Understanding Risk and Behaviours in Individuals

This stream of EPICentre's work focuses on the assessment, management and understanding of risk in both societal and scientific contexts. The work can be divided broadly into two complementary strands: one focusing on risk representation and behavioural response in individuals and communities, and the other on risk quantification.

Our work on risk representation and behavioural response examines how people in seismically vulnerable cultures represent the earthquake threat. We have done extensive work in understanding how the public sees earthquake risk in Seattle, USA, Osaka, Japan and Izmir, Turkey. We have found that a range of emotional factors shape their conceptualisations, as do socio-culturally specific factors. Thus we are well versed in the barriers and facilitators of preparedness, in peoples' mind-sets, in a range of cultures. Recently, we have branched out in a number of ways. We have moved into looking at multi-hazard responses, like earthquakes and fires, and more pertinently, into how to intervene to better prepare people at-risk for these threats. To this end we have devised the fix-it intervention for earthquake and fire preparedness in Seattle and Izmir [12] and have implemented this. Furthermore, we have also branched into examining expert understandings, particularly how seismology itself is conceived by seismologists. What is its aim? What should its focus be? We have found that prediction is regarded as a stigmatised endeavour by many working in seismology; they emphasise the importance of working towards mitigation rather than prediction [3].

To complement the research above, we work on the quantification of risk using modern statistical methods. A key challenge here is the treatment of uncertainty, which arises either from intrinsic unpredictability of a system or from imperfect understanding. Uncertainty assessment, often within a modern Bayesian framework, therefore runs through much of the work carried out by Epicentre staff. Expertise includes the modelling of probability distributions governing the frequency of occurrence of natural disasters such as earthquakes; accounting for the limitations of both data and models when determining the impacts of natural disasters upon the built environment and the resulting economic and societal consequences; and the combination of sometimes conflicting information from multiple sources in a transparent and defensible manner so that the results can support decision-making. Close connections with colleagues in the UCL Statistical Science Department ensure that Epicentre's approaches to uncertainty quantification are state-of-the-art.

Most recently, we have moved into the field of resilient recovery from disasters. From a mix of a psycho-social, WASH (water-and-sanitation-hygiene) and structural engineering perspective, we address how to make communities and structures more resilient in disaster prone areas. We use schools as hubs in relation to which our measures can be applied due to their importance in local people’s lives [4]. We have developed a range of tools for assessing the needs and perspectives of local stakeholders, and have developed successful interventions to improve psycho-social as well as WASH and structural resilience. We have found song to be a particularly powerful resource in developing psycho-social resilience in teenage girls in disaster prone areas.

Epicentre Projects in this field: