Integrating multidisciplinary science, modelling and impact data into evolving, syn-event volcanic hazard mapping and communication: A case study from the 2012 Tongariro eruption crisis, New Zealand
Introduction
Natural hazard impact assessment and communication mechanisms are important inter-related factors in improving community resilience or adaptive capacity for Disaster Risk Reduction (DRR) or Disaster Risk Management (DRM) (Paton et al., 2001a, Paton et al., 2001b). Disaster risk reduction and management is core to the New Zealand Civil Defence and Emergency Management Act (CDEM Act, 2002) and is part of the Hyogo Framework for Action 2005–2015's second priority of “Identify, assess and monitor disaster risks and enhance early warning” (United Nations, 2005), which explicitly includes the mapping of natural hazards. Here we use the 2012 Te Maari eruptions as a case study of the use of impact data to support communication mechanisms (with a focus on hazard maps) during a volcanic crisis.
Effective risk management relies on communication between scientists, emergency managers and the public based on clear, consistent and reliable information (Barclay et al., 2008, Haynes et al., 2008a, Haynes et al., 2008b, McGuire et al., 2009). Adequate risk perception is an important step towards the public and emergency managers choosing to take appropriate risk management actions. Perception varies depending on individual exposure to impacts (Johnston et al., 1999) along with the quality and nature of information received. Developing and maintaining trust and a common understanding of hazard and risk are both important as a basis for appropriate risk mitigation decisions (Paton et al., 2008), and are also challenging (De la Cruz-Reyna and Tilling, 2008). If confusion, misunderstandings or lack of trust occur the public will seek alternative, often informal, information instead (Haynes et al., 2008b, Paton et al., 2008). During a crisis there are now rapidly expanding pathways for the public and relevant agency staff to request hazard information, and for emergency managers and scientists to disseminate information. These pathways follow now-ubiquitous internet access and social media use in homes and over mobile devices (Mersham, 2010, Mersham and Theunissen, 2011). The speed of that communication is virtually instantaneous, and with that speed comes an increasing expectation for information to be current, or at least frequently updated, and for requests to be quickly answered. With the expanding variety of communication channels, message consistency has become increasingly important and increasingly challenging.
At the time of the 2012 Te Maari eruptions, the volcanic risk management context in New Zealand had been dominated by a sustained (nearly 20 year) drive to increase volcanic eruption preparedness in New Zealand following the 1995–96 Ruapehu eruptions (Johnston et al., 2000). Substantial advances had been made in volcanic surveillance coverage and capability with the introduction of the GeoNet project (Section 2.1). Knowledge gaps were being addressed through central and regional government investment in research programmes investigating eruptive histories for New Zealand volcanoes; and in better defining and understanding volcanic impacts on critical infrastructure and key economic sectors, such as primary industries (Section 4.4; Wilson et al., 2012). Improvements occurred in both emergency management arrangements and delivery of science information to inform end-user decision making. These involved science integration with emergency management planning and operations (Paton et al., 1998) and, particularly in years before and after a large breakout lahar from Ruapehu's Crater Lake, better integration between responding agencies (Keys and Green, 2008). Hence, once the small scale of the Te Maari eruptions became apparent, beyond the on-going life safety issues at proximal distances, it was still an ideal opportunity for government, communities, emergency managers and scientists to work together operationally to test response arrangements. The event allowed the positioning of resources, and identified the level of additional planning and support necessary to manage future larger or more sustained eruptions.
Hazard and risk maps are widely used for natural hazard planning and communication (e.g. flooding, Handmer and Milne, 1981, Osti et al., 2008; landslides, Pachauri and Pant, 1992; seismic/earthquake, Frankel, 1995; seismic triggering of landslides, Jibson et al., 2000). The comparatively limited volcano hazard mapping literature has been mostly focussed on the method used for defining hazard zones (e.g. Orsi et al., 2004, Saucedo et al., 2005, Alberico et al., 2008, Macías et al., 2008, Alatorre-Ibargüengoitia et al., 2012, MIAVITA, 2012) despite the importance of maps for communication and the difficulties in designing clear maps for that purpose (Haynes et al., 2007, Leonard et al., 2008, Nave et al., 2010). For example MIAVITA (2012; the European Commission ‘MItigate and Assess risk from Volcanic Impact on Terrain and human Activities’ project) propose a standardised method for mapping each volcanic hazard in detail. However, there rarely is a discussion of the role of hazard maps in communication and emergency response. Evacuation zones are often mapped in a volcanic crisis, but there is little pre-event planning of the linkage to volcanic hazard mapping. This remains a disparity in both the literature and in practice during volcanic crises. MIAVITA (2012) discuss evacuation mapping under another section, without linkage to the extensive hazard mapping section. Given that published hazard maps are one of the first and most widely accessed information resources for warnings and emergency response, the linkage to end-users and their varying needs should be paramount.
