UCL IRDR 5th Annual Conference, 25th June 2015 Poster Presentations on Calculating, Portraying, Understanding and Reducing Risk Chair: Dr Gordon Ross, UCL IRDR and UCL Statistical Sciences 1. Structural geology and recurrence intervals of active normal faults in the central and northern Italian Apennines. Zoë K. Watson1, Gerald P. Roberts2, Joanna P. Faure Walker1, Peter Sammonds1, 1 UCL IRDR 2 Department of Earth and Planetary Sciences, Birkbeck, University of London Large earthquakes occur in the central Italian Apennines along active normal faults that are exposed at the surface as bedrock faults scarps. These scarps formed during the Holocene, and by undertaking detailed structural mapping (<10m resolution) along them variability in strike, dip, throw and kinematics has been recognised. For one example, the Mt Le Scalette fault, a systematic relationship is seen between the strike, dip and coseismic throw (visible as a pale unweathered stripe). The existence of this relationship highlights that the unweathered stripe must have a tectonic origin and the surface fault scarp is connected to the active seismogenic fault at depth. This has important implications for seismic hazard. The orientation data collected will be used to model the Coulomb stress transfer following large earthquakes within the historical record or instrumental record. It has previously been hypothesised that Coulomb stress transfer plays a role in triggering aftershocks or subsequent earthquakes following a large earthquake. Currently faults are modelled in Coulomb as straight planar faults, this does not accurately reflect the variability of the observed bedrock scarps. From preliminary modelling presented here, it has been shown that by introducing variability of the source fault (earthquake-­‐inducing), the Coulomb stress pattern generated is different to that of a planar fault. The geometry of the receiver faults (other faults nearby) is also important, as it affects the magnitude of the stress transferred. Further modelling of variable normal faults will be undertaken in the future, utilising structural data collected from mapping. This approach is novel, and will aid understanding of stress transfer and seismic hazard (especially following large earthquakes) in the central Apennines. 2. The effect of ice rubble on ice-­‐ice sliding Sally Scourfield1, Peter Sammonds1, Ben Lishman1, Kaj Riska2 1 UCL IRDR 2 TOTAL Ice deformation processes in the Arctic often generate ice rubble, and many situations arise where ice fragments of varying size separate sea ice floes. While the shear forces between sea ice floes in direct contact with each other are controlled by ice-­‐ice friction, what is not known is how the slip of the floes is affected by the presence of rubble between the sliding surfaces. The project aims to describe ice-­‐ice sliding in the presence of ice rubble in terms of a “rate and state” friction law, which considers the rate of movement and the state of thermal consolidation and sliding history. Experiments designed to investigate this were carried out during two field trips in the Arctic, and future laboratory experiments will build on this in a more controlled environment. The next objective will be to incorporate our rate-­‐and-­‐state friction law into a Discrete Element Model that describes ice-­‐ice and ice-­‐structure interactions. Arctic shipping also generates ice rubble. When ships repeatedly transit the same channel in icy waters this channel can fill with broken ice known as brash ice. The sheer volume of accumulations can be problematic in itself, but also if the brash ice is allowed to refreeze, its physical properties can be rather different from the original sea ice. This complicates shipping around ports in the Arctic. This research is expected impact on safe shipping and offshore engineering operations in the Arctic, which is especially relevant given the interest in oil exploration in this region. UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 1 3. Validation of ground motion simulations through spectral-­‐shape and duration proxies for the nonlinear response of engineered systems Alexandra Tsioulou and Carmine Galasso Both at: Dept. of Civil Environmental & Geomatic Engineering and Institute for Risk and Disaster Reduction, UCL Physics-­‐based simulated (or “synthetic”) ground motions (GMs) capturing complex source features (such as spatially variable slip distributions, rise-­‐time, and rupture velocities), path effects (geometric spreading and crustal damping), and site effects (wave propagation through basins and shallow site response) provide a valuable supplement to recorded GMs, fulfilling a variety of engineering needs. Among engineers the general concern is that simulated records may not be equivalent to real records in estimating seismic demand, and hence, in estimating the induced damages to structures. In the present study, we summarize the results of an ongoing study within the Southern California Earthquake Center (SCEC) Technical Activity Group (TAG) focused on Ground Motion Simulation Validation (GMSV). The aim here is to address, on a statistical basis, whether simulated GMs for four historical earthquakes are biased in terms of median spectral-­‐shape-­‐ and duration-­‐related intensity measures (IMs) in comparison with real records. In particular, the study considers the Graves and Pitarka’s (2010) hybrid broadband ground motion simulation methodology for the following historical events: 1979 M 6.5 Imperial Valley, 1989 M 6.8 Loma Prieta, 1992 M 7.2 Landers, and 1994 M 6.7 Northridge. We also compare the intra-­‐event dispersion of recorded and simulated IMs. Hypothesis tests are carried out to assess the statistical significance of the results found for each IM and each event. The results from this study show that the simulation methodology matches well the recorded IMs and the observed differences are generally not statistically significant, providing confidence in using the simulation methodology for engineering application. 4. Modeling fluid-­‐induced seismicity Georgios Michas1, Filippos Vallianatos1,2, Peter Sammonds1 1 UCL IRDR 2 Laboratory of Geophysics and Seismology, Technological Educational Institute of Crete Induced seismicity is frequently observed in the Earth’s crust as response to pore-­‐pressure perturbations caused by fluid diffusion phenomena in the Earth’s crust. In a critically stressed crust such phenomena are known to decrease the effective normal stress along pre-­‐existing cracks and trigger earthquakes. The efficient modeling of fluid-­‐induced seismicity has direct applications in earthquake hazard assessments, in oil and gas explorations, in waste storage and geothermal industry. Characteristic cases can be drawn by fluid-­‐induced earthquake swarm sequences that are typically characterized by strong variations and clustering effects in time and space and can neither be described by a dominant earthquake nor by any simple scaling relation, as the Omori scaling known for aftershock sequences. Here we study such a sequence that occurred on 2001 at the SW part of Corinth rift (central Greece) and its possible relation to pore-­‐
pressure diffusion. We consider earthquakes as a point process in time and space and study their spatiotemporal evolution in terms of Continuous Time Random Walk (CTRW) theory that presents joint probability density functions of inter-­‐event times and jumps between the successive earthquakes. The spatiotemporal analysis of seismicity indicates a correlated anomalous diffusion process that migrates slowly with time. Inter-­‐event times exhibit a broad probability distribution while jump lengths a finite variance, properties that correspond to the subdiffusive regime in terms of CTRW theory. In normal (Brownian) diffusion the mean squared displacement of seismicity, in regard to an initial point in time and space, will be linearly time dependent. In our case we get a power-­‐law time dependency of the mean squared displacement that grows slower than time, indicating the anomalous (subdiffusive) process. In addition, we show that the fractional diffusion equation can be used to model the propagation of the earthquake activity with time. 2 UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 5. Self-­‐protective behaviour during earthquake shaking Gillian Dacey UCL IRDR It is well documented that earthquakes have the power to cause tens of thousands of deaths with many more thousands injured, however the relationship between earthquakes, the environment and the human population is a complex one. Countries at risk from earthquakes issue protective action advice and earthquake preparedness guidance for their citizens. This guidance typically varies in context and details from one country to the next, however the “Drop, Cover, Hold” action has been adopted by many countries as the primary action to take during shaking. But do people actually perform these actions during earthquakes, and what are the outcomes of their actions? By researching the relationship between earthquake action advice, self-­‐protective behaviour and injury patterns, it will be possible to further understand the link between earthquakes, the environment and people. This poster outlines early results from a survey conducted on residents of Christchurch, New Zealand about their actions and experiences during the February 2011 earthquake, and presents an initial interpretation of the data. 6. Is the deep groundwater in Bangladesh at risk of arsenic contamination? Mohammad Shamsudduha1, Anwar Zahid2, William Burgess3 1 UCL IRDR 2 Ground Water Hydrology, Bangladesh Water Development Board, Dhaka, Bangladesh 3 Department of Earth Sciences, UCL Widespread arsenic (As) contamination in shallow groundwater is posing a serious risk to public health and food security in Bangladesh and other Asian Mega-­‐Deltas. Over the past two decades, the use of deep (>150 m below ground level) tubewells has become the most widely used mitigation strategy for the groundwater As crisis in Bangladesh. Also over this time, much has been learned about the deep groundwater of the Bengal Aquifer System through field investigation, experiments and computer modelling. The deep wells are ultimately vulnerable to contamination by As drawn down from its shallow sources, but the magnitude and timing of the As ingress and the security of alternative groundwater pumping strategies are highly uncertain. Vulnerability to increased salinity in groundwater in some coastal areas of Bangladesh is an additional concern particularly due to rising sea levels and frequency of cyclones under global warming. Policy direction and monitoring strategies are urgently needed as a basis for effective management of deep groundwater abstraction. Both requirements may be guided by predictive modelling of alternative groundwater abstraction strategies and scenarios, accounting for the heterogeneous character of the aquifer, the spatial distribution of deep pumping, and the likely scale of increasing demand. Current study applies a numerical groundwater modelling approach to map the security of deep groundwater in southeastern Bangladesh. 7. Permeability in volcanic systems at elevated temperatures and pressures: An Experimental Approach Amy Chadderton1, Peter Sammonds1, Philip Meredith2, Rosanna Smith1, and Hugh Tuffen3 1 UCL IRDR 2 Department of Earth Sciences, UCL 3 Lancaster Environment Centre, Lancaster University Understanding the dynamics of volcanoes and what makes them erupt holds more than purely scientific interest. Being able to characterize and accurately predict the conditions/triggers of eruptive behavior will ultimately save many lives, therefore advancing our understanding of the internal processes of volcanoes holds a very real emotive interest to many people. The permeability of magma exerts a major influence on volcanic activity and we have long held the ability to experimentally determine the permeability of volcanic material via various techniques. These observations have provided the basis for numerous theories of magmatic degassing. Mount St Helens has provided excellent opportunities to observe degassing and has UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 3 given us plenty of sampling opportunities to investigate further. New insights have been afforded by the recent rhyolitic eruptions at Cordón Caulle and Chaitén Volcano in Chile, which both displayed simultaneous explosive and effusive activity, with both lava and ash plumes emitted from the same vents. The permeability of fracture networks that act as fluid flow pathways are the key to understanding such eruptive behaviour, and are being investigated systematically at magmatic temperatures and pressures in the presence of pore fluids, using our newly-­‐developed experimental capability. Over the past year, permeability experiments conducted under these high temperature/pressure conditions have shown that the permeability of lava dome material decreases as temperature increases. The experimental program continues and the permeability of samples will be investigated further with the addition of sample deformation during tests. 8. An interacting natural hazards approach to assessing disaster risk Melanie Duncan1, Mirianna Budimir2 and Joel Gill3 1 British Geological Survey 2 Independent Consultant 3 King’s College London Many populated areas, both rural and urban, are affected by a wide variety of natural hazards. In many cases, however, the spatial distribution of these hazards overlap and interrelationships between different hazards can exist. Here we present qualitative and quantitative evidence of the implications of the following hazard interaction types: (i) one hazard triggering one or more secondary hazard events (e.g., an earthquake triggering a landslide); (ii) a series of triggering relationships forming a cascade/domino event (e.g., an earthquake triggering a landslide, which then triggers flooding); (iii) one hazard changing the probability of an event occurring (e.g., ground subsidence increasing the probability of flooding); and (iv) the spatial/temporal coincidence of natural hazards, resulting in an impact greater than the sum of its component parts. A growing evidence base, including our own research, demonstrates that assessments ignoring relationships between hazards have the potential to underestimate or increase risk. However, while many regions are exposed to multiple or interrelated natural hazards, many assessments, applications and research focuses on single hazards, treating them as being independent. Part of the problem relates to confusion about what a multi-­‐hazard approach could entail, along with a reluctance to fully adopt such an approach, perhaps due to perceived complexities in doing so. Whilst the term multi-­‐hazard is often used, a clear definition and uniform conceptual approach are rarely included. In addition, acquiring sufficiently detailed and separate data for each of the hazards to qualitatively assess interacting multi-­‐hazard risk is still an obstacle; currently the majority of loss data are attributed to the primary hazard event in global hazard databases. A holistic assessment of hazard potential must not only include the identification of all spatially relevant hazards, but also the interactions between them. Such a 'multi-­‐hazard' approach gives a better understanding of the hazard potential faced by communities. 9. Accumulation of Risk in Urban India: Cities provide transformational opportunity for risk reduction Garima Jain and Teja Malladi Indian Institute for Human Settlements, Bangalore, India India is one of the more vulnerable and high-­‐risk countries in the world. Disaster loss is rising every year with grave consequences for the survival, dignity and livelihood of individuals, particularly the poor, and poses a severe setback to hard-­‐won development gains. Growing concentrations of people, built and economic assets in cities is exponentially increasing their propensity to disaster risk. While disasters are known to compound vulnerabilities, with increasing densities in cities and access to services and resources becoming more contested, people’s vulnerability to disasters is growing, making it a vicious cycle. Despite proven correlation between changing climate and increasing frequencies and intensities of hazards, the level of preparedness for such events is still very low. With limited resources available to direct towards planning and resilience building, developing economies like India end up prioritizing 4 UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 rehabilitation and rescue in the face of an event. Lack of data, access to technology and lack of technical and institutional capacities exacerbates this situation further. 76 per cent of the Indian population is exposed to high or medium hazard risk, of which nearly 30% live in one lakh (0.1 million) plus cities. While cities are collectors of risk, they also provide an immense opportunity to reduce this risk, if managed well in time. The number of cities where people and assets are concentrated is still very small—according to the 2011 Census, the number of rural settlements (0.64 million) exceeds that of urban areas (0.008 million) by nearly two orders of magnitude. More risk can be mitigated by directing relevant research and resources to the most vulnerable urban centres. Planned development and growth in second-­‐tier cities would not just safeguard their future, but could also provide for enhanced lives for their citizens. This paper looks at where amongst these cities is risk concentrated in order to prioritise risk reduction practices. It is important to know what aspects are increasing the cities’ propensity to risk—hazard exposure, higher vulnerabilities or lower capacities—so that specific efforts can be made to address these issues. This paper is an attempt to analyse the current status of risk in India – how is risk defined, where is it concentrated and what are the current institutional and policy gaps – which if addressed holistically, can build a more resilient society. Methodology. This paper builds on the fundamental risk equations (Peduzzi, Dao, Herold, & Mouton, 2009; Wisner, Blaikie, Cannon, & Davis, 2003) with an additional attribution to capabilities inherent to cities and their institutions. It uses data for ten hazards, thirteen indicators of vulnerabilities, four for exposure and about twelve indicators for capacities to cope. It uses disaggregated micro-­‐level geographical data collected by various agencies such as ISRO-­‐
Bhuvan and Census at the national level, and global sources such as UNISDR’s GAR 2013 (DesInventar, etc.) to arrive at an understanding of geographical concentration of risk in India. Geographical Information Systems (GIS) are used further to go beyond its city administrative boundaries, helping to identify urban agglomerations exposed to such risks. This paper systematically arrives at a priority set of locations, using hazard sensitive and vulnerability sensitive equations, where risk can be addressed by either providing better early warning systems and planning, vulnerability reduction, exposure management or by building capacities. 10. Insufficient Historical Earthquakes for Probabilistic Seismic Hazard Assessment – Why Active Faults are needed Luke N.J. Wedmore1, Gerald P. Roberts2, Joanna P. Faure Walker1, Ioannis Papanikolaou3, Ken McCaffrey4, Peter Sammonds1, Patience A. Cowie5, Laura C. Gregory6, Zoe K. Watson1, Alessandro Maria Michetti7 1 UCL IRDR 2 Department of Earth and Planetary Sciences, Birkbeck College, University of London 3 Laboratory Mineralogy – Geology, Agricultural University of Athens, Greece 4 Department of Earth Sciences, Durham University 5 University of Bergen, Department of Earth Science, Bergen, Norway 6 School of Earth and Environment, University of Leeds 7 Dipartimento di Scienza e Alta Tecnologia, Università degli Studi dell’Insubria, Como, Italy Historical and instrumental seismic records are primarily used around the world as the key source of data within probabilistic seismic hazard maps. These maps are used by national governments, engineers, the insurance industry and civil protection agencies to calculate, understand and warn people about seismic hazard and risk. However, in areas where earthquake recurrence intervals are very long, it is not clear if instrumental and historical records of earthquake are long enough to accurately describe the seismic hazard of a region. We address this problem by analyzing historical records of earthquake damage in individual villages and towns in the central Apennines over 650 years. This record is almost unique in its length and detail. We compare it to a map of maximum expected damage based on locations of faults that have been active in the last 15,000 years. We analyze the historical record and palaeoseismic records to identify faults that ruptured in the historical earthquakes. We show that there are a large number of faults that have not ruptured during the period of the historical records. Furthermore, there are large areas including population centres that consequently have not suffered the level of shaking possible were a nearby fault to cause an earthquake. Notably many of these areas lie outside the highest hazard identified on the national seismic hazard map of UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 5 Italy. We use this example to illustrate that probabilistic seismic hazard maps should incorporate the location and slip rates of all known active faults and not just those that have experienced earthquakes in the recent past. Current research by the IRDR and collaborators aims to improve our understanding of the geography of seismic hazard through mapping and measuring slip-­‐rates of the active faults and in central Italy. 11. The need for detailed fault measurements across active faults for seismic hazard Joanna Faure Walker UCL IRDR Earthquakes occur across active faults in response to a build up of strain across them. In order to improve our understanding of earthquake hazard and move towards being able to forecast which faults may be close to rupture, we need to know how much strain has accumulated across individual faults as well as gain a better understanding how, why and when this strain is released. To achieve the former, in areas where there is sufficient fault exposure, we need to create detailed maps of active fault locations showing their detailed geometry and measure long-­‐
term displacement-­‐rates across the faults. The Italian Apennines provides such a study area as here active faults are exposed at the surface and offset landforms and sediments with known ages allowing deformation rates to be calculated. Previous work has shown that local 3d fault geometries and the offsets across these active faults can be highly variable. This demonstrates that detailed measurements are needed at multiple sites along these faults. Measurements of fault geometry, fault length, fault displacement-­‐rates and slip vectors can be combined in order to calculate long-­‐term strain-­‐rates across faults and hence to measure at what rate the strain is accumulating across them. Detailed measurements are needed because sites with variable geometry may represent sites of higher (or lower) deformation rates and hence seismic hazard; the precise locations of active faults has important implications for seismic risk, particularly if these faults are close to high population centres. 12. Integrating Disaster Risk Reduction concept into chemistry curriculum in secondary high school in Indonesia. A case study in Banda Aceh, Indonesia. Nurmalahayati Nurdin UCL IRDR Indonesia has adapted various natural calamities such as landslides, floods, droughts, fire forests, and coastal erosions as an impact of climate change. In this regard, education can take a pivotal role in educating the young generation to reduce the increasing impact of climate issue in recent decade. This study therefore aims to develop the effective teaching method in integrating the disaster risk reduction concept into chemistry curriculum through understanding climate change impact on ocean system and water quality in the Senior Secondary High School, Banda Aceh, Indonesia. The study is designed to help learners in understanding the context by using a specific unit and module to develop knowledge, skill and attitude toward climate change-­‐disaster related risks. The finding will contribute to the development of chemistry curriculum and dissemination of disaster risk reduction knowledge through school lesson in the Secondary High School level in Indonesia. It can be a model for other schools across Indonesia in developing their curriculum to improve student’ awareness on climate’s impacts. 13. Are earthquakes predictable? A study on magnitude correlations in California. Katerina Stavrianaki, Gordon J. Ross, Peter R. Sammonds UCL IRDR While the clustering of earthquakes in time and space is widely accepted, much more questionable is the existence correlations in earthquake magnitudes. The standard approach generally assumes that magnitudes are independent and therefore in principle unpredictable. Some research studies have shown the existence of correlation between earthquakes and the earthquakes that occur after them, while others conclude that such correlations lay within the range of uncertainty introduced by the incompleteness of the earthquake catalogues. In this study we will search for the existence of magnitude correlations, not between two subsequent earthquakes but between an earthquake and its aftershocks. To distinguish between 6 UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 main shocks and aftershocks the earthquake catalogue of California has been stochastically de-­‐
clustered using the Epidemic Type Aftershock Sequence (ETAS) model. Once the catalogue has been de-­‐clustered, conditional averages for aftershock magnitude can be computed by grouping the magnitudes of aftershocks that follow earthquakes of comparable magnitude. The aim of this study is to observe any trends of dependency between the magnitudes of aftershock earthquakes and the earthquakes that trigger them. 14. Engineering Applications of Earthquake Early Warning Systems Omar Velazquez Ortiz and Carmine Galasso Dept. of Civil Environmental & Geomatic Engineering and IRDR, UCL The development of earthquake early warning (EEW) systems has recently received much investment and research interest worldwide. EEW usually involves the early detection of an incoming earthquake, the estimation in real-­‐time of the ground shaking/damage/loss likely to result, and the issuance of an alert to people in the region where the ground shaking is likely to be hazardous. The last decade has seen the rapid development of methodologies for EEW systems. So far, most of the research in this field has been led by seismologists as the potential benefits of EEW hinge on successfully producing real-­‐time estimation of ground shaking intensity from a given earthquake source. Typical warning times range from a few seconds to tents of second, with an upper limit of about 1 min. Once available, this information can be used in a variety of ways. Identified applications include personal protective measures (e.g., individuals moving to a safe-­‐zone, for example under a table), automated actions (e.g., stopping trains, isolating sensitive and hazardous machinery and chemicals) and situation awareness by large organization that can help prevent cascading failures. The information provided by an EEW system could also be used by earthquake engineers to design structural systems that can respond better to an incoming earthquake. Design approaches for structure-­‐specific EEW have not been extensively investigated so that their potential and effectiveness largely remains open questions. Based on these preliminary remarks, this project will investigate and address the issues faced in the design of engineering applications of EEW, in a performance-­‐based (i.e., loss-­‐based) approach. In particular, the project will address the design of controlled structural systems using the early warning information combined with semi-­‐active devices, in particular semi-­‐active variable stiffness (SAVS) control systems and magnetorheological dampers, i.e., devices using the EEW information to modify, in real-­‐time, the mechanical properties of strategic structures (e.g., bridges) to better withstand the expected shaking. 15. Enabling Disaster Risk Reduction (DRR) Knowledge by Working Together (?): Insights from non-­‐European Contexts Funda Atun1, Scira Menoni1, Patrick Pigeon2, Fernando Briones3, María-­‐José Jiménez4, Mariano García-­‐Fernández4 1 Politecnico di Milano, Department of Architecture and Urban Studies (DAStU), Italy 2 Université de Savoie-­‐Mont-­‐Blanc, Chambéry, France 3 Center of Research and Higher Studies in Social Anthropology (CIESAS), Mexico DF, Mexico 4 Institute of Geosciences (CSIC,UCM), Madrid, Spain Despite the increasing accumulation of knowledge on natural hazards and vulnerability assessment in recent decades, human and economic losses due to disasters continue increasing. Projections prepared by several institutions also forecast a further increase in losses due to trends such as an on-­‐going concentration of human activities in risk-­‐prone areas and the projected effects of global warming. At this point, disaster risk reduction (DRR) and adaptation policies gain importance and attract even more the attention of various actors. DRR policies to be effective require the collaboration and co-­‐operation of various actors among them scientists, decision-­‐makers and practitioners from the private sector, the public sector and the civil society. In addition, they call for a better employment of knowledge for decision making and implementation. In practice, although knowledge is available, is often fragmented within and among these groups and not adequately integrated into decisions on DRR and the further implementation of risk reducing measures. UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 7 However, the application of knowledge is not automatic as there are barriers and pathways in the process of employing the previously gathered knowledge. It is a complex and complicated process involving multiple levels and scales, as well as a wide range of stakeholders and competencies dealing with constantly changing risks. This work provides an analysis of main knowledge fragmentation issues within and across the four stakeholder groups identified by the consortium: scientists, public sector, private sector, and civil society and identified a range of barriers and pathways in knowledge production, sharing maintenance and usage. This work includes a mixed survey methodology, including exploratory interviews and questionnaires with 17 experts from Istanbul and focus group interviews conducted in Mexico City with 13 experts coming from Latin America. The purpose is not to provide statistical data, but to enrich the findings achieved at the Know-­‐4-­‐DRR project by collecting insights from non-­‐European cases 16. The impact of disasters on political trust and satisfaction with the government Frederike Albrecht Department of Government and Centre for Natural Disaster Science (CNDS), Uppsala University The impact of extreme events on political attitudes held by citizens is a challenging yet crucial part of understanding disasters and their relevance for the political system. Investigating to what extent and under what conditions political trust and satisfaction with the government – as indicators for the perceived governmental performance – are affected by natural disasters, the author provides evidence that can explain change in political attitudes following disasters. Twelve cases of disasters in Europe, caused by various natural hazards, are examined with a mixed-­‐methods approach. A unique research design was developed for this research project; it allows using cross-­‐sectional data to investigate immediate changes in attitudes and overcomes problems that disaster researchers in similar situations face when suitable panel data is not available. Results indicate that changes in satisfaction with the government are more likely, while political trust tends to remain stable. A crucial explanatory factor for change in political attitudes is media coverage during and after the disasters, which is examined in a qualitative analysis of all cases. When disaster management is framed positively in media, satisfaction with the government increases during and after the disaster. Positive framing of disaster management took place when the government reacted quickly and efficiently. However, there is also evidence that positive framing can occur despite a comparatively low activity of the government. Here, comparisons with other countries that appeared to manage the same disaster event worse were used to improve the perception of performance in the own country. This type of positive framing in media can lead to an equally positive effect on satisfaction with the government as efficient disaster management itself. 17. Challenges in Collaborative Disaster Risk Reduction and Management Jenni Koivisto1 and Helena Hermansson2 1 Centre for Climate and Safety / Centre for Natural Disaster Studies, Karlstad University 2 Department of Government, Uppsala University, Swedish National Defence College / Centre for Natural Disaster Studies The complexity and boundary spanning nature of disasters calls for multi-­‐organisational approaches to disaster management and risk reduction. Yet such collaboration is often problematic owing to problems of trust, coordination, power imbalance, varying interests etc. There is therefore a need to understand actor-­‐behaviour and the benefits and drawbacks of different collaborative structures. This is particularly the case outside of the Global North, where studies of collaborative subsystems and networks are scarce. Drawing on two qualitative case studies the authors highlight two important aspects of disaster management and risk reduction: i) how network structure impacts upon collaborative conditions (case Turkey) and ii) the importance of understanding the policy landscape (case Mozambique). Results from Turkey suggest that assumptions regarding how inter-­‐organisational structure affects collaborative conditions may not hold across administrative systems. Despite existing hurdles to collaboration in Turkey, including centralised network governance and a political-­‐administrative system where collaboration is far from the dominating modus operandi, progress in collaborative disaster management has been made. The results suggest that ‘tailor fitting’ inter-­‐organisational 8 UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 network designs to the political-­‐administrative culture and the wider governance system partially explain this. The study informs a theoretical discussion about how political-­‐
administrative system attributes may influence inter-­‐organisational collaboration in the disaster management domain. The Mozambique case study utilises a stakeholder analysis to empirically define disaster risk reduction (DRR) subsystem boundaries and to identify belief systems and the key actors therein. The results indicate that actors can be divided into two advocacy coalitions, formed around extant approaches to DRR, that seek to shape the country’s DRR regime. The study provides insights into how a stakeholder analysis can be used as a tool to pinpoint conflict issues causing stalemate among actors within the policy process. The study thus contributes to efforts to overcome several known challenges for DRR governance. 18. Understanding and reducing Cascading Disasters: Addressing the vulnerability path Gianluca Pescaroli UCL IRDR The consequences of the 2010 eruption of Eyjafjallajökull and the 2011 Tohoku earthquake persuaded the global community to consider more realistically the problem of "cascading disasters". Since then, the concept has been widely used among scholars and practitioners but its definition and distinguishing features remain vague. In order to explain a chain-­‐sequence of interconnected failures, the word 'cascading' is often associated with the metaphor of toppling dominoes, which may have a bearing on the cause-­‐and-­‐effect relationship that is a feature of most catastrophic events. This poster aims to avoid this grey area: it offers a clear definition of the phenomena and it addresses the specific vulnerability path that emerges. A review of the literature is employed to point out the specific features that differentiate "cascading disasters" and "cascading effects" from other forms and dynamics of disaster. Our conclusions suggest that interdependencies, vulnerability, amplification, secondary disasters and critical infrastructure are important factors that accentuate the spreading of chain effects. A new operational function of cascading will be indicated as starting point for future researches. 19. Collaborative risk management across the London health economy Liz Clark and Chloe Sellwood Both at NHS England (London) and London Local Health Resilience Partnership The London Local Health Resilience Partnership (LHRP) is committed to its strategic objective to facilitate health sector preparedness and planning for emergencies across London. This is supported through a strategic objective to have oversight of key health Emergency Preparedness, Resilience and Response (EPRR) risks across London via a London wide health EPRR risk register. This is underpinned and supplemented by the London LHRP EPRR Risk Management Strategy and Policy. The London LHRP recognises that the proactive management of health EPRR risks can bring benefits and opportunities to the London health economy. Good risk management supports London health EPRR professionals in the delivery of their duties and contributes towards better care for patients. The London LHRP integrated risk management approach for health EPRR risks considers the potential impact of all types of health EPRR risks on processes, activities, stakeholders and health services. The London LHRP Risk Management Strategy and Policy, drafted through a multi-­‐agency forum, provides the strategic direction, guidance and good management practice regarding embedding an integrated risk management approach, ensuring it is central to all health EPRR related activities, detailing clear lines of accountability, responsibilities and arrangements. The identification of health EPRR risks follows a concurrent “top down” and “bottom up” approach ensuring that the London LHRP, Patch LHRPs, Patch EPRR Networks and individual organisations are actively engaged in the process of identifying health EPRR risks. Identification of these risks comes from the National, Regional and Community risk registers; national and regional reports; responses to adverse events and escalation from individual organisations. This poster illustrates the risk management process and the current stage of development of the London LHRP Risk Register. UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 9 20. Fuzzy classification of the disaster risk within the socio-­‐economic setting of the affected area Sara Saremi Brunel Business School The severity of the disaster impact varies based on the socio-­‐economic characteristics of the affected area. The higher is the vulnerability of the population, the higher is the impact of the disaster and therefore the more humanitarian aid is required. The failure to assess this severity, may impose risk to affected population and lead to the loss of lives. The existing severity assessment tools employ various criteria such as intensity, frequency, vulnerability and capability. However the socio-­‐economic characteristics of the affected area, which influence the coping capability of the population is seldom investigated. The present research emphasize on the importance of these factors by bringing evidence from equally intense earthquakes which effects Japan (with the high coping capability) differently from Haiti (with the low coping capability). To that end a Disaster Severity Assessment (DSA) tool is provided which accommodates physical and socio-­‐economic impacts of the disaster on the affected population to diagnose the severity of the disaster. This tool is built upon the fuzzy classification of the previous records of disasters and based on six criteria including impact time, fatality, casualty, relative financial damage, Human Development Index (HDI) and Disaster Risk Index (DRI). This decision support tool may assist decision makers to reduce the risk of failure in disaster response operations by taking into account the socio-­‐economical characteristics of the affected population. 