposure and vulnerability in space and time, prior, during and following volcanic activity (Hicks and Few 2015; Few et  al. 2017). Yet to date, impact and risk assessment frame- works are typically limited to producing static, single-phase estimations of impact and risk, despite global recognition of the systemic and dynamic nature of disaster risk (Jenkins et al. 2014b; World Bank 2014; Loughlin et al. 2015; UNDRR 2019). Cascading hazards and vulnerabilities exacerbate the impacts observed from disasters and exhibit strong spatio-temporal dependencies (Zuccaro and De Gregorio 2013; Pescaroli and Alexander 2018; Zuccaro et al. 2018; Dunant et al. 2021a), particularly in a volcanic risk context (Biass et al. 2014, 2016; Hicks and Few 2015; Phillips et al. 2019). Indirect losses during and following disasters are well documented (Galderisi et al. 2011; Cherry et  al. 2015; Armijos et  al. 2017; Few et  al. 2017; National 2017; Thacker et  al. 2017a; Phillips et al. 2019; Schweizer and Renn 2019; UNDRR 2019) and are often witnessed far beyond the spatial and temporal extent of the hazard. Critical infrastructure networks are recognised to be essential for the well-being of communities and socio-economic activi- ties and can be highly interdependent (Rinaldi et al. 2001). Despite global recognition of the contribution of indirect infrastructure loss of service to the total observed impact dur- ing disaster events (Zorn and Shamseldin 2015; Zorn 2017b; UNDRR 2019), there is little to no incorporation of this dimension in volcanic impact and risk assessment frameworks to date. Yet the inclusion of indirect impacts in volcanic impact assessment frameworks would likely enhance their quality and thus better inform resilience-building initiatives and inform critical infrastructure network management during volcanic crises. Increasingly, these complex challenges are being addressed by multi-disciplinary teams of scientists and practitioners, that partner in the production of disaster risk reduction (DRR) knowledge and disaster risk management (DRM) strategies (Davies et  al. 2015; Barton et  al. 2020; Hayes et  al. 2020; Weir et  al. 2022). Effective disaster risk manage- ment (DRM) can be strengthened by partnerships between practitioners and scientists to 9329Natural Hazards (2024) 120:9327–9360 1 3 develop an understanding of what is legitimate, relevant and credible for the local risk con- text (Cash et al. 2003; Davies et al. 2015; Fearnley and Beaven 2018; Barton et al. 2020). It follows that collaboration with stakeholders with the relevant expertise and knowledge can greatly increase the effectiveness of impact assessments (Whybark 2015; Thompson et al. 2017b; Orchiston et al. 2018; Barton et al. 2020; Mach et al. 2020). This study develops and presents a framework for impact assessment under multi-phase, multi-hazard volcanic events and tests the effectiveness of the framework for mitigation planning and risk management interventions. The framework incorporates direct and indi- rect impacts to interdependent infrastructure networks and was developed in partnership with local end-user agencies. The framework addresses the growing demand for impact- based decision-support tools to facilitate more targeted, considered response and mitigation planning, ultimately contributing to reducing losses associated with multi-hazard volcanic eruptions. We apply the framework to interdependent critical infrastructure sectors in the Taranaki region of Aotearoa New Zealand, where Taranaki Mounga1 volcano is estimated to have a 33–42% likelihood of erupting within the next 50 years (Damaschke et al. 2018). 1.1 Multi‑hazard impact and risk assessment: approaches Multi-hazard impact and risk assessment has been identified as an important requirement for disaster risk assessment (AghaKouchak et al. 2018; Ordaz et al. 2019; UNDRR 2019), but is a research area in need of further development. Previous studies have generally been focused on hazard, particularly the quantification of cascading and interacting hazard rela- tionships and their contribution to potential exposure and impacts (Zuccaro et  al. 2008; Deligne et al. 2017a, b; Bathrellos et al. 2017; Yousefi et al. 2020; Dunant et al. 2021a, b; Skilodimou and Bathrellos 2021). There has been comparatively less attention given to integrating exposure, vulnerability and resilience elements of impact and risk within these frameworks (Ordaz et al. 2019; Phillips et al. 2019; Wang et al. 2020; Dominguez et al. 2021; Kc et al. 2021). Various studies have attempted and noted the difficulty in integrating the physical and spatio-temporal characteristics of different hazards to produce an appro- priate common metric which can be used within multi-hazard impact and risk assessment frameworks (Kappes et al. 2012; Tilloy et al. 2019; Wang et al. 2020). Attempts to address this have led to contributions in quantifying multi-hazard potential, generally either quali- tatively or semi-quantitatively as hazard classifications or indices, where multi-hazards are typically related by intensity, duration and/or recurrence interval to deduce the ‘hazard level’ at a given location (e.g. “high ashfall and high lahar hazard”) (Tilloy et  al. 2019; Wang et al. 2020). There has also been the strong appreciation that such frameworks and their resultant outputs can be highly technical, so a key consideration is also to ensure the resultant impact and risk metrics are suitable for end-user purposes (Wachinger et al. 2013; Hicks et al. 2014; Hudson-Doyle et al. 2014b, a; Davies and Davies 2018). Few studies have considered the relative contribution from multi-hazards to observed or potential impact (e.g. damage and disruption) and how this manifests throughout a multi- hazard cascading initiation sequence. Notable exceptions include Zuccaro et al. (2008) and Zuccaro et al. (2018). Additionally, due to the uncertainty and complexity of probabilistic approaches, they have proven to be difficult for many DRM users to adopt, with the notable 1 The term Mounga is the te reo Māori language term for mountain, mount or peak and is a local variation of the more commonly used Maunga. 9330 Natural Hazards (2024) 120:9327–9360 1 3 exception of insurance, building code regulation and construction, and some land use plan- ning applications (Haynes et al. 2008; Global Facility for Disaster Reduction and Recovery 2014; Hudson-Doyle et  al. 2014b; World Bank 2014; Davies and Davies 2018). Hybrid impact-led approaches that enhance the value of both deterministic (scenario) and prob- abilistic outputs are finding increasing favour (Thompson et  al. 2017b; Ang et  al. 2020; Hayes et al. 2020), especially in settings where scientists and users have attempted to col- laboratively partner in DRR activities that distil the complexity of potential impacts in a credible, yet useable way (Haynes et  al. 2008; Barclay et  al. 2008; Hudson-Doyle et  al. 2015; Ang et al. 2020; Hayes et al. 2020). 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. 2019, 2021) providing valuable insight into best-practice approaches and research develop- ment required. 2 Longitudinal multi‑hazard volcanic impact assessment: design and requirements 2.1 Partnering with DRR knowledge end‑users: needs and requirements The collaborative development and management of impact assessment frameworks with information end-users can facilitate knowledge sharing and increase the relevance and uptake of scientific information (Davies 2015, Barton 2020). Well-managed collaborations between science, practice and policy have been found to result in enduring, partnerships that contribute to scientifically robust knowledge that meets the requirements of the deci- sion-makers (Wyborn et  al. 2017; Fearnley and Beaven 2018; Barton et  al. 2020; Mach et al. 2020). This study was conducted alongside Weir et al. (2022) and shares the collabo- rative, co-production approach with stakeholders and research partners (refer to Weir et al. (2022) for more detail). As in Weir et  al. (2022), ‘co-production of [knowledge in] this study [was achieved via three main mechanisms]: embedding the lead science author of this study (Weir) in the most appropriate policy and practice agency (the local emergency management authority) for several weeks (1), regular semi-structured meetings between actors either side of the SPPI (2), and iterative ongoing validation of the outcomes with the emergency management agency’s wider stakeholder networks (3). These mechanisms were used to establish the key objectives of this study (1), clearly identify the scope (1, 2), and ensure that the format and content of the study was relevant and applicable (1, 2, 3)’ (p. 4, Weir et al. 2022). During and after the embedding period, the lead author coordinated and facilitated monthly semi-structured meetings between volcanic hazard and risk scientists and emer- gency management practitioners involved in this study to identify and develop solutions for a series of key issues surrounding volcanic risk assessment and management for future Taranaki Mounga volcanism. This decision-making process had two main stages of meth- odological development: a knowledge stock take (Stage 1) and evaluation of the suitability of available volcanic impact assessment frameworks (Stage 2) (Table 1). Both the scien- tists and the practitioners were motivated to identify problems and develop solutions; how- ever, some aspects were more strongly motivated by scientists (e.g. the development of 9331Natural Hazards (2024) 120:9327–9360 1 3 Ta bl e 1 T he k ey sc op in g an d pr ob le m id en tifi ca tio n st ag es in th e co -p ro du ct io n pr oc es s  Th e ta bl e sh ow s th e is su es id en tifi ed b y Ta ra na ki C iv il D ef en ce a nd E m er ge nc y M an ag em en t ( Ta ra na ki C D EM , t he lo ca l e m er ge nc y m an ag em en t a ut ho rit y) a nd th e au th or s co -c re at iv el y, a nd th e re su lta nt a ct io ns ta ke n, d oc um en te d in th is st ud y Re le va nt lo ca l c on te xt Sc op in g st ag e Pr ob le m So lu tio n Ta ra na ki M ou ng a ha s a h ig h lik el ih oo d of pr od uc in g an e xp lo si ve e ru pt io n (D am as ch ke et  a l. 20 18 ) St ag e 1: K no w le dg e sto ck -ta ke a nd k no w l- ed ge g ap id en tifi ca - tio n La ck o f u nd er st an di ng o f t he tr an sl at io n of po te nt ia l h az ar ds (W ei r e t a l. 20 22 ) i nt o po te nt ia l i m pa ct C on du ct a v ol ca ni c im pa ct a ss es sm en t f or c rit i- ca l i nf ra str uc tu re se ct or s i n th e Ta ra na ki re gi on (th is st ud y) Th e Ta ra na ki re gi on h ou se s n at io na lly a nd re gi on al ly si gn ifi ca nt c rit ic al in fr as tru ct ur e La ck o f i m pa ct a ss es sm en t s tu di es fo r c rit ic al in fr as tru ct ur e se ct or s i n th e Ta ra na ki re gi on (W ei r e t a l. 20 22 ) Ta ra na ki M ou ng a ha s a h ig h po te nt ia l o f p ro - du ci ng a lo ng -d ur at io n, m ul ti- ph as e, m ul ti- ha za rd e ru pt io n (W ei r e t a l. 20 22 ) St ag e 2: E va lu at io n of av ai la bl e vo lc an ic im pa ct a ss es sm en t fr am ew or ks La ck o f a va ila bl e vo lc an ic im pa ct a ss es sm en t fr am ew or ks fo r m ul ti- ph as e, c om pl ex e ru p- tio ns D ev el op a m ul ti- ph as e vo lc an ic im pa ct a ss es s- m en t f ra m ew or k (th is st ud y) C rit ic al in fr as tru ct ur e se ct or s i n th e Ta ra na ki re gi on a re h ig hl y in te rd ep en de nt La ck o f v ol ca ni c im pa ct a ss es sm en t f ra m e- w or ks th at in co rp or at e in di re ct in fr as tru ct ur e im pa ct D ev el op a v ol ca ni c im pa ct a ss es sm en t f ra m e- w or k fo r d ire ct a nd in di re ct im pa ct s ( th is stu dy ) Lo ca l a ut ho rit ie s a re ta sk ed w ith p la nn in g fo r an d re sp on di ng to v ol ca ni c ac tiv ity (C D EM A ct 2 00 2) La ck o f v ol ca ni c im pa ct a ss es sm en t f ra m e- w or ks th at te st de ci si on -m ak in g sc en ar io s D ev el op a v ol ca ni c im pa ct a ss es sm en t f ra m e- w or k th at a llo w s t es tin g of d ec is io n- m ak in g sc en ar io s a nd o pt im al in te rv en tio n tim e (th is stu dy ) 9332 Natural Hazards (2024) 120:9327–9360 1 3 novel, sophisticated volcanic hazard and risk assessment tools), and some by practitioners (e.g. tools to support decision-making for volcanic risk management and volcanic crisis response). Balancing the scientific credibility and operational relevance of the resultant volcanic impact assessment framework enhanced the legitimacy of the research outcomes, as it well reflects the knowledge requirements of both scientists and practitioners. 