Three types of hazard map — background, crisis, and ashfall prediction hazard maps — have previously been used within Tongariro National Park (TNP, which includes Tongariro, Ngauruhoe and Ruapehu volcanoes, Fig. 1), providing a basis for their application across northern Tongariro Volcano (hereafter ‘Tongariro’, which includes the Te Maari vent) in 2012. The latest iterations of the Ruapehu background hazard maps were produced from 2003 onwards (GNS Science (compiler), 2004, GNS Science (compiler), 2005, GNS Science (compiler), 2008), and include hazard zones and instructional text. Ruapehu hazard maps are based only on the single Crater Lake vent as active — it is the only vent from which lava flows or pyroclastic deposits have been erupted from since the mid-Holocene. A background hazard map for Tongariro was developed from 2005 based on multiple ‘active’ vents (GNS Science (compiler), 2007a; see Section 5.1).
Background hazard maps are widely posted throughout TNP, at local businesses and within communities to provide information on the noticeable precursors and locations of volcanic hazard, and of the best life-safety actions to take, in the event of sudden onset of unrest or an eruption. They intend to cover the mapped extent of hazards, what those hazards are, and what action to take and when. The maps form a series and all have a consistent layout template. They are produced by GNS Science in conjunction with DOC and with input from university scientists and other emergency managers. These background hazard maps were tested in small Ruapehu eruptions in 2006 and 2007 (Kilgour et al., 2010) and generally they were understood well. A key outcome was recognition in 2007 that there was a need for a distinct, event-specific crisis hazard map to give messages specific to short-term heightened risks particular to the crisis, with instructions to stay away from the highest hazard zones (GNS Science (compiler), 2007b). Finally, ashfall prediction maps were in use during the 1995–96 eruption sequence at Ruapehu, which included the last significant New Zealand ashfalls prior to 2012. They are produced for all New Zealand volcanoes daily and available to use for communication in the event of an eruption or eruption forecast.
Three main categories of hazard zone are generally used to draw hazard and risk maps:
- (1)
Past event zones envelop the spatial distribution of one or many previous hazard events, usually determined geologically (forensically) or by historic records.
- (2)
Deterministic zones take a potential future scenario, and model what would happen for that given vent location, event size, eruption parameters and topography.
- (3)
Probabilistic zones are produced by aggregating a very large number of scenarios, each generated by sampling a reasonable range of each models' input parameters (e.g., vent location, event sizes, eruption parameters and topography); each iteration is effectively one deterministic scenario. Collectively, these probabilistically-generated scenarios effectively span the full expected suite of outcomes that may occur in a future eruption. The stacked scenarios can then be used to calculate probability of a particular hazard impacting a given point; for example, if 10% of the probabilistically-generated models produce lava flow in an area, that area could be interpreted to have a 10% probability of lava flow inundation in a future eruption. The resulting zone may be coloured or shaded by the varying likelihood of the hazard affecting parts of the zone; alternatively, a specific probability of being affected may be selected (e.g., 5% chance of being impacted by lava flows) and a zone contouring the area affected at that or higher probability. The latter approach is used in the calculations described in Jolly et al. (2014a).
In practice, output maps that claim to be probabilistic are often limited to a few representative values for vent location, event size, and/or eruption parameters because of the vast number of scenarios that occur when multiplying the combination of these variables. For example, Alberico et al. (2008) take into account a wide probabilistic distribution of vent locations but use only two representative event scales at each location.
Definitions of hazard, risk and communication can vary widely. In this paper we define a ‘volcanic hazard’ as a volcanic process with the potential to affect humans or their assets, including the probability of its occurrence (Fournier d'Albe, 1979, UNISDR, 2009). We define ‘risk’ as a function of the probability of a hazardous process and its consequences (based on Fournier d'Albe, 1979). This is consistent with the risk management approach of Jolly et al. (2014a). We use the term ‘communications’ to include (1) a documented role within emergency management teams (within ‘incident management systems’), (2) ‘public awareness’ (UNISDR, 2009) and (3) ‘risk communication’ (summaries in Bier, 2001, Fischhoff, 1995).
