The increase of the world's population located near areas prone to natural disasters has given rise to new ‘mega risks’; the rebuild after disasters will test the governments’ capabilities to provide appropriate responses to protect the people and businesses. During the aftermath of the Christchurch earthquakes (2010-2012) that destroyed much of the inner city, the government of New Zealand set up a new partnership between the public and private sector to rebuild the city’s infrastructure. The new alliance, called SCIRT, used traditional risk management methods in the many construction projects. And, in hindsight, this was seen as one of the causes for some of the unanticipated problems. This study investigated the risk management practices in the post-disaster recovery to produce a specific risk management model that can be used effectively during future post-disaster situations. The aim was to develop a risk management guideline for more integrated risk management and fill the gap that arises when the traditional risk management framework is used in post-disaster situations. The study used the SCIRT alliance as a case study. The findings of the study are based on time and financial data from 100 rebuild projects, and from surveying and interviewing risk management professionals connected to the infrastructure recovery programme. The study focussed on post-disaster risk management in construction as a whole. It took into consideration the changes that happened to the people, the work and the environment due to the disaster. System thinking, and system dynamics techniques have been used due to the complexity of the recovery and to minimise the effect of unforeseen consequences. Based on an extensive literature review, the following methods were used to produce the model. The analytical hierarchical process and the relative importance index have been used to identify the critical risks inside the recovery project. System theory methods and quantitative graph theory have been used to investigate the dynamics of risks between the different management levels. Qualitative comparative analysis has been used to explore the critical success factors. And finally, causal loop diagrams combined with the grounded theory approach has been used to develop the model itself. The study identified that inexperienced staff, low management competency, poor communication, scope uncertainty, and non-alignment of the timing of strategic decisions with schedule demands, were the key risk factors in recovery projects. Among the critical risk groups, it was found that at a strategic management level, financial risks attracted the highest level of interest, as the client needs to secure funding. At both alliance-management and alliance-execution levels, the safety and environmental risks were given top priority due to a combination of high levels of emotional, reputational and media stresses. Risks arising from a lack of resources combined with the high volume of work and the concern that the cost could go out of control, alongside the aforementioned funding issues encouraged the client to create the recovery alliance model with large reputable construction organisations to lock in the recovery cost, at a time when the scope was still uncertain. This study found that building trust between all parties, clearer communication and a constant interactive flow of information, established a more working environment. Competent and clear allocation of risk management responsibilities, cultural shift, risk prioritisation, and staff training were crucial factors. Finally, the post-disaster risk management (PDRM) model can be described as an integrated risk management model that considers how the changes which happened to the environment, the people and their work, caused them to think differently to ease the complexity of the recovery projects. The model should be used as a guideline for recovery systems, especially after an earthquake, looking in detail at all the attributes and the concepts, which influence the risk management for more effective PDRM. The PDRM model is represented in Causal Loops Diagrams (CLD) in Figure 8.31 and based on 10 principles (Figure 8.32) and 26 concepts (Table 8.1) with its attributes.
Natural hazard disasters often have large area-wide impacts, which can cause adverse stress-related mental health outcomes in exposed populations. As a result, increased treatment-seeking may be observed, which puts a strain on the limited public health care resources particularly in the aftermath of a disaster. It is therefore important for public health care planners to know whom to target, but also where and when to initiate intervention programs that promote emotional wellbeing and prevent the development of mental disorders after catastrophic events. A large body of literature assesses factors that predict and mitigate disaster-related mental disorders at various time periods, but the spatial component has rarely been investigated in disaster mental health research. This thesis uses spatial and spatio-temporal analysis techniques to examine when and where higher and lower than expected mood and anxiety symptom treatments occurred in the severely affected Christchurch urban area (New Zealand) after the 2010/11 Canterbury earthquakes. High-risk groups are identified and a possible relationship between exposure to the earthquakes and their physical impacts and mood and anxiety symptom treatments is assessed. The main research aim is to test the hypothesis that more severely affected Christchurch residents were more likely to show mood and anxiety symptoms when seeking treatment than less affected ones, in essence, testing for a dose-response relationship. The data consisted of mood and anxiety symptom treatment information from the New Zealand Ministry of Health’s administrative databases and demographic information from the National Health Index (NHI) register, when combined built a unique and rich source for identifying publically funded stress-related treatments for mood and anxiety symptoms in almost the whole population of the study area. The Christchurch urban area within the Christchurch City Council (CCC) boundary was the area of interest in which spatial variations in these treatments were assessed. Spatial and spatio-temporal analyses were done by applying retrospective space-time and spatial variation in temporal trends analysis using SaTScan™ software, and Bayesian hierarchical modelling techniques for disease mapping using WinBUGS software. The thesis identified an overall earthquake-exposure effect on mood and anxiety symptom treatments among Christchurch residents in the context of the earthquakes as they experienced stronger increases in the risk of being treated especially shortly after the catastrophic 2011 Christchurch earthquake compared to the rest of New Zealand. High-risk groups included females, elderly, children and those with a pre-existing mental illness with elderly and children especially at-risk in the context of the earthquakes. Looking at the spatio-temporal distribution of mood and anxiety symptom treatments in the Christchurch urban area, a high rates cluster ranging from the severely affected central city to the southeast was found post-disaster. Analysing residential exposure to various earthquake impacts found that living in closer proximity to more affected areas was identified as a risk factor for mood and anxiety symptom treatments, which largely confirms a dose-response relationship between level of affectedness and mood and anxiety symptom treatments. However, little changes in the spatial distribution of mood and anxiety symptom treatments occurred in the Christchurch urban area over time indicating that these results may have been biased by pre-existing spatial disparities. Additionally, the post-disaster mobility activity from severely affected eastern to the generally less affected western and northern parts of the city seemed to have played an important role as the strongest increases in treatment rates occurred in less affected northern areas of the city, whereas the severely affected eastern areas tended to show the lowest increases. An investigation into the different effects of mobility confirmed that within-city movers and temporary relocatees were generally more likely to receive care or treatment for mood or anxiety symptoms, but moving within the city was identified as a protective factor over time. In contrast, moving out of the city from minor, moderately or severely damaged plain areas of the city, which are generally less affluent than Port Hills areas, was identified as a risk factor in the second year post-disaster. Moreover, residents from less damaged plain areas of the city showed a decrease in the likelihood of receiving care or treatment for mood or anxiety symptoms compared to those from undamaged plain areas over time, which also contradicts a possible dose-response relationship. Finally, the effects of the social and physical environment, as well as community resilience on mood and anxiety symptom treatments among long-term stayers from Christchurch communities indicate an exacerbation of pre-existing mood and anxiety symptom treatment disparities in the city, whereas exposure to ‘felt’ earthquake intensities did not show a statistically significant effect. The findings of this thesis highlight the complex relationship between different levels of exposure to a severe natural disaster and adverse mental health outcomes in a severely affected region. It is one of the few studies that have access to area-wide health and impact information, are able to do a pre-disaster / post-disaster comparison and track their sample population to apply spatial and spatio-temporal analysis techniques for exposure assessment. Thus, this thesis enhances knowledge about the spatio-temporal distribution of adverse mental health outcomes in the context of a severe natural disaster and informs public health care planners, not only about high-risk groups, but also where and when to target health interventions. The results indicate that such programs should broadly target residents living in more affected areas as they are likely to face daily hardship by living in a disrupted environment and may have already been the most vulnerable ones before the disaster. Special attention should be focussed on women, elderly, children and people with pre-existing mental illnesses as they are most likely to receive care or treatment for stress-related mental health symptoms. Moreover, permanent relocatees from affected areas and temporarily relocatees shortly after the disaster may need special attention as they face additional stressors due to the relocation that may lead to the development of adverse mental health outcomes needing treatment.