21. Frequency-­‐Magnitude Distribution and Heat Flow in Greece GiorgosPapadakis1, FilipposVallianatos2,1, Peter Sammonds1 1 UCL IRDR 2 Technological Educational Institute of Crete, Laboratory of Geophysics and Seismology, Crete, Greece This study investigates seismicity in Greece and its relation to heat flow, based on the science of complex systems. Non-­‐extensive statistical physics formalism is a generalisation of Boltzmann-­‐
Gibbs statistical physics and has been successfully used for the analysis of a variety of complex systems, where fractality and long-­‐range interactions are important. Consequently, in this study, the frequency-­‐magnitude distribution analysis was performed in a non-­‐extensive statistical physics context, and the non-­‐extensive parameter, qM, which is related to the frequency-­‐
magnitude distribution, was used as an index of the physical state of the studied area. Examination of the spatial distribution of the thermo-­‐statistical parameter qM revealed its accordance with the spatial distribution of seismicity during the period 1976-­‐2009. We propose that high qM values are consistent with high seismicity for focal depths ≤ 40 km. Although it seems that, in the region of the central Aegean Sea and the volcanic arc, there is an agreement between low seismicity (low qM values) and heat flow, the eastern part of the volcanic arc exhibits high seismic activity and consequently presents high qM values. It should be noted, however, that the central Aegean region is characterised by a thin and deformed crust and high heat flow values, whereas continental Greece is characterised by a thick crust and low heat flow. In conclusion, the intention of this study is to examine the spatial distribution of the non-­‐
extensive parameter in relation to heat flow anomalies in Greece and the adjacent areas contributing, thus, to wider understanding of the evolution of seismicity in tectonically active regions. 22. The deformation characteristics and response relationship with the influencing factors of the typical colluvial landslide in the Three Gorges Reservoir Ying CAO1,2, Kunlong YIN1 1. Department of Engineering Faculty, China University of Geosciences, Wuhan, China 2 UCL IRDR Because of the proper environmental conditions, the deformations of the colluvial landslides in the Three Gorges Reservoir are influenced by combined factors, such as geotechnical conditions, precipitation and the fluctuation of the reservoir level, the shape of the cumulative displacement-­‐
time curves of which follows a step function. The loading/unloading response ratio theory has 1
0 UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 been widely used in the kinetic processes research of the colluvial landslides. However, most of the researches selected individual influencing factor in calculating the value of the loading/unloading response ratio, such as rainfall or groundwater level, the research of landslides synthetically influenced by several factors have been seldom involved. To this end, the typical bank landslide-­‐Baijiabao landslide was selected as an example in this paper, the deformation characteristics, mechanism and the influence factors of which were studied based on investigation and monitoring. Then the rainfall and reservoir water level were chosen as the loading/unloading parameters to establish a multi-­‐factor loading/unloading response ratio model. The results showed that, the superimposition of the continuous heavy rainfall and the rapid decline of the reservoir was the main inducing factor of intense deformations of the Baijiabao landslide; The values of the load-­‐unload response ratio fluctuated up or down by 1, this indicated that the Baijiabao landslide was in a constant deformation stage, which was in a generally good agreement with the monitoring data and the actual results of field investigations. Therefore, the improved loading/unloading response ratio considering multiple factors of landslide displacement can well reflect the landslide stability condition and the periodical deformation characteristic. 23. Fire Smart Territory: a wildfire risk management concept based on a social-­‐ecological approach Fantina Tedim1, Vittorio Leone2, Gavriil Xanthopoulos3 1 University of Porto, Geography Department, Porto, Portugal, 2 University of Basilicata, Department of Crop Systems, Potenza, Italy 3 Hellenic Agricultural Organization "Demeter", Institute of Mediterranean Forest Ecosystems, Athens, Greece In Europe, the fire suppression policy (command and control chain approach) even though it has been able, for the time being, to stop the increase in annually burned area, it has not addressed the diversity and complexity of fire causes and the appearance of extreme fire events that result in major disasters. Signs of the necessary change are already going on but they still focus on “trees” (the discrete factors) and not on the “forest” (the entire socio-­‐ecological system), i.e. not following a holistic perspective of problem solving. Human activities in earth’s systems, make wildfires inherently tied to social systems, thus the challenge of sustainable solutions to most fire-­‐related problems is impossible without understanding the links and interdependencies between humans and ecosystems. Coping with a phenomenon which is mainly anthropogenic can be done with and not without or, worst, against people.As an answer to this challenge we propose the concept of Fire Smart Territory (FST). Territory and not landscape is its target. This poster presents FST which reframes the relationship between society and territory with the aim to reduce fire incidence by making of a wise and knowledgeable fire use a broad, allowed and accepted practice, and not by putting fire out of law. FST can be defined as a fire prone territory in which the integration of economic and social activities aimed at risk reduction and conservation on natural values and ecosystem services is accomplished by aware and well trained empowered communities able to decide the objectives and practices for preventing, controlling or utilizing fires. 24. Government-­‐to-­‐Citizens (G2C) communications in the post-­‐disaster reconstruction phase: Emilia Romagna earthquakes case study Serena Tagliacozzo UCL IRDR Communicating effectively with citizens is being taken increasingly seriously by government agencies. Moreover, the number of available communication channels is increasing rapidly. Recent research has shown that Web 2.0 technologies, such as social media, have brought new opportunities for government agencies to communicate with the public in an engaging, collaborative and transparent manner. This observation may be particularly relevant when considering that long-­‐term post-­‐disaster reconstruction (PDR) needs the implementation of strategies for collaborative communication between government agencies involved in the recovery efforts and affected citizens. This poster presents the preliminary results of a study UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 1
1 designed to investigate government-­‐to-­‐citizens (G2C) communications during the recovery phase of the Emilia-­‐Romagna earthquakes of May 2012 in northern Italy. The study used a mix of data collection methods, including field notes, a multiple-­‐choice questionnaire and an analysis of official websites. Results suggest that local and regional government agencies have adopted several channels and strategies to communicate with citizens about the reconstruction process. Important elements include information about reconstruction of housing and infrastructure, and availability of new funds and refunds. This information was communicated mainly through face-­‐
to-­‐face interaction and on official websites. In most cases a mixture of communication channels was employed to convey the messages. Social media platforms have been used infrequently to provide information about the reconstruction process and usually only for informational purposes, rather than to engage citizens in two-­‐way discussions. However, analysis of the broader communications context revealed that the citizens' input and opinions about rebuilding were sought and gathered using other means (for example, public meetings and community projects), thus underrating the potential of social media for these tasks. 25. The importance of effective hazard communication and data sensitivity: hazard mapping of Campi Flegrei Danielle Charlton, Christopher Kilburn, Stephen Edwards, Rosa Sobradelo UCL Hazard Centre & UCL IRDR Campi Flegrei is one of the most populated active calderas on Earth. Situated immediately west of Naples, in southern Italy, it has been in episodic unrest since the late 1950s. The unrest is the first recorded at Campi Flegrei since its last eruption in 1538 and suggests that the caldera may have re-­‐entered a state with an increased probability of eruption. Since the last caldera-­‐forming eruption 15,500 years ago, approximately 60 eruptions have occurred across the caldera floor. These have mostly been explosive magmatic or phreatomagmatic events with VEIs of 3-­‐5. Several hazard maps have been published, focussing on the expected distributions of pyroclastic density currents (PDCs) and tephra fall deposits. Although superficially similar, the maps show significant differences in detail, notably in the potential runout distances of PDCs and the distances and directions of major tephra fall. The civil authorities have thus a choice of hazard maps for designing their emergency response and, hence, the potential exists for a confused response. We show here that the final forms of the hazard maps are sensitive to the starting assumptions of each study. Focusing on the hazards from PDCs, these assumptions include; the preferred location and size of a future eruption, the particular model used to determine heights of column collapse and the maximum potential runout distance of PDCs. The results illustrate the importance of highlighting the starting assumptions when communicating hazard information to the civil authorities, so that more-­‐informed decisions can be made about the choice of hazard map to be used when preparing mitigating responses to volcanic unrest. 26. Gendering risk communication – lessons for emergency managers Kevin Blanchard DRR Dynamics In the period before, during and after a disaster event, the role of gender plays significantly in the understanding of the risks faced by the affected population. Evidence has shown that depending on a person’s gender, a message or communication regarding a predicted risk can be interpreted differently and as such, can impact on the wider understanding of that risk. A frequently cited example of this would be the use of the term ‘storm surge’ during Typhoon Haiyan in the Philippines. Whilst the term ‘storm surge’ is widely understood within Western countries, it was not understood by residents in the path of Haiyan. As such, warnings to those communities were not as widely understood and acted upon as they could have been, leading to an increased risk. In addition, it is easy to see how this lack of nuanced communications can then be exaggerated by gender divisions in a community. Within some societies in developing nations, women and girls will often been required to remain within close proximity to the house and as such, may miss wider communication channels centred around community centres or market places. This further exacerbates the risk to those marginalised or at-­‐risk sectors of the community 1
2 UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 In order to reduce risks typically associated with poor methods of risk communication, I have been arguing that risk communicators and emergency planners need to consider their audiences in a much more nuanced way. My suggested poster would examine this area in more detail and specifically through a gender lens. 27. Many Strong Voices: Dealing with change in the Arctic and on small island developing states (SIDS) Ilan Kelman1, John Crump2, Tiina Kurvits2, Stavros Mavrogenis3 1 UCL IRDR and IGH 2 GRID-­‐Arendal, Norway 3 Panteion University of Athens, Greece The Many Strong Voices (MSV) programme at http://www.manystrongvoices.org brings together vulnerable regions in the Arctic and Small Island Developing States (SIDS) in a collaborative effort to document and respond to the common challenges that climate change presents to the regions. MSV supports a consortium of policy-­‐makers, researchers, and implementers in the regions according to the needs that they have identified. This poster summarises MSV's scientific programme and publications, documenting vulnerability and community-­‐ based adaptation to climate change in the Arctic and SIDS. UCL IRDR 5th Annual Conference Poster Abstracts, 25th June 2015 1
3