2.2 Longitudinal multi‑hazard volcanic impact assessment for interdependent infrastructure: framework design This study presents a multi-phase, multi-hazard impact assessment framework that car- ries previously incurred component damage through to the next volcanic activity phase, assesses downstream, systemic outages associated with the physical damage of infrastruc- ture assets, and allows the evaluation of mitigation or decision-making efficacy during an evolving multi-phase event sequence (Fig. 1). By alternating volcanic activity phases with decision-making opportunities, the framework allows testing of the timing of risk manage- ment actions and considers how different elements of systemic risk cascade to provoke losses in the event of a disaster. The study utilises pre-existing hazard, exposure and vul- nerability tools and information as building blocks (or modules) in the framework, and applies the resultant framework to conduct a longitudinal impact assessment for distributed Fig. 1 The dynamic impact assessment framework comprised of five modules: (1) volcanic multi-hazard, multi-phase footprints (E1, E3, E5…En), (2) risk management interventions (I2, I4 … In), (3) exposed assets and their characteristics, (4) direct impact to assets, and (5) indirect impact to assets 9333Natural Hazards (2024) 120:9327–9360 1 3 interdependent infrastructure networks surrounding Taranaki Mounga volcano, in the Taranaki region of Aotearoa New Zealand. 2.2.1 Framework modules The impact assessment framework is comprised of five key modules, detailed in the fol- lowing sections. The modules are: (1) volcanic activity events (VAEs) (2) risk management interventions (RMIs), (3) exposed interdependent infrastructure assets, (4) volcanic multi- hazard direct impact and (5) indirect impact. The framework and its five constituent mod- ules are illustrated in Fig. 1. The proposed dynamic volcanic impact assessment framework requires volcanic hazard footprints for a multi-phase eruption. The VAEs (E1, E3, E5…En) are interspersed with potential opportunities for mitigation (RMIs; I2, I4, I6…In) (Fig. 1). VAEs can consist of multiple volcanic hazards, such as ashfall concurrent with pyroclastic density currents (PDC) or lava flows. Spatial determination of hazard footprints is required in order to deduce the assets directly exposed to the volcanic hazards. In this framework, accumulated direct impact to an infrastructure asset from volcanic hazards is carried through to the next event in the framework. For each VAE, relevant vul- nerability models are applied to calculate the direct impact to exposed assets. As differ- ent volcanic hazards possess differing physical behaviours and properties, the best hazard intensity metric (HIM) to evaluate impact for one hazard may not be appropriate for use when assessing impact from a different hazard. An Impact State (IS) metric accounts for differing physical parameters and is used in this multi-hazard framework. Volcanic impacts can be expressed using different IS metrics, such as physical damage, service disruption or economic loss (Marzocchi and Woo 2009; Wilson 2015; Mcdonald et al. 2017). Following the determination of direct impact, indirect impact through an interdependent network graph of critical infrastructure is then calculated. For the next VAE, the expo- sure inventory is comprised of the direct and indirect impact dataset from the previous VAE, thus carrying previously incurred damage and disruption throughout to the subse- quent phases. RMIs exist between each VAE. If the RMI following a VAE is populated with a risk management action or decision, the exposure inventory is modified and carried through to the next event (e.g. if the RMI is roof cleaning before further ash deposition, the ashfall thickness at the start of the next VAE is 0 mm, rather than accumulating the thick- ness between phases). 2.3 Direct impact: accounting for multi‑hazards Volcanic ash physical vulnerability models have been derived from post-event observa- tions, laboratory experimentation and expert elicitation (Spence et al. 2005; Stewart et al. 2006; Wilson et al. 2011, 2012, 2017; Jenkins et al. 2014b; Craig 2015; Thompson et al. 2017a), resulting in a suite of vulnerability models that typically assess vulnerability to ashfall using mathematical functions, e.g. fragility functions (Wilson et  al. 2014, 2017; Blake et  al. 2017). Vulnerability models regarding impacts from volcanic surface flows, such as lahars, are generally derived from post-event surveying (Baxter et al. 2005; Kün- zler et al. 2012; Jenkins et al. 2013, 2014a, 2015) and tend to determine threshold values that lead to gradational or absolute impact of the exposed asset, therefore generating a suite of binary or threshold vulnerability model types (Spence et al. 2004; Baxter et al. 2005; Wilson et al. 2014; Deligne et al. 2017b), though some fragility functions are available for building responses to flow hazards (Zuccaro and De Gregorio 2013; Jenkins et al. 2015). 9334 Natural Hazards (2024) 120:9327–9360 1 3 Despite recognition that volcanic eruptions are inherently multi-phase events that often produce multiple hazards (Jenkins et al. 2007; Bebbington and Jenkins 2019; Lerner et al. 2019), there are few, if any, vulnerability models that account for the compounding and cascading nature of volcanic multi-hazard impacts (Zuccaro et  al. 2008; Deligne et  al. 2017b). In this framework, we applied a simplistic approach to account for multi-hazard interactions in volcanic contexts. As available vulnerability models for volcanic surface flows and critical infrastructure assets (excluding buildings) assume maximum impact upon contact (Wilson et  al. 2014, 2017; Wilson 2015), we follow this approach in this framework. Surface flow vulnerability models can therefore be simply coupled with ashfall vulnerability models, by considering the accumulated impact, which is maximal if a lahar is involved. For multiple instances of ashfall, the HIM (in many cases, deposit thickness) is accumulated and the total HIM related to IS using an available physical vulnerability model. Remobilisation and cleaning of ashfall deposits can be applied in this framework, during the risk management intervention phases, and the resultant phases can be modified accordingly. 2.4 Indirect (cascading) impact to interdependent infrastructure networks The proposed dynamic volcanic impact assessment framework requires an understanding of the critical infrastructure downstream dependencies and interdependencies between infrastructure sectors and asses in the exposed networks. Direct damage to assets reduces the functionality of dependent assets (Rinaldi et al. 2001; Setola et al. 2009; Zorn 2017a; Hempel et  al. 2018). Recent advances in systems and network analysis quantify system dependencies for electricity, water supply and telecommunications in particular (Zorn and Shamseldin 2015, 2016, 2017; Thacker et al. 2017b; Zorn 2017b; He and Cha 2018) and allow the incorporation of these tools in impact assessment frameworks. The cascading (indirect) infrastructural impacts following direct impacts to assets can be estimated, by propagating infrastructure disruption downstream in an interdependent network graph (Weir 2021). Here we apply that method to the multi-hazard multi-phase eruption suite developed by Weir et al. (2022) and use it to quantify the systemic vulner- ability of the region to volcano events. Then, by introducing the opportunity for dynamic risk management intervention in the framework, we explore the degree to which this vul- nerability can realistically be mitigated. 3 Case study risk context: Taranaki Mounga volcano, Aotearoa New Zealand Taranaki Mounga (Mount Taranaki) is an active stratovolcano, located in the Taranaki region on the west coast of Te Ika-a-Māui the North Island of Aotearoa New Zealand (Fig. 2B). Volcanism at Taranaki Mounga is characterised by cycles of edifice growth and collapse (Zernack et al. 2011; Lerner et al. 2019; Cronin et al. 2021). The cone growth peri- ods are characterised by effusive and lava dome-forming activity, interspersed with explo- sive sub-Plinian to Plinian eruptions (Platz et  al. 2007; Turner et  al. 2011; Lerner et  al. 2019; Cronin et al. 2021). Typical far-reaching volcanic hazards include column-collapse and block-and-ash flow pyroclastic density currents (PDCs), ashfall, debris avalanches caused by edifice collapse, and frequent remobilisation of volcanic debris in the form of lahars (Fig. 2D). Proximal hazards, such as volcanic ballistic projectiles (VBPs) and lava 9335Natural Hazards (2024) 120:9327–9360 1 3 flows, are relatively infrequent in Taranaki Mounga’s eruptive history (Turner et al. 2011) and are typically constrained to the current extent of the Te Papakura o Taranaki National Park. Lerner et al. (2019) dated the most recent eruptive activity at Taranaki Mounga to 1780–1800 AD. This activity produced many small effusive and explosive eruptions, and a final dome-building event built the current summit topography (Neall 1972; Platz et al. 2012). Damaschke et al. (2018) estimates there to be a 33–42% probability of an eruption within the next 50 years. The Taranaki region has the second highest Gross Domestic Product (GDP) per capita in Aotearoa New Zealand (NZD $68,427 in 2018; NZD $58,778 was the national average) (Stats NZ 2018). Dairy, together with oil and gas production and processing, are the two key economic activities in the Taranaki region (Venture Taranaki 2017). State highways circumnavigate Taranaki Mounga, providing vital transport links along the western part of the North Island (Fig. 2C). Other critical networks include freight-only railway lines, elec- tricity transmission lines and oil and gas pipelines. The majority of local infrastructure is located in the three main population centres [New Plymouth (pop: 74,184), Stratford (pop: 9474) and Hawera (pop: 9810)] (Fig. 2C), with collection and transmission infrastructure ensuring continued supply to processing plants, and municipal supply of critical infrastruc- ture to rural consumers. The Taranaki region has 530 named rivers, most of which extend radially from Taranaki Mounga. Critical infrastructure such as telecommunications lines, water pipes, and oil and gas pipelines are often bound beneath road bridges or cross rivers as pipe bridges, mean- ing they are highly exposed to volcanic surface flows such as lahars and PDCs. Further, subaerial infrastructure such as electricity transmission and distribution lines are often co-located parallel to road transportation. Taranaki Mounga is within the Te Papakura o Taranaki National Park, which is managed by Te Papa Atawhai the Department of Con- servation (DoC). The national park extends ~ 10km radially from the volcanic summit Fig. 2 A The location of Aotearoa-New Zealand; B the location of the Taranaki region and volcanic cen- tres of Aotearoa New Zealand; C urban centres and selected critical infrastructure of the region; D simpli- fied geology of the Taranaki volcanic complex; E land use of the Taranaki region. Modified from Weir et al. (2022). The geographic extent of C, D and E is − 38.713711, 173.252031 (top left) to − 39.929024, 175.071420 (bottom right) 9336 Natural Hazards (2024) 120:9327–9360 1 3 (Fig. 2E), which limits urban development close to the volcanic edifice, thus ensuring min- imal exposure of the built environment to volcanic hazards. Regional Civil Defence and Emergency Management (CDEM) groups in Aotearoa New Zealand are entrusted with developing plans and strategies for the management of hazards relevant to their geographic region. Further, they are tasked with facilitating rela- tionships, discussion and knowledge transfer between regional agencies, stakeholders and groups (MCDEM 2005, 2010, 2015). The Taranaki CDEM group is therefore responsible for volcanic risk reduction, response and resilience planning in the Taranaki region. The Taranaki CDEM group recognises that strong partnerships with researchers is a key ele- ment of producing actionable science for disaster risk assessment and management (New Zealand Government 2015). Direct economic losses associated with a volcanic event at Taranaki only comprise a fraction of total expected loss: McDonald et al. (2017) calculates an ‘initial shock’ change in gross domestic product (GDP) of − 66% (− $4106 million NZD) in the Taranaki region alone for a high-intensity volcanic eruption scenario. Indirect impacts, such as supply chain distribution, staffing shortage, air travel disruption, electricity/water outages, and business non-continuity, could cause substantial economic losses and are highly variable both spa- tially and temporally (Mcdonald et al. 2017). 4 Case study application: direct impact assessment data and models Weir et al. (2022) developed a suite of multi-phase, multi-hazard volcanic eruption scenar- ios for Taranaki Mounga, using a modular scenario development framework. The scenario suite consists of nine hypothetical eruption scenarios (comprised of ‘small’ (S1, S2 and S3), ‘medium’ (M1, M2 and M3) and ‘large’ (L1, L2 and L3) sub-suites), each with six potential phases of volcanic activity [named phases 1d, 2p, 3d, 4p, 5d and 6p; where d is discrete (occurring over 1 day) and p is prolonged (occurring over weeks to months)] (Weir et al. 2022). The scenarios cover a credible range of volcanic magnitudes, styles, durations and hazards and were co-produced with end-users and decision-makers. Weir et al. 2022) includes many volcanic hazards as part of the scenario suite, but only provides spatial foot- prints for ashfall and lahar hazards. This is in part due to limited asset exposure within 10km of the summit (i.e. within Te Papakura o Taranaki National Park), which also con- strains other hazards, such as pyroclastic density currents (PDCs) (Procter et al. 2010). We therefore consider direct impact from only ashfall and lahars in this initial impact assess- ment application. Critical infrastructure exposure data were collected and compiled for six sectors: elec- tricity, water supply, road transportation, wastewater, energy (oil and gas) and telecom- munications. These six sectors are recognised to be of vital importance for national secu- rity and public welfare (Taranaki Civil Defence and Emergency Management Group 2018; Ministry of Civil Defence & Emergency Management (MCDEM) 2019; New Zealand Life- lines Council 2020; New Zealand Infrastructure Commission 2022). Membership of the Taranaki Civil Defence and Emergency Management (CDEM) Lifelines Advisory Group (LAG) is principally comprised of representatives from these six sectors, and the scope of critical infrastructure considered in this study was validated by end-users (Taranaki CDEM and the Taranaki CDEM LAG). Spatial and asset characteristic data for these six sectors were acquired from a variety of sources, some from open-source data repositories (Land Information New Zealand (LINZ); 9337Natural Hazards (2024) 120:9327–9360 1 3 Koordinates) and others were shared by the data owner(s). Data sharing by infrastructure providers was facilitated by Taranaki CDEM, the LAG, and their wider stakeholder net- work. Where appropriate, data sharing agreements were entered into, and where possible, any inconsistencies across data sources were addressed through consultation with data pro- viders. This process of data collection, unification and quality checking was undertaken during the aforementioned embedding and engagement process with Taranaki CDEM. In several instances, asset data were created or edited by the first author by referring to offi- cial open-source documentation and validated using satellite imagery (such as water supply inlets, inferred from the district council Water Asset Management Plans) (Table 3). Asset classes are broadly distinguished as either nodes (e.g. electricity substations or water treat- ment plants) or lines (e.g. electricity lines and water supply pipes) (Table  3). Line-type infrastructure was broken into 100m spans for the purpose of analysis, consistent with haz- ard data resolution (Weir et al. 2022). Several review studies compile global accounts and records of volcanic impacts to infrastructure and present an overview of the current global knowledge base (Wilson et al. 2014, 2017; Wilson 2015). Most vulnerability models used in the present application to Taranaki Mounga are to be found in these studies, supplemented by targeted studies on roads (Blake et al. 2017) and energy (Juniper 2018) sector vulnerability to volcanic haz- ards (Table  4). The vulnerability models used are either fragility functions (Blake et  al. 2017; Wilson et al. 2017) or damage matrices (Juniper 2018; Wilson et al. 2014). When applying fragility functions, a randomly generated number was used to generate one impact outcome, consistent with other recent fragility function application approaches (Williams et al. 2020, 2022). In Weir et  al. (2022), the hazard intensity metrics (HIMs) presented in the eruptions scenarios are deposit thickness (mm) for ashfall, and presence/absence for lahars, consist- ent with lahar vulnerability models for critical infrastructure assets in Wilson et al. (2014). Lahar vulnerability models available in Juniper (2018) use dynamic pressure (kPa) as the HIM, which is absent from the Weir et al. (2022) scenario suite. Therefore, in this study we assume energy assets exposed to lahars incur total damage, consistent with Wilson et al. (2014), where across all critical infrastructure sectors studied, lahar flow velocities > 0 m/s and lahar flow depth > 0 m result in the asset reaching the maximum impact state (in Wil- son et al. (2014), Damage Level 3: Replacement or financially expensive repair). When assets are exposed to multiple hazards, we apply the coupled physical vulner- ability approach for ashfall and lahars (Table 2). Typically, when an asset is exposed to a Table 2 Logic treatment of sequential ashfall and surface flow hazards in vulnerability model Hazard 1 Hazard 2 Ashfall Surface flow Ashfall Accumulate HIM (e.g. deposit thickness; ash loading) and apply in vulnerability model Assume ISmax; no change with suc- cessive ashfall or lahars Surface flow Assume ISmax; no change with successive ashfall or lahars Assume ISmax; no change with suc- cessive ashfall or lahars 9338 Natural Hazards (2024) 120:9327–9360 1 3 Ta bl e 3 T he c rit ic al in fr as tru ct ur e se ct or s, pr ov id er s, as se t c la ss es , a nd (g eo sp at ia l a nd a ss et c ha ra ct er ist ic ) d at a so ur ce s i n th is st ud y, w he re ‘D SA ’ i s d at a sh ar in g ag re em en t w ith p ro vi de r, ‘O S’ is o pe n- so ur ce d at a an d ‘F A S’ is fi rs t-a ut ho r- su pp le m en te d da ta . A ls o sh ow n ar e th e vu ln er ab ili ty m od el s ap pl ie d fo r ea ch a ss et c la ss , f or a sh fa ll an d la ha rs Se ct or Su b se ct or Pr ov id er A ss et C la ss Ty pe (n od e or lin e) D at a so ur ce A sh fa ll V ul ne ra bi lit y M od el La ha r V ul ne ra bi lit y M od el El ec tri ci ty Tr an sm is si on Tr an sp ow er N ew Ze al an d Lt d El ec tri ci ty L in es Li ne O S W ils on e t a l. (2 01 7) n/ a Su bs ta tio ns N od e O S W ils on e t a l. (2 01 7) W ils on e t a l. (2 01 4) Po le s N od e O S n/ a W ils on e t a l. (2 01 4) D ist rib ut io n Po w er co L td El ec tri ci ty li ne s Li ne D SA W ils on e t a l. (2 01 7) n/ a Su bs ta tio ns N od e D SA W ils on e t a l. (2 01 7) W ils on e t a l. (2 01 4) Po le s N od e D SA n/ a W ils on e t a l. (2 01 4) W at er su pp ly C ol le ct io n D ist ric t c ou nc ils (N PD C ; S D C ; ST D C ) Su rfa ce w at er in ta ke s; gr ou nd w at er in ta ke s (b or es ) N od e O S; F A S1 W ils on e t a l. (2 01 4) W ils on e t a l. (2 01 4) Tr ea tm en t D ist ric t c ou nc ils (N PD C ; S D C ; ST D C ) W at er tr ea tm en t p la nt s N od e O S; F A S1 W ils on e t a l. (2 01 7) W ils on e t a l. (2 01 4) D ist rib ut io n D ist ric t c ou nc ils (N PD C ; S D C ; ST D C ) W at er p ip es (i nc lu di ng ae ria l c ro ss in gs ) Li ne O S; F A S1 n/ a W ils on e t a l. (2 01 4) St or ag e D ist ric t c ou nc ils (N PD C ; S D C ; ST D C ) Re se rv oi rs N od e O S; F A S1 W ils on e t a l. (2 01 4) W ils on e t a l. (2 01 4) Ro ad tr an sp or ta tio n St at e H ig hw ay s W ak a K ot ah i N Z Tr an sp or t Ro ad s Li ne O S B la ke e t a l. (2 01 7) W ils on e t a l. (2 01 4) B rid ge s N od e O S; F A S2 As su m ed a s r oa ds ; B la ke e t a l. (2 01 7) W ils on e t a l. (2 01 4) O th er D ist ric t C ou nc ils (N PD C ; S D C ; ST D C ) Ro ad s Li ne O S B la ke e t a l. (2 01 7) W ils on e t a l. (2 01 4) B rid ge s N od e O S; F A S2 As su m ed a s r oa ds ; B la ke e t a l. (2 01 7) W ils on e t a l. (2 01 4) 9339Natural Hazards (2024) 120:9327–9360 1 3 Ta bl e 3 (c on tin ue d) Se ct or Su b se ct or Pr ov id er A ss et C la ss Ty pe (n od e or lin e) D at a so ur ce A sh fa ll V ul ne ra bi lit y M od el La ha r V ul ne ra bi lit y M od el W as te w at er Tr ea tm en t D ist ric t c ou nc ils (N PD C ; S D C ; ST D C ) W as te w at er tr ea tm en t pl an ts N od e O S W ils on e t a l. (2 01 7) W ils on e t a l. (2 01 4) Tr an sm is si on D ist ric t c ou nc ils (N PD C ; S D C ; ST D C ) W as te w at er p ip es (in cl ud in g ae ria l cr os si ng s) Li ne O S n/ a W ils on e t a l. (2 01 4) W as te w at er p um pi ng st at io ns N od e O S W ils on e t a l. (2 01 4) W ils on e t a l. (2 01 4) En er gy U ps tre am Va rio us e xp lo ra tio n an d pr od uc tio n co m - pa ni es (~ 20 ) W el l H ea ds N od e D SA n/ a Ju ni pe r ( 20 18 ) Pr od uc tio n fa ci lit ie s N od e D SA Ju ni pe r ( 20 18 ) Ju ni pe r ( 20 18 ) Tr an sm is si on (m id - str ea m ) Fi rs t G as L td (a nd pa rtn er s) Pi pe lin es (i nc lu di ng ae ria l c ro ss in gs ) Li ne D SA n/ a Ju ni pe r ( 20 18 ) Pi pe lin e su bs ta tio n N od e D SA n/ a Ju ni pe r ( 20 18 ) St or ag e Va rio us tr an sp or t, sto r- ag e an d pr oc es si ng co m pa ni es (~ 6) Ta nk s N od e D SA Ju ni pe r ( 20 18 ) Ju ni pe r ( 20 18 ) D ist rib ut io n Po w er co L td Pi pe lin es Li ne D SA n/ a Ju ni pe r ( 20 18 ) Te le co m m un ic at io ns C ol le ct io n Vo da fo ne N Z Lt d, Sp ar k N Z Lt d, 2 D eg re es M ob ile L td C el lu la r t ow er s N od e O S; F A S1 W ils on e t a l. (2 01 4) W ils on e t a l. (2 01 4) Tr an sm is si on C ho ru s N Z Lt d Ex ch an ge s N od e O S; F A S1 W ils on e t a l. (2 01 4) W ils on e t a l. (2 01 4) 1 D at a in co ns ist en ci es b et w ee n di str ic t c ou nc ils w er e su pp le m en te d us in g pu bl ic al ly av ai la bl e as se t m an ag em en t p la ns a nd sa te lli te im ag er y 2 D at a su pp le m en te d by in fe rr in g br id ge lo ca lit ie s a t t he in te rs ec tio n of ro ad s a nd ri ve rs (p ub lic al ly av ai la bl e da ta se ts ) 9340 Natural Hazards (2024) 120:9327–9360 1 3 Ta bl e 4 Im pa ct m et ric s an d ha za rd in te ns ity m et ric (H IM ) re qu ire m en ts fo r t he v ul ne ra bi lit y m od el s us ed in th is im pa ct a ss es sm en t a pp lic at io n, a nd d et ai ls o f co nv er si on s m ad e fro m a va ila bl e H IM s i n W ei r e t a l. (2 02 2) . A ls o sh ow n ar e co nv er si on s m ad e be tw ee n im pa ct m et ric s i n av ai la bl e stu di es , a nd th e im pa ct m et ric u se d in th is st ud y (W il- so n et  a l. 20 17 ) St ud y D at a ty pe Im pa ct m et ric de sc rip tio n H IM re qu ire - m en ts (a sh fa ll) H IM re qu ire - m en ts (l ah ar ) Im pa ct m et ric ti er s an d cl as si fic at io ns H IM c on ve rs io n fo r t hi s a pp lic a- tio n (a sh fa ll) H IM c on ve r- si on fo r t hi s ap pl ic at io n (la ha r) IS c on ve rs io n fo r th is a pp lic at io n Ju ni pe r ( 20 18 ) Ex pe rt ju dg e- m en t Im pa ct S ta te (I S) ; 4 -ti er A sh fa ll lo ad (k Pa ) D yn am ic p re s- su re (M Pa ) D 0 N o ph ys ic al da m ag e C on ve rt th ic k- ne ss (m m ) t o lo ad (k Pa ), as su m in g a de po si t de ns ity o f 10 00  k g  m − 3 A ss um e pr es - en ce o f l ah ar (> 0  M Pa ) re su lts in D 3 D 0 IS 0 ( W ils on e t a l. 20 17 ) D 1 M in or (n o str uc tu ra l) da m ag e D 1 IS 1 ( W ils on e t a l. 20 17 ) D 2 M aj or (s tru c- tu ra l) da m ag e D 2 IS 2 ( W ils on e t a l. 20 17 ) D 3 To ta l p hy si ca l as se t d es tru c- tio n D 3 IS 3 ( W ils on e t a l. 20 17 ) B la ke e t a l. (2 01 7) Em pi ric al (im pa ct da ta ); ex pe rt ju dg em en t; la bo ra to ry ex pe rim en ta - tio n Im pa ct S ta te (I S) ; 4 -ti er ; di sr up tio n D ep os it th ic k- ne ss (m m ) n/ a IS 0 N o di sr up tio n n/ a n/ a IS 0 IS 0 ( W ils on e t a l. 20 17 ) IS 1 Sk id re si st an ce re du ct io n IS 1 IS 1 ( W ils on e t a l. 20 17 ) IS 2 Im pa ss ab le fo r so m e ve hi cl es IS 2 IS 2 ( W ils on e t a l. 20 17 ) IS 3 Im pa ss ab le if te ph ra u nc on - so lid at ed IS 3 IS 3 ( W ils on e t a l. 20 17 ) 9341Natural Hazards (2024) 120:9327–9360 1 3 Ta bl e 4 (c on tin ue d) St ud y D at a ty pe Im pa ct m et ric de sc rip tio n H IM re qu ire - m en ts (a sh fa ll) H IM re qu ire - m en ts (l ah ar ) Im pa ct m et ric ti er s an d cl as si fic at io ns H IM c on ve rs io n fo r t hi s a pp lic a- tio n (a sh fa ll) H IM c on ve r- si on fo r t hi s ap pl ic at io n (la ha r) IS c on ve rs io n fo r th is a pp lic at io n W ils on e t a l. (2 01 7) Em pi ric al (im pa ct d at a) ; ex pe rt ju dg e- m en t Im pa ct S ta te (I S) ; 4 -ti er ; ag gr eg at e of p hy si ca l da m ag e an d di sr up tio n D ep os it th ic k- ne ss (m m ) n/ a IS 0 N o da m ag e n/ a n/ a n/ a IS 1 C le an in g re qu ire d IS 2 Re pa ir re qu ire d IS 3 Re pl ac em en t or fi na nc ia lly ex pe ns iv e re pa ir W ils on e t a l. (2 01 4) Em pi ric al (im pa ct d at a) ; ex pe rt ju dg e- m en t Im pa ct /D am - ag e Le ve l (D L) ; 4 -ti er ; ag gr eg at e of p hy si ca l da m ag e an d di sr up tio n D ep os it th ic k- ne ss (m m ) Fl ow d ep th (m ) or fl ow v el oc - ity (m /s ) D L0 N o da m ag e n/ a A ss um e pr es - en ce o f l ah ar re su lts in fl ow de pt h > 0  m n/ a D L1 C le an in g re qu ire d D L2 Re pa ir re qu ire d D L3 Re pl ac em en t or fi na nc ia lly ex pe ns iv e re pa ir 9342 Natural Hazards (2024) 120:9327–9360 1 3 lahar, total asset damage is incurred (Wilson et al. 2014), and thus in this study, we assume the maximum impact state until the end of the scenario, regardless of further exposure to ashfall and lahar hazards. For multiple instances of ashfall, the thickness is accumulated and the total thickness is related to an impact state using the vulnerability models listed in Table 3. The impact metrics used in the physical vulnerability models applied in this study are shown in Table 4. For this application, we adopt the four-tiered ‘Impact State’ (IS) met- ric used by (Wilson et al. 2017) and Wilson et al. (2014) and make necessary conversions from Blake et al. (2017) and Juniper (2018) (Table 4). To quantify indirect impact on infrastructure sectors, we use the graphical approach and tools developed for the Taranaki region in Weir (2021). The graphical approach for interde- pendent infrastructure determines the reach of source to sink resource flow following direct impact to assets. In this application, we assume assets are functional under IS0 and IS1 (no damage and cleaning required, respectively) and complete loss of functionality under IS2 and IS3 (repair required and replacement required (or financially expensive repair), respectively). Assets determined to be at IS2 or IS3 following direct impact from volcanic hazards are therefore absent in the resultant network graph configuration, and the source to sink disruption is recorded and carried through to the next event in the framework. Indirect impact (i.e. loss of service due to physical damage upstream of resource flow) is assumed to be binary, and assets indirectly impacted are recorded as being ‘indirectly disrupted’ (ID). 5 Longitudinal volcanic multi‑hazard impact assessment results 5.1 Direct impact assessment: results across sectors The direct, physical impact to assets for six infrastructure sectors (electricity, water supply, waste water, oil and gas, transportation and telecommunications) were calculated for all nine eruption scenarios in the scenario suite. An IS is given for each asset in the inventory for each phase of volcanic activity (1d, 2p, 3d, 4p, 5d and 6p; where d is ‘discrete’ (phase duration 1 day), and p is ‘prolonged’ [phase duration is user specified) (Weir et al. 2022)], providing a longitudinal direct impact analysis under variable, credible volcanic regimes. Direct impact data is available for each sector (n = 6), at each phase (n = 6) of each scenario (n = 9), resulting in 324 data frames (though 30 of these data frames show no change in direct impact due to volcanic inactivity). The direct impact to the electricity sector (trans- mission and distribution) is shown in Fig. 3, and the direct impact to water supply in Fig. 4. Figure 5 shows the direct impact to electricity infrastructure at each volcanic activity phase during the scenario (phases 1d, 2p, 3d and 4p), using scenario M2 as a representative mid- range scenario. Across all sectors, 58% of nodes achieve IS3 in at least one scenario (Table 5), with 40% of nodes at IS3 for at least two of the eruption scenarios. One water supply inlet in South Taranaki district, upstream of a large municipal supply catchment, reaches IS3 in 8 of the 9 eruption scenarios. Of the 21 nodes that reach IS3 for 5 or more scenarios, 11 of these nodes are water supply nodes (9 surface water inlets, 1 water treatment plant and 1 reservoir), 5 are oil and gas production or transmission sites, 2 are electricity distribution substations, 2 are cell towers and 1 is a waste water treatment plant. There are 563 occurrences of a node asset reaching IS3 across all sectors, for all scenarios. Of these, 63% are due to exposure to lahars (Table 6). There are 167 node assets that reach 9343Natural Hazards (2024) 120:9327–9360 1 3 IS2, corresponding to ‘requires repair’, at least once throughout the scenario suite from expo- sure to ashfall. Of these nodes 28% are water supply nodes, 19% are waste water, 19% are telecommunications, 17% are electricity, and 17% are oil and gas. For infrastructure lines, the total length in different ISs was determined for the subaerial electricity and transportation sec- tors. For other sectors with buried line infrastructure (water supply, waste water and oil and gas), the total number of breakage points were summed, as impact was assumed only in topo- graphic depressions upon contact with a lahar, where buried infrastructure is generally carried by road or pipe bridges. The total length and number of breakage points for each scenario are shown in Table 7. Fig. 3 The direct impact to electricity transmission and distribution infrastructure in the Taranaki region of Aotearoa New Zealand. The impact is shown at volcanic activity phase 6p (i.e. at the end of the eruption sequence) for each of the nine scenarios (S1–L3) 9344 Natural Hazards (2024) 120:9327–9360 1 3 5.2 Indirect impact assessment for interdependent infrastructure The indirect impact is quantified in terms of number of assets (nodes) indirectly disrupted (ID) and the number of private dwellings without service (PDWS), as in Weir  (2021). The indirect impact was calculated for each phase (n = 6) of each scenario (n = 9), across all sectors, resulting in 54 data frames. Figure 6 shows the direct impact to all infrastruc- ture sectors across the phases of volcanic activity, and Fig. 7 shows the direct and indirect impact for the same phases. Inter-sector indirect impact occurs to all infrastructure sec- tors during Scenario M2, with the exception of the transportation sector, as transportation is not dependent on any other sectors included in the infrastructure systems tool. During Scenario M2 the greatest indirect impact is to the oil and gas sector. Due to the sector’s strong dependence on electricity, and the pinchpoints present in the network, supply is dis- rupted from the first phase of volcanic activity (M2_1d) (Fig. 7). Electricity supply across west and south Taranaki is disrupted due to structural damage of high voltage transmission Fig. 4 The direct impact to water supply infrastructure in the Taranaki region of Aotearoa New Zealand. The impact is shown at volcanic activity phase 6p (i.e. at the end of the eruption sequence) for each of the nine scenarios (S1–L3) 9345Natural Hazards (2024) 120:9327–9360 1 3 Fig. 5 The direct impact to electricity transmission and distribution infrastructure for phases 1d, 2p, 3d and 4p of eruption scenario M2. Scenario M2 has no volcanic activity (including post-eruptive lahars) during phases 5d and 6p, therefore map M2_4p is the final phase for direct impact in scenario M2 Table 5 The number of node assets per sector in IS2 (requires repair) or IS3 (requires replacement) at end of each eruption scenario Taranaki Mounga volcanic eruption scenario (Weir et al. 2022) Critical infrastructure sector S1 S2 S3 M1 M2 M3 L1 L2 L3 Electricity (n = 50) 0 0 0 7 18 14 28 21 14 Water supply (n = 96) 1 3 6 18 33 33 49 57 24 Waste water (n = 80) 0 0 0 7 31 9 32 37 5 Oil and gas (n = 61) 0 7 4 4 20 21 26 26 23 Transportation (n = 16) 0 0 1 2 5 3 12 10 10 Telecommunications (n = 111) 0 1 4 6 19 19 41 36 20 Total 1 11 14 42 121 96 176 177 85 9346 Natural Hazards (2024) 120:9327–9360 1 3 infrastructure on the southern side of Taranaki Mounga, coupled with disruption of a transmission substation in Hawera. Dependent assets across all sectors in south and west Taranaki are therefore indirectly impacted. Inter-sector indirect impact is observed for all sectors besides transportation (Fig.  7), due to lack of transportation interdependence with other sectors. Intra-sector indirect impact is observed across all sectors, but particularly in electricity, water supply, waste water, and oil and gas, due to their heavy reliance on resource flow from a limited number of nodes. Physical impact to the electricity sector is the most common cause of inter-sector indirect impact, due to the strong interdependencies within the network. Generally, low-intensity scenarios (S1–S3) result in minimal indirect impact and are largely restricted to the water supply and oil and gas sectors. Scenario S2, however, causes IS2 physical impact to transmission and distribution lines supplying New Plymouth and hence encompasses high indirect impact relative to the other scenarios in the ‘S’ (small) suite. For the larger scenarios (M1 – L3), total network disruption is often observed, with near-complete disruption of the electricity and oil and gas sector observed for Scenarios Table 6 The number of node assets per sector that achieve IS3 (requires replacement) as a result of lahars for each eruption scenario Taranaki Mounga volcanic eruption scenario (Weir et al. 2022) Critical infrastructure sector S1 S2 S3 M1 M2 M3 L1 L2 L3 Electricity (n = 50) 0 0 0 2 2 13 13 15 3 Water supply (n = 96) 1 2 6 7 8 32 29 37 16 Waste water (n = 80) 0 0 0 1 1 8 19 4 2 Oil and gas (n = 61) 0 4 4 1 4 20 16 19 8 Transportation (n = 16) 0 0 1 2 5 3 12 10 10 Telecommunications (n = 111) 0 1 0 0 0 14 20 20 7 Total 1 7 10 9 15 87 97 95 36 Table 7 Infrastructure lines for each sector in IS2 (requires repair) or IS3 (requires replacement) at the end of each scenario. For electricity lines and transportation, the length in IS2 or IS3 is given. For water sup- ply, wastewater and oil and gas, the total number of breakage points is given, as impact by lahar to buried infrastructure was assumed in topographic depressions (where the lines are generally carried by road or pipe bridges) Taranaki Mounga volcanic eruption scenario (Weir et al. 2022) Critical infrastructure sector S1 S2 S3 M1 M2 M3 L1 L2 L3 Length (km) Electricity (915 km) 0 km 18 km 0 km 98 km 277 km 176 km 406 km 450 km 214 km Road (390 km) 0 km 0 km 4 km 12 km 23 km 19 km 58 km 47 km 56 km Breakage point count Water supply (176 km) 1 6 9 8 13 19 28 35 15 Waste water (154 km) 0 2 5 8 10 17 23 27 11 Oil and gas (531 km) 0 4 7 3 6 12 12 15 8 Telecommunications (n/a) – – – – – – – – – 9347Natural Hazards (2024) 120:9327–9360 1 3 M2, M3, L1 and L2, with near-complete disruption of oil and gas (but not electricity) observed in Scenario L3. Physical impact to Stratford grid exit point (GXP), a transmission substation located in Stratford, which services nearly the entire region with electricity, is the cause of the most substantial indirect impact and is observed in scenarios M2, M3, L1 and L2. A summary of the total nodes directly or indirectly impacted by each scenario is shown in Table 8. The highest count of nodes directly or indirectly impacted occurs in scenario L2, where both Stratford GXP and Hawera electricity transmission substation are both at IS3. This is also the case for scenario M2, the scenario with the second highest count of disrupted and damaged nodes. Generally, the most extensive indirect impact is seen during volcanic activity phases 1d and 2p across the entire scenario suite, with ashfall provoking direct and indirect impact during phase 1d and lahars provoking direct and indirect impact in phase 2p. Across all scenarios, the indirect impact in phases 3d, 4p, 5d and 6p is mini- mal relative to phases 1d and 2p. Of all nodes that reach IS2 or IS3 during the analysis, the Fig. 6 The direct impact to all six infrastructure sectors for phases 1d, 2p, 3d and 4p of scenario M2. Sce- nario M2 has no volcanic activity (including post-eruptive lahars) during phases 5d and 6p, therefore map M2_4p is the final phase for direct impact in scenario M2 9348 Natural Hazards (2024) 120:9327–9360 1 3 Fig. 7 The direct and indirect impact to all six infrastructure sectors for phases 1d, 2p, 3d and 4p of sce- nario M2. Scenario M2 has no volcanic activity (including post-eruptive lahars) during phases 5d and 6p, therefore map M2_4p is the final phase for direct and indirect impact in scenario M2 Table 8 The number of nodes in Impact State 2 (IS2; requires repair), Impact State 3 (IS3; requires replace- ment) or indirectly disrupted (ID) in the last phase of volcanic activity (6p) for each scenario Taranaki Mounga volcanic eruption scenario (Weir et al. 2022) Critical infrastructure sector S1 S2 S3 M1 M2 M3 L1 L2 L3 Electricity (n = 50) 0 19 4 23 48 33 48 50 40 Water supply (n = 96) 1 27 11 19 93 55 86 95 71 Waste water (n = 80) 0 22 7 24 70 53 71 73 69 Oil and gas (n = 61) 0 34 56 56 56 49 55 58 37 Transportation (n = 16) 0 0 1 2 5 3 12 10 10 Telecommunications (n = 111) 0 31 5 7 101 46 89 109 68 Total 1 133 85 129 368 236 349 385 285 9349Natural Hazards (2024) 120:9327–9360 1 3 five single nodes which result in the most considerable indirect impact are Stratford trans- mission GXP, Hawera electricity transmission substation, Ōpunake electricity transmission substation, the gas transmission pipe bridge in Hangatahua (Stony) river catchment (SW of volcanic edifice), and Kapuni water treatment plant (WTP). Infrastructure lines servicing these nodes are also frequently reaching IS2 and IS3, resulting in extensive indirect impact. 5.3 Risk management scenarios: testing efficacy The timing and nature of risk management action(s) is/are crucial in reducing the adverse impact of hazardous events. The decision-making phases of the framework (I2, I4 … In) allow the testing of mitigation strategies. Figure 8 uses a hypothetical volcanic eruption scenario to illustrate the potential reduction in the impact state of assets given different tim- ing of intervention strategies. We include risk management interventions in Scenario M2 to demonstrate the full extent of the dynamic impact assessment framework. These risk man- agement interventions were identified through end-user consultation and were determined to be feasible measures to implement during an ongoing volcanic sequence at Taranaki Mounga. Here, we test two hypothetical risk management decisions: cleaning of electric- ity substations and lines in IS1 (‘cleaning required’) and IS0 (‘no damage’) during I2, and the physical covering of Stratford grid exit point (GXP) transmission substation, therefore Fig. 8 An illustration of the testing of mitigation strategies. For Assets 1 and 2 (both subject to multi-haz- ards), the timing of the ‘mitigation strategy’ (black exclamation icon) affects the resultant impact state of the asset 9350 Natural Hazards (2024) 120:9327–9360 1 3 mitigating ashfall impacts to this site, during I4 (Table 9). The volcanic activity phases are renamed in Table 9 to suit the inclusion of risk management interventions. Figure 9 shows the direct and indirect impact of volcanic eruption scenario M2 with the addition of risk management phases outlines in Table 9. When comparing Fig. 7 with Fig. 9, we see that the cleaning of electricity lines and substations (during I2) and the phys- ical mitigation of Stratford GXP (during I4) results in a large reduction in the potential direct and indirect impact. This is particularly apparent in the difference between the indi- rect impact for both figures, due to the high system dependency on Stratford GXP. 6 Discussion This study integrates the growing knowledge base surrounding the co-production of dis- aster knowledge (Hicks et al. 2014; Davies et al. 2015; Whybark 2015; Fearnley and Bea- ven 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 vulnerabil- ity to volcanic hazards (Wilson et  al. 2012, 2014; Dominguez 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. The framework presented in this study meets the grow- ing end-user need for impact-based decision-support tools and allows the testing of mitiga- tion strategies under different volcanic regimes and can be customised for applications to other volcanic risk contexts. 6.1 Longitudinal modular impact assessment Designing flexibility and modularity into impact assessment frameworks enhances the rel- evance of the outputs in practice and policy, but also in science, where the ever-chang- ing state of knowledge necessitates the continual refinement and modification of previous work. Given volcanic uncertainty, there is a need for rapid volcanic response tools that are dependent on the current (or immediate previous) state of the volcano, in appreciation of the temporal dependencies of activity (Bebbington and Jenkins 2019; Jenkins et al. 2007). Table 9 Framework nomenclature for the application of risk management interventions to volcanic eruption scenario M2. Volcanic activity phases are renamed and two risk management interventions are included in the framework (I2 and I4). No risk management intervention occurs during I6 E1 I2 E3 I4 E5 I6 E7 Volcanic activity events in scenario M2 Phase name (Weir et al. 2022) M2_1d – M2_2p – M2_3d – M2_4p Risk management interventions in scenario M2 Description – Cleaning of electricity infrastructure in IS0 and IS1 – Mitigation of ash- fall at Stratford GXP – – – 9351Natural Hazards (2024) 120:9327–9360 1 3 Furthermore, given the potential long-duration of volcanic activity and resultant impacts, there is a need for quick-to-implement testing tools for decision-making, response and miti- gation planning. The frameworks presented here have the potential for rapid modification and re-design to suit different operational and research contexts, and support decision-mak- ers in the dynamic highly charged emergency planning and response environment. The inclusion of risk management interventions in this framework was spurred from recent advances in the science field, which recognise the utility of iterative mitigation test- ing and planning (Dominguez et al. 2021; UNDRR 2019; UNISDR 2015; Zuccaro et al. 2018). Disruption of critical infrastructure service has implications for emergency manage- ment, particularly for welfare and building habitability, and implications for national-scale disruption of resource supply. Direct damage of assets introduces higher repair costs and Fig. 9 The direct and indirect impact to all six infrastructure sectors for scenario M2 with the implementa- tion of mitigation measures. Phase E1 shows volcanic activity phase M2_1d, phase I2 involves the cleaning of electricity lines and substations in IS0 and IS1, phase E3 shows volcanic activity phase M2_2p, phase I4 involves the physical mitigation of Stratford GXP to volcanic ashfall, phase E5 show volcanic activity phase M2_3d, phase I6 involves no risk management decisions, and phase E7 shows volcanic activity phase M2_4p. Scenario M2 has no volcanic activity (including post-eruptive lahars) during phases M2_5d and M2_6p, therefore map M2_4p is the final phase for direct and indirect impact in scenario M2 9352 Natural Hazards (2024) 120:9327–9360 1 3 requires labour, access and time to restore operations, which may not be possible during periods of volcanic uncertainty. This framework provides a platform for this exploration and enables and empowers risk managers to take a more active role in science production. The modularity of the framework facilitates application in other volcanic risk contexts. Volcanic hazard footprints, risk management interventions, exposure inventories, direct impact and indirect impact modules can be substituted in the impact assessment frame- work. These modifications can be made at various times during the disaster planning, response and recovery life cycle (i.e. pre-event, syn-event and post-event). 6.2 Implications for risk management at Taranaki Mounga The results of this study illustrate that the indirect impact to infrastructure networks from volcanic hazards is potentially far more extensive and pervasive than direct impact to infrastructure assets during the majority of modelled scenarios. Generally, the majority of nodes in the interdependent network graph without supply are due to indirect supply out- ages, rather than due to physical impact of infrastructure, and the number of nodes sys- temically disrupted, and the spatial reach of the disruption, remains fairly consistent across many of the volcanic eruption scenarios. The commonalities in provoking widespread indi- rect impact seem to be direct damage (IS2 or IS3) to Stratford GXP, coupled with direct damage to Hawera electricity transmission substation, or a combination of neighbouring electricity distribution substations. This highlights the strong dependency of the distributed infrastructure network on the electricity sector in particular. We find the oil and gas sector to be highly systemically vulnerable to volcanic haz- ards due to several factors: the strong dependence on electricity supply, the exposure of resource-carrying pipes in known lahar channels, and the lack of redundancy in resource transport. The oil and gas sector is of great national significance (New Zealand Lifelines Council 2020; New Zealand Lifelines Council 2017), and further exploration of the down- stream impacts of oil and gas disruption is required. The water supply sector is found to be most vulnerable in the South Taranaki District, where spatially extensive municipal supply schemes regularly traverse topographic depres- sions. The propensity for municipal supply schemes in Taranaki to be fed by surface water inlets introduces high systemic vulnerability, which cascades through interdependent sec- tors, such as waste water and the oil and gas sector. In consideration of these issues, the merit of the risk management modules of the sce- nario framework becomes apparent. The study shows that for potential eruptive activity at Taranaki Mounga, mitigation measures see the most potential benefit after the initial phases of volcanic activity (before phase 3d); however, the study also shows that this is a sensitive, uncertain environment that requires further investigation and thorough explora- tion of possible risk treatment strategies, and optimal intervention opportunities. 6.3 Recommendations for future volcanic multi‑hazard impact assessment There is considerable scope for the framework to be adopted and tested in other volcanic risk contexts, nationally and globally. Future studies could pair the impact assessment framework with information regarding periods of volcanic unrest (and associated deci- sions regarding emergency management and infrastructure management), and response and recovery decisions. However, applications and iterations of this work will need to take 9353Natural Hazards (2024) 120:9327–9360 1 3 similar care to mitigate the risks associated with engaging practitioners, carefully manag- ing response capacity due to limited access, staffing and resources. In future work, it must be acknowledged that periods of quiescence can vary hugely, and this should be incorporated in future iterations of this work, through the mechanisms employed in Weir et al. (2022) to develop the eruption scenario suite. In tandem, the dura- tion and timing of risk management interventions and of mitigation measure implemen- tation should be carefully considered in partnership with stakeholders and users. Further facets of vulnerability could be included in the framework. Here, physical and systemic vulnerability were considered with regards to distributed infrastructure, but in the future, social and economic vulnerability could be incorporated. We also encourage the investi- gation of habitability following lifeline disruption due to volcanic activity, and how life- safety and wellbeing concerns (due to volcanic hazard and critical infrastructure disrup- tion, respectively) will inform evacuations and emergency supply distribution. Further interdependencies could be considered as part of this work, such as fuel supply, supply chains, landline telecommunications, storm water and rail, port and airports. In par- ticular, we suggest the more granular analysis of road network disruption, as road function- ality is such as crucial component of emergency response. 7 Conclusions The framework presented and applied in this study shows the benefit of longitudinal vol- canic multi-hazard impact assessment. Volcanic eruptions can be, and often are, multi- phase, multi-hazard, complex events that necessitate a similarly multi-phase societal response. Allowing for risk management interventions between periods of volcanic activity allows more targeted, informed mitigation planning and meets the identified end-user need for impact-based metrics and tools when planning for disasters. This study utilised the growing knowledge base surrounding co-production of disaster knowledge, spatio-temporal dependencies in volcanic systems, and infrastructure vulner- ability to volcanic hazards to conduct a dynamic impact assessment for distributed infra- structure surrounding Taranaki Mounga. This study applied a suite of multi-phase volcanic multi-hazard eruption scenarios to estimate the direct and indirect impact of these sce- narios for infrastructure. The dynamic, modular, flexible nature of these impact scenarios allows for investigation of the spatio-temporal linkages inherent in volcanic systems, and how they manifest in impacts to the built environment and surrounding communities. The volcanic impact assessment outputs in this study could be paired with other tools to investigate the national-level disruption from Taranaki volcanism, which aligns well with Aotearoa New Zealand’s risk profile (Willis 2014). Of particular note is the independence and recovery model developed for national-scale critical infrastructure network disruption (Zorn 2017a, 2017b). The frameworks could feasibly include other dimensions of systemic risk besides cascading infrastructure failure. The could align well with the identified need for more targeted analysis of social, economic and environment aspects of vulnerability (Few et  al. 2017; Hicks and Few 2015; Kappes et  al. 2012). The feasibility of applica- tion of these frameworks to other geological hazards is unknown, but has the potential to offer great reward to practitioners that manage large risk profiles in an evolving hazard landscape. For the Taranaki region, these tools provide a test bed for mitigation strategies. These hazard and impact scenarios should be viewed as demonstrative of the diversity of impacts 9354 Natural Hazards (2024) 120:9327–9360 1 3 expected from future activity at Taranaki Maunga, and as such, should be used to test and plan for the best response and resilience-building strategies. Costing of different risk man- agement actions can and should be considered in order to make an evidence-based case for infrastructure investment. Acknowledgements We would like to thank all those that contributed to the development of the impact assessment framework and the application to Taranaki Mounga: Todd Velvin and the wider Taranaki Civil Defence and Emergency Management Group, the Taranaki Lifelines Advisory Group, New Plymouth Dis- trict Council, Stratford District Council, South Taranaki District Council, Taranaki Regional Council, Pow- erco and Firstgas and Dr Natalia I. Deligne (United States Geological Survey, GNS Science Te Pū Ao). Author contributions All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Dr. AMW. The first draft of the manuscript was written by Dr. AMW and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Funding Open Access funding enabled and organized by CAUL and its Member Institutions. We gratefully acknowledge funding support from the Natural Hazard Research Platform (NHRP) (Grant 2015-MAU-01- NHRP), the He Mounga Puia Transitioning Taranaki to a Volcanic Future programme (Grant UOAX1913) and the Resilience to Natures Challenges (RNC) Volcano (Grant GNS-RNC047), Rural (Grant GNS- RNC045) and Multihazard Risk (Grant GNS-RNC043) programmes. Declarations Conflict of interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Com- mons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. References AghaKouchak A, Huning LS, Chiang F et  al (2018) How do natural hazards cascade to cause disasters? Nature 561:458–460 Ang PS, Bebbington MS, Lindsay JM, Jenkins SF (2020) From eruption scenarios to probabilistic volcanic hazard analysis: an example of the Auckland volcanic field, New Zealand. J Volcanol Geoth Res 397:106871. https:// doi. org/ 10. 1016/j. jvolg eores. 2020. 106871 Armijos MT, Phillips J, Wilkinson E et al (2017) Adapting to changes in volcanic behaviour: formal and informal interactions for enhanced risk management at Tungurahua Volcano, Ecuador. Glob Environ Chang 45:217–226. https:// doi. org/ 10. 1016/j. gloen vcha. 2017. 06. 002 Barclay J, Haynes K, Mitchell T et  al (2008) Framing volcanic risk communication within disaster risk reduction: finding ways for the social and physical sciences to work together. Geol Soc Spec Pub 305:163–177. https:// doi. org/ 10. 1144/ SP305. 14 Barton TM, Beaven SJ, Cradock-Henry NA, Wilson TM (2020) Knowledge sharing in interdisciplinary dis- aster risk management initiatives: cocreation insights and experience from New Zealand. Ecol Soc 25:1–18. https:// doi. org/ 10. 5751/ ES- 11928- 250425 Bathrellos GD, Skilodimou HD, Chousianitis K et al (2017) Suitability estimation for urban development using multi-hazard assessment map. Sci Total Environ 575:119–134. https:// doi. org/ 10. 1016/j. scito tenv. 2016. 10. 025 http://creativecommons.org/licenses/by/4.0/ https://doi.org/10.1016/j.jvolgeores.2020.106871 https://doi.org/10.1016/j.gloenvcha.2017.06.002 https://doi.org/10.1144/SP305.14 https://doi.org/10.5751/ES-11928-250425 https://doi.org/10.1016/j.scitotenv.2016.10.025 https://doi.org/10.1016/j.scitotenv.2016.10.025 9355Natural Hazards (2024) 120:9327–9360 1 3 Baxter PJ, Boyle R, Cole P et  al (2005) The impacts of pyroclastic surges on buildings at the eruption of the Soufrière Hills volcano, Montserrat. Bull Volcanol 67:292–313. https:// doi. org/ 10. 1007/ s00445- 004- 0365-7 Bebbington MS, Jenkins SF (2019) Intra-eruption forecasting. Bull Volcanol 81:34. https:// doi. org/ 10. 1007/ s00445- 019- 1294-9 Bebbington MS, Jenkins SF (2022) Intra-eruption forecasting using analogue volcano and eruption sets. J Geophys Res Solid Earth 127:e2022JB024343. https:// doi. org/ 10. 1029/ 2022J B0243 43 Biass S, Scaini C, Bonadonna C et al (2014) A multi-scale risk assessment for tephra fallout and airborne concentration from multiple Icelandic volcanoes - Part 1: hazard assessment. Nat Hazard 14:2265– 2287. https:// doi. org/ 10. 5194/ nhess- 14- 2265- 2014 Biass S, Falcone JL, Bonadonna C et al (2016) Great balls of fire: a probabilistic approach to quantify the hazard related to ballistics—a case study at la Fossa volcano, Vulcano Island, Italy. J Volcanol Geoth Res 325:1–14. https:// doi. org/ 10. 1016/j. jvolg eores. 2016. 06. 006 Blake DM, Deligne NI, Wilson TM, Wilson G (2017) Improving volcanic ash fragility functions through laboratory studies: example of surface transportation networks. J Appl Volcanol 6:16. https:// doi. org/ 10. 1186/ s13617- 017- 0066-5 Cash DW, Clark WC, Alcock F et al (2003) Knowledge systems for sustainable development. Proc Natl Acad Sci USA 100:8086–8091. https:// doi. org/ 10. 1073/ pnas. 12313 32100 Cherry KE, Sampson L, Nezat PF et al (2015) Long-term psychological outcomes in older adults after disaster: relationships to religiosity and social support. Aging Ment Health 19:430–443. https:// doi. org/ 10. 1080/ 13607 863. 2014. 941325 Coppola DP (2011) Introduction to international disaster management, 2nd edn. Butterworth-Heine- mann, Oxford, pp 377–423 Craig HM (2015) Agricultural vulnerability to tephra fall impacts. University of Canterbury, Canterbury Cronin SJ, Zernack AV, Ukstins IA et al (2021) The geological history and hazards of a long-lived stra- tovolcano, Mt. Taranaki, New Zealand. N Z J Geol Geophys. https:// doi. org/ 10. 1080/ 00288 306. 2021. 18952 31 Damaschke M, Cronin SJ, Bebbington MS (2018) A volcanic event forecasting model for multi- ple tephra records, demonstrated on Mt. Taranaki. N Z Bull Volcanol. https:// doi. org/ 10. 1007/ s00445- 017- 1184-y Davies TRH, Davies AJ (2018) Increasing communities’ resilience to disasters: an impact-based approach. Int J Disaster Risk Reduct 31:742–749. https:// doi. org/ 10. 1016/j. ijdrr. 2018. 07. 026 Davies T, Beaven S, Conradson D et al (2015) Towards disaster resilience: a scenario-based approach to co-producing and integrating hazard and risk knowledge. Int J Disaster Risk Reduct 13:242–247. https:// doi. org/ 10. 1016/j. ijdrr. 2015. 05. 009 Deligne NI, Fitzgerald RH, Blake DM et al (2017a) Investigating the consequences of urban volcanism using a scenario approach I: development and application of a hypothetical eruption in the Auck- land Volcanic Field, New Zealand. J Volcanol Geotherm Res. https:// doi. org/ 10. 1016/j. jvolg eores. 2017. 02. 023 Deligne NI, Horspool N, Canessa S et  al (2017b) Evaluating the impacts of volcanic eruptions using RiskScape. J Appl Volcanol. https:// doi. org/ 10. 1186/ s13617- 017- 0069-2 Dominguez L, Bonadonna C, Frischknecht C et  al (2021) Integrative post-event impact assessment framework for volcanic eruptions: a disaster forensic investigation of the 2011–2012 Eruption of the Cordón Caulle Volcano (Chile). Front Earth Sci. https:// doi. org/ 10. 3389/ feart. 2021. 645945 Dunant A, Bebbington M, Davies T (2021a) Probabilistic cascading multi-hazard risk assessment meth- odology using graph theory, a New Zealand trial. Int J Disaster Risk Reduct 54:102018. https:// doi. org/ 10. 1016/j. ijdrr. 2020. 102018 Dunant A, Bebbington M, Davies T, Horton P (2021b) Multihazards scenario generator: a network-based simulation of natural disasters. Risk Anal. https:// doi. org/ 10. 1111/ risa. 13723 Elissondo M, Baumann V, Bonadonna C et al (2016) Chronology and impact of the 2011 Cordón Caulle eruption, Chile. Nat Hazard 16:675–704. https:// doi. org/ 10. 5194/ nhess- 16- 675- 2016 Fearnley CJ, Beaven S (2018) Volcano alert level systems: managing the challenges of effective volcanic crisis communication. Bull Volcanol. https:// doi. org/ 10. 1007/ s00445- 018- 1219-z Few R, Armijos MT, Barclay J (2017) Living with Volcan Tungurahua: the dynamics of vulnerability during prolonged volcanic activity. Geoforum 80:72–81. https:// doi. org/ 10. 1016/j. geofo rum. 2017. 01. 006 Galderisi A, Kropp JP, Ceudech A, Kallache M (2011) From global to local and from local to global: examples of event scenarios in Europe BT-inside risk: a strategy for sustainable risk mitigation. In: Menoni S, Margottini C (Eds). Springer Milan, Milano, pp 245–285 https://doi.org/10.1007/s00445-004-0365-7 https://doi.org/10.1007/s00445-004-0365-7 https://doi.org/10.1007/s00445-019-1294-9 https://doi.org/10.1007/s00445-019-1294-9 https://doi.org/10.1029/2022JB024343 https://doi.org/10.5194/nhess-14-2265-2014 https://doi.org/10.1016/j.jvolgeores.2016.06.006 https://doi.org/10.1186/s13617-017-0066-5 https://doi.org/10.1186/s13617-017-0066-5 https://doi.org/10.1073/pnas.1231332100 https://doi.org/10.1080/13607863.2014.941325 https://doi.org/10.1080/13607863.2014.941325 https://doi.org/10.1080/00288306.2021.1895231 https://doi.org/10.1080/00288306.2021.1895231 https://doi.org/10.1007/s00445-017-1184-y https://doi.org/10.1007/s00445-017-1184-y https://doi.org/10.1016/j.ijdrr.2018.07.026 https://doi.org/10.1016/j.ijdrr.2015.05.009 https://doi.org/10.1016/j.jvolgeores.2017.02.023 https://doi.org/10.1016/j.jvolgeores.2017.02.023 https://doi.org/10.1186/s13617-017-0069-2 https://doi.org/10.3389/feart.2021.645945 https://doi.org/10.1016/j.ijdrr.2020.102018 https://doi.org/10.1016/j.ijdrr.2020.102018 https://doi.org/10.1111/risa.13723 https://doi.org/10.5194/nhess-16-675-2016 https://doi.org/10.1007/s00445-018-1219-z https://doi.org/10.1016/j.geoforum.2017.01.006 https://doi.org/10.1016/j.geoforum.2017.01.006 9356 Natural Hazards (2024) 120:9327–9360 1 3 Global Facility for Disaster Reduction and Recovery (2014) Guide to developing disaster recovery frameworks: world reconstruction conference version. Global Facility for Disaster Reduction and Recovery, International Recovery Platform Hayes JL, Wilson TM, Deligne NI et al (2020) Developing a suite of multi-hazard volcanic eruption sce- narios using an interdisciplinary approach. J Volcanol Geoth Res 392:106763. https:// doi. org/ 10. 1016/j. jvolg eores. 2019. 106763 Haynes K, Barclay J, Pidgeon N (2008) Whose reality counts? Factors affecting the perception of vol- canic risk. J Volcanol Geoth Res 172:259–272. https:// doi. org/ 10. 1016/j. jvolg eores. 2007. 12. 012 He X, Cha EJ (2018) Modeling the damage and recovery of interdependent critical infrastructure sys- tems from natural hazards. Reliab Eng Syst Saf 177:162–175. https:// doi. org/ 10. 1016/j. ress. 2018. 04. 029 Hempel L, Kraff BD, Pelzer R (2018) Dynamic interdependencies: problematising criticality assessment in the light of cascading effects. Int J Disaster Risk Reduct 30:257–268. https:// doi. org/ 10. 1016/j. ijdrr. 2018. 04. 011 Hicks A, Barclay J (2018) Citizen-centric risk communication. Geoscientist. https:// doi. org/ 10. 1093/ actra de/ 97801 99576 203. 003. 0007 Hicks A, Few R (2015) Trajectories of social vulnerability during the Soufrière Hills volcanic crisis. J Appl Volcanol. https:// doi. org/ 10. 1186/ s13617- 015- 0029-7 Hicks A, Barclay J, Simmons P, Loughlin S (2014) An interdisciplinary approach to volcanic risk reduction under conditions of uncertainty: a case study of Tristan da Cunha. Nat Hazard 14:1871–1887. https:// doi. org/ 10. 5194/ nhess- 14- 1871- 2014 Hudson-Doyle EE, McClure J, Johnston DM, Paton D (2014a) Communicating likelihoods and probabilities in forecasts of volcanic eruptions. J Volcanol Geotherm Res 272:1–15. https:// doi. org/ 10. 1016/j. jvolg eores. 2013. 12. 006 Hudson-Doyle EE, McClure J, Paton D, Johnston DM (2014b) Uncertainty and decision making: volcanic crisis scenarios. Int J Disaster Risk Reduct 10:75–101. https:// doi. org/ 10. 1016/j. ijdrr. 2014. 07. 006 Hudson-Doyle EE, Paton D, Johnston DM (2015) Enhancing scientific response in a crisis: evidence-based approaches from emergency management in New Zealand. J Appl Volcanol 4:1. https:// doi. org/ 10. 1186/ s13617- 014- 0020-8 Jenkins SF, Magill CR, McAneney KJ (2007) Multi-stage volcanic events: a statistical investigation. J Vol- canol Geoth Res 161:275–288. https:// doi. org/ 10. 1016/j. jvolg eores. 2006. 12. 005 Jenkins S, Komorowski JC, Baxter PJ et al (2013) The Merapi 2010 eruption: an interdisciplinary impact assessment methodology for studying pyroclastic density current dynamics. J Volcanol Geotherm Res 261:316–329. https:// doi. org/ 10. 1016/j. jvolg eores. 2013. 02. 012 Jenkins SF, Spence RJS, Fonseca JFBD et al (2014a) Volcanic risk assessment: quantifying physical vul- nerability in the built environment. J Volcanol Geotherm Res 276:105–120. https:// doi. org/ 10. 1016/j. jvolg eores. 2014. 03. 002 Jenkins SF, Phillips JC, Price R et al (2015) Developing building-damage scales for lahars: application to Merapi volcano, Indonesia. Bull Volcanol. https:// doi. org/ 10. 1007/ s00445- 015- 0961-8 Jenkins SF, Wilson TM, Magill CR, et al (2014b) Volcanic ash fall hazard and risk: technical background paper for the UN-ISDR 2015 global assessment report on disaster risk reduction. United Nations Office for Disaster Risk Reduction Juniper Z (2018) Risk assessment of the petroleum exploration and production industry from volcanic haz- ards in the Taranaki region, New Zealand Kappes MS, Keiler M, von Elverfeldt K, Glade T (2012) Challenges of analyzing multi-hazard risk: a review. Nat Hazards 64:1925–1958. https:// doi. org/ 10. 1007/ s11069- 012- 0294-2 Kc B, Shepherd JM, King AW, Johnson Gaither C (2021) Multi-hazard climate risk projections for the United States. Nat Hazards 105:1963–1976. https:// doi. org/ 10. 1007/ s11069- 020- 04385-y Künzler M, Huggel C, Ramírez JM (2012) A risk analysis for floods and lahars: case study in the cordillera central of Colombia. Nat Hazards 64:767–796. https:// doi. org/ 10. 1007/ s11069- 012- 0271-9 Lerner GA, Cronin SJ, Bebbington MS, Platz T (2019) The characteristics of a multi-episode volcanic regime: the post-AD 960 Maero eruptive period of Mt. Taranaki (New Zealand). Bull Volcanol. https:// doi. org/ 10. 1007/ s00445- 019- 1327-4 Lindsay J, Marzocchi W, Jolly G et al (2010) Towards real-time eruption forecasting in the Auckland vol- canic field: application of BET_EF during the New Zealand national disaster exercise ‘Ruaumoko.’ Bull Volcanol 72:185–204. https:// doi. org/ 10. 1007/ s00445- 009- 0311-9 Loughlin SC, Vye-Brown C, Sparks RSJ, Brown SK (2015) Global volcanic hazards and risk: Summary background paper for the Global Assessment Report on Disaster Risk Reduction 2015 Mach KJ, Lemos MC, Meadow AM et al (2020) Actionable knowledge and the art of engagement. Curr Opin Environ Sustain 42:30–37. https:// doi. org/ 10. 1016/j. cosust. 2020. 01. 002 https://doi.org/10.1016/j.jvolgeores.2019.106763 https://doi.org/10.1016/j.jvolgeores.2019.106763 https://doi.org/10.1016/j.jvolgeores.2007.12.012 https://doi.org/10.1016/j.ress.2018.04.029 https://doi.org/10.1016/j.ress.2018.04.029 https://doi.org/10.1016/j.ijdrr.2018.04.011 https://doi.org/10.1016/j.ijdrr.2018.04.011 https://doi.org/10.1093/actrade/9780199576203.003.0007 https://doi.org/10.1093/actrade/9780199576203.003.0007 https://doi.org/10.1186/s13617-015-0029-7 https://doi.org/10.5194/nhess-14-1871-2014 https://doi.org/10.5194/nhess-14-1871-2014 https://doi.org/10.1016/j.jvolgeores.2013.12.006 https://doi.org/10.1016/j.jvolgeores.2013.12.006 https://doi.org/10.1016/j.ijdrr.2014.07.006 https://doi.org/10.1186/s13617-014-0020-8 https://doi.org/10.1186/s13617-014-0020-8 https://doi.org/10.1016/j.jvolgeores.2006.12.005 https://doi.org/10.1016/j.jvolgeores.2013.02.012 https://doi.org/10.1016/j.jvolgeores.2014.03.002 https://doi.org/10.1016/j.jvolgeores.2014.03.002 https://doi.org/10.1007/s00445-015-0961-8 https://doi.org/10.1007/s11069-012-0294-2 https://doi.org/10.1007/s11069-020-04385-y https://doi.org/10.1007/s11069-012-0271-9 https://doi.org/10.1007/s00445-019-1327-4 https://doi.org/10.1007/s00445-009-0311-9 https://doi.org/10.1016/j.cosust.2020.01.002 9357Natural Hazards (2024) 120:9327–9360 1 3 Marzocchi W, Bebbington MS (2012) Probabilistic eruption forecasting at short and long time scales. Bull Volcanol 74:1777–1805. https:// doi. org/ 10. 1007/ s00445- 012- 0633-x Marzocchi W, Woo G (2009) Principles of volcanic risk metrics: theory and the case study of Mount Vesu- vius and Campi Flegrei, Italy. J Geophys Res 114:B03213. https:// doi. org/ 10. 1029/ 2008J B0059 08 Marzocchi W, Sandri L, Selva J (2008) BET_EF: a probabilistic tool for long- and short-term eruption fore- casting. Bull Volcanol 70:623–632. https:// doi. org/ 10. 1007/ s00445- 007- 0157-y Marzocchi W, Sandri L, Selva J (2010) BET_VH: A probabilistic tool for long-term volcanic hazard assess- ment. Bull Volcanol 72:705–716. https:// doi. org/ 10. 1007/ s00445- 010- 0357-8 MCDEM (2005) Recovery management: Director’s guidelines for CDEM groups (M. of C. D. and E. Man- agement, Ed.). http:// www. civil defen ce. govt. nz/ assets/ Uploa ds/ publi catio ns/ dgl- 04- 05- recov ery- manag ement. pdf MCDEM (2010) Community Engagement in the CDEM context [BPG 04/10] (M. of C. D. and E. Management,Ed.). http:// www. civil defen ce. govt. nz/ cdem- sector/ cdem- frame work/ guide lines/ commu nity- engag ement- in- the- cdem- conte xt/ MCDEM (2015) CDEM national capability assessment report: December 2015 (M. of C. D. and E. Man- agement, Ed.). http:// www. civil defen ce. govt. nz/ assets/ FINAL- Natio nal- Capab ility- Asses sment- Report- 2015. pdf Mcdonald GW, Cronin SJ, Kim J et  al (2017) Computable general equilibrium modelling of economic impacts from volcanic event scenarios at regional and national scale, Mt. Taranaki, New Zealand. Bull Volcanol 79:1–18 Ministry of Civil Defence & Emergency Management (MCDEM) (2019) National Disaster Resilience Strat- egy Rautaki ā-Motu Manawaroa Aituā National MG-S (2017) Recording Disaster Losses for improving risk modelling Neall VE (1972) Tephrochronology and tephrostratigraphy of western Taranaki (N108–109), New Zealand. N Z J Geol Geophys 15(4):507–557. https:// doi. org/ 10. 1080/ 00288 306. 1972. 10423 983 New Zealand Government (2015). National Civil Defence Emergency Management Plan Order 2015. 2015(April), p 96 New Zealand Lifelines Council (2017) New Zealand Infrastructure Vulnerability Assessment : Stage 1. Sep- tember, pp 1–64 New Zealand Infrastrcture Commission (2022) Rautaki Hanganga o Aotearoa New Zealand Infrastructure Strategy New Zealand Lifelines Council (2020) New Zealand Critical Lifelines Infrastructure National Vulnerability Assessment Orchiston C, Mitchell J, Wilson T et  al (2018) Project AF8: developing a coordinated, multi-agency response plan for a future great Alpine Fault earthquake. NZ J Geol Geophys 61:389–402. https:// doi. org/ 10. 1080/ 00288 306. 2018. 14557 16 Ordaz M, Salgado-Gálvez MA, Huerta B et al (2019) Considering the impacts of simultaneous perils: the challenges of integrating earthquake and tsunamigenic risk. Disaster Prev Manag Int J 28:823–837. https:// doi. org/ 10. 1108/ DPM- 09- 2019- 0295 Pescaroli G, Alexander D (2018) Understanding compound, interconnected, interacting, and cascading risks: a holistic framework. Risk Anal. https:// doi. org/ 10. 1111/ risa. 13128 Phillips J, Barclay PJ, Pyle PD, Armijos MT (2019) Dynamic and extensive risk arising from volcanic ash impacts on agriculture. 1–30 Platz T, Cronin SJ, Cashman KV, Stewart RB, Smith IEM (2007) Transition from effusive to explosive phases in andesite eruptions - A case-study from the AD1655 eruption of Mt. Taranaki, New Zealand. J Volcanol Geoth Res 161(1–2):15–34. https:// doi. org/ 10. 1016/j. jvolg eores. 2006. 11. 005 Platz T, Cronin SJ, Procter JN, Neall VE, Foley SF (2012) Non-explosive, dome-forming eruptions at Mt. Taranaki, New Zealand. Geomorphology 136(1):15–30. https:// doi. org/ 10. 1016/j. geomo rph. 2011. 06. 016 Procter JN, Cronin SJ, Platz T et al (2010) Mapping block-and-ash flow hazards based on Titan 2D sim- ulations: a case study from Mt. Taranaki. NZ Nat Hazards 53:483–501. https:// doi. org/ 10. 1007/ s11069- 009- 9440-x Rinaldi SM, Peerenboom JP, Kelly TK (2001) Identifying, understanding, and analyzing critical infrastruc- ture interdependencies. IEEE Control Syst Mag 21:11–25. https:// doi. org/ 10. 1109/ 37. 969131 Sandri L, Thouret JC, Constantinescu R et al (2014) Long-term multi-hazard assessment for El Misti vol- cano (Peru). Bull Volcanol 76:1–26. https:// doi. org/ 10. 1007/ s00445- 013- 0771-9 Schweizer P, Renn O (2019) Governance of systemic risks for disaster prevention and mitigation. Glob Assess Rep Disaster Risk Reduct 28(6):862–874 https://doi.org/10.1007/s00445-012-0633-x https://doi.org/10.1029/2008JB005908 https://doi.org/10.1007/s00445-007-0157-y https://doi.org/10.1007/s00445-010-0357-8 http://www.civildefence.govt.nz/assets/Uploads/publications/dgl-04-05-recovery-management.pdf http://www.civildefence.govt.nz/assets/Uploads/publications/dgl-04-05-recovery-management.pdf http://www.civildefence.govt.nz/cdem-sector/cdem-framework/guidelines/community-engagement-in-the-cdem-context/ http://www.civildefence.govt.nz/cdem-sector/cdem-framework/guidelines/community-engagement-in-the-cdem-context/ http://www.civildefence.govt.nz/assets/FINAL-National-Capability-Assessment-Report-2015.pdf http://www.civildefence.govt.nz/assets/FINAL-National-Capability-Assessment-Report-2015.pdf https://doi.org/10.1080/00288306.1972.10423983 https://doi.org/10.1080/00288306.2018.1455716 https://doi.org/10.1080/00288306.2018.1455716 https://doi.org/10.1108/DPM-09-2019-0295 https://doi.org/10.1111/risa.13128 https://doi.org/10.1016/j.jvolgeores.2006.11.005 https://doi.org/10.1016/j.geomorph.2011.06.016 https://doi.org/10.1016/j.geomorph.2011.06.016 https://doi.org/10.1007/s11069-009-9440-x https://doi.org/10.1007/s11069-009-9440-x https://doi.org/10.1109/37.969131 https://doi.org/10.1007/s00445-013-0771-9 9358 Natural Hazards (2024) 120:9327–9360 1 3 Setola R, De Porcellinis S, Sforna M (2009) Critical infrastructure dependency assessment using the input- output inoperability model. Int J Crit Infrastruct Prot 2:170–178. https:// doi. org/ 10. 1016/j. ijcip. 2009. 09. 002 Skilodimou HD, Bathrellos GD (2021) Natural and technological hazards in urban areas: assessment, plan- ning and solutions. Sustainability 13(15):8301 Spence R (2004) Risk and regulation: can improved government action reduce the impacts of natural disas- ters? Build Res Inf 32:391–402. https:// doi. org/ 10. 1080/ 09613 21042 00022 1043 Spence RJS, Zuccaro G, Petrazzuoli S, Baxter PJ (2004) Resistance of buildings to pyroclastic flows: ana- lytical and experimental studies and their application to Vesuvius. Nat Hazard Rev 5:48–59. https:// doi. org/ 10. 1061/ (ASCE) 1527- 6988(2004)5: 1(48) Spence RJS, Kelman I, Baxter PJ et al (2005) Residential building and occupant vulnerability to tephra fall. Nat Hazards Earth Syst Sci 5:477–494 Stats NZ (2018) Census Map. https:// datafi nder. stats. govt. nz/ data/ categ ory/ census/ 2018/?_ ga=2. 14334 772. 20517 07801. 16315 70339- 13163 90401. 16224 15496 Stewart C, Johnston DM, Leonard GS et al (2006) Contamination of water supplies by volcanic ashfall: a literature review and simple impact modelling. J Volcanol Geotherm Res 158:296–306. https:// doi. org/ 10. 1016/j. jvolg eores. 2006. 07. 002 Taranaki Civil Defence and Emergency Management Group (2018) Civil Defence Emergency Man- agement Group Plan for Taranaki 2018–2023 Te Mahere Rōpū Tauira a Taranaki Civil Defence Emergency Thacker S, Barr S, Pant R et al (2017a) Geographic hotspots of critical national infrastructure. Risk Anal 37:2490–2505. https:// doi. org/ 10. 1111/ risa. 12840 Thacker S, Pant R, Hall JW (2017b) System-of-systems formulation and disruption analysis for multi- scale critical national infrastructures. Reliab Eng Syst Saf 167:30–41. https:// doi. org/ 10. 1016/j. ress. 2017. 04. 023 Thompson MA, Lindsay JM, Wilson TM et  al (2017a) Quantifying risk to agriculture from volcanic ashfall: a case study from the Bay of Plenty, New Zealand. Nat Hazards 86:31–56. https:// doi. org/ 10. 1007/ s11069- 016- 2672-7 Thompson MA, Owen S, Lindsay JM et al (2017b) Scientist and stakeholder perspectives of transdisci- plinary research: early attitudes, expectations, and tensions. Environ Sci Policy 74:30–39. https:// doi. org/ 10. 1016/j. envsci. 2017. 04. 006 Tierz P, Woodhouse MJ, Phillips JC et al (2017) A framework for probabilistic multi-hazard assessment of rain-triggered lahars using Bayesian belief networks. Front Earth Sci 5:1–23. https:// doi. org/ 10. 3389/ feart. 2017. 00073 Tilloy A, Malamud BD, Winter H, Joly-Laugel A (2019) A review of quantifi