Section snippets
Volcano monitoring and emergency management in New Zealand
During a volcanic crisis there are a wide range of New Zealand agencies that need to communicate clearly, consistently and promptly with each other. We provide a brief overview of agencies discussed here; the detailed roles of these agencies are explained in Section 2.1 of Jolly et al. (2014a). New Zealand operates under a two-tier government structure: central government (ministries and departments) and local government (local authorities — regional councils and territorial authorities). The
Sequence of events
Tongariro, sited within the UNESCO World Heritage Area of TNP, New Zealand, erupted briefly in 2012 after about 80 years of inactivity (Scott et al., 2014) and following volcanic earthquakes in preceding weeks (Hurst et al., 2014). At 23:52 (local time) on 6 August the Te Maari vent (Fig. 1) produced a series of explosions and a minor volume of tephra (~ 5 × 105 m3, Pardo et al., 2014). The eruption initially produced surges directed to the east then west in extensive contact with the ground for 1.5
Science communication during the 2012 crisis
We describe the main science communication mechanisms used during the 2012 eruptions, and discuss current shortcomings and future improvements. We also summarise the material (which included impact information and hazard maps) communicated amongst scientists, and between scientists and emergency managers; health agencies; lifelines; DOC; aviation; media and the public, all of whom had distinct information needs.
Hazard maps
Maps are an important and widely used tool for hazard communication. Volcanic hazard maps for the volcanoes in TNP have been developed over the last decade and cover three distinct situations:
- (1)
Background — for awareness during periods of no eruption with background volcanic unrest levels,
- (2)
Crisis — for life safety advice during specific eruption crises, and
- (3)
Ashfall — predicting the potential volcanic ashfall area during a crisis.
Here we focus on the evolution and interrelationship of these types of
Discussion and recommendations
The 2012 Te Maari eruptions of Tongariro were small (< VEI 3) discrete events with minor impacts (Section 3) but they generated a huge demand for information, which was taxing on limited communication resources (Section 4). The communication amongst agencies, media and the public, the science response, and overall emergency management response were relatively well managed, but this finding should be considered with caution given the relatively small event size and duration. The eruptions have
Conclusions
While small, the 2012 Te Maari eruptions were high-profile and created complex information demands and communication needs. Impacts outside the national park boundary were minor but are documented here to help define damage and mitigation thresholds for future eruptions. Impact information, including that impacts were minor, was an important component of the science communication. A collaborative and collegial scientific environment was actively encouraged, and facilitated agreement on
Acknowledgements
This work was supported by Ministry of Business, Innovation and Employment (MBIE) core science funding to GNS Science, Massey University and the University of Canterbury via the NHRP; by EQC funding to the GeoNet project; and by DOC funding towards social research within TNP. Natalia Deligne, Mark Constable, Jo Horrocks and Richard Smith provided valuable editing of specific sections, and reviews by Dougal Townsend and Maureen Coomer enhanced the manuscript. Jim Cole and Jan Lindsay are thanked
References (112)
- et al.
Volcanic hazard and risk assessment from pyroclastic flows at Ischia island (southern Italy)
J. Volcanol. Geotherm. Res.
(2008) On the state of the art: risk communication to the public
Reliab. Eng. Syst. Saf.
(2001)- et al.
Using the spatial distribution and lithology of ballistic blocks to interpret eruption sequence and dynamics: August 6 2012 Upper Te Maari eruption, New Zealand
J. Volcanol. Geotherm. Res.
(2014) - et al.
Environmental hazards of fluoride in volcanic ash: a case study from Ruapehu volcano, New Zealand
J. Volcanol. Geotherm. Res.
(2003) - et al.
Volcanic ash leachate compositions and assessment of health and agricultural hazards from 2012 hydrothermal eruptions, Tongariro, New Zealand
J. Volcanol. Geotherm. Res.
(2014) - et al.
Dual polarisation C-band weather radar imagery of the 6 August 2012 Te Maari Eruption, Mount Tongariro, New Zealand
J. Volcanol. Geotherm. Res.
(2014) - et al.
Scientific and public responses to the ongoing volcanic crisis at Popocatépetl Volcano, Mexico: importance of an effective hazards-warning system
J. Volcanol. Geotherm. Res.
(2008) - et al.
Vegetation dieback as a proxy for temperature within a wet pyroclastic density current: A novel experiment and observations from the 6th of August 2012 Tongariro eruption
J. Volcanol. Geotherm. Res.
(2014) - et al.
The application of a calibrated 3D ballistic trajectory model to ballistic hazard assessments at Upper Te Maari, Tongariro
J. Volcanol. Geotherm. Res.
(2014) - et al.
Detecting complex eruption sequence and directionality from high-rate geodetic observations: The August 6, 2012 Te Maari eruption, Tongariro, New Zealand
J. Volcanol. Geotherm. Res.
(2014)