INTRODUCTION: Connections between environmental factors and mental health issues have been postulated in many different countries around the world. Previously undertaken research has shown many possible connections between these fields, especially in relation to air quality and extreme weather events. However, research on this subject is lacking in New Zealand, which is difficult to analyse as an overall nation due to its many micro-climates and regional differences.OBJECTIVES: The aim of this study and subsequent analysis is to explore the associations between environmental factors and poor mental health outcomes in New Zealand by region and predict the number of people with mental health-related illnesses corresponding to the environmental influence.METHODS: Data are collected from various public-available sources, e.g., Stats NZ and Coronial services of New Zealand, which comprised four environmental factors of our interest and two mental health indicators data ranging from 2016 up until 2020. The four environmental factors are air pollution, earthquakes, rainfall and temperature. Two mental health indicators include the number of people seen by District Health Boards (DHBs) for mental health reasons and the statistics on suicide deaths. The initial analysis is carried out on which regions were most affected by the chosen environmental factors. Further analysis using Auto-Regressive Integrated Moving Average(ARIMA) creates a model based on time series of environmental data to generate estimation for the next two years and mental health projected from the ridge regression.RESULTS: In our initial analysis, the environmental data was graphed along with mental health outcomes in regional charts to identify possible associations. Different regions of New Zealand demonstrate quite different relationships between the environmental data and mental health outcomes. The result of later analysis predicts that the suicide rate and DHB mental health visits may increase in Wellington, drop-in Hawke's Bay and slightly increase in Canterbury for the year 2021 and 2022 with different environmental factors considered.CONCLUSION: It is evident that the relationship between environmental and mental health factors is regional and not national due to the many micro-climates that exist around the nation. However, it was observed that not all factors displayed a good relationship between the regions. We conclude that our hypotheses were partially correct, in that increased air pollution was found to correlate to increased mental health-related DHB visits. Rainfall was also highly correlated to some mental health outcomes. Higher levels of rainfall reduced DHB visits and suicide rates in some areas of the country.
Rapid, reliable information on earthquake-affected structures' current damage/health conditions and predicting what would happen to these structures under future seismic events play a vital role in accelerating post-event evaluations, leading to optimized on-time decisions. Such rapid and informative post-event evaluations are crucial for earthquake-prone areas, where each earthquake can potentially trigger a series of significant aftershocks, endangering the community's health and wealth by further damaging the already-affected structures. Such reliable post-earthquake evaluations can provide information to decide whether an affected structure is safe to stay in operation, thus saving many lives. Furthermore, they can lead to more optimal recovery plans, thus saving costs and time. The inherent deficiency of visual-based post-earthquake evaluations and the importance of structural health monitoring (SHM) methods and SHM instrumentation have been highlighted within this thesis, using two earthquake-affected structures in New Zealand: 1) the Canterbury Television (CTV) building, Christchurch; 2) the Bank of New Zealand (BNZ) building, Wellington. For the first time, this thesis verifies the theoretically- and experimentally validated hysteresis loop analysis (HLA) SHM method for the real-world instrumented structure of the BNZ building, which was damaged severely due to three earthquakes. Results indicate the HLA-SHM method can accurately estimate elastic stiffness degradation for this reinforced concrete (RC) pinched structure across the three earthquakes, which remained unseen until after the third seismic event. Furthermore, the HLA results help investigate the pinching effects on the BNZ building's seismic response. This thesis introduces a novel digital clone modelling method based on the robust and accurate SHM results delivered by the HLA method for physical parameters of the monitored structure and basis functions predicting the changes of these physical parameters due to future earthquake excitations. Contrary to artificial intelligence (AI) based predictive methods with black-box designs, the proposed predictive method is entirely mechanics-based with an explicitly-understandable design, making them more trusted and explicable to stakeholders engaging in post-earthquake evaluations, such as building owners and insurance firms. The proposed digital clone modelling framework is validated using the BNZ building and an experimental RC test structure damaged severely due to three successive shake-table excitations. In both structures, structural damage intensifies the pinching effects in hysteresis responses. Results show the basis functions identified from the HLA-SHM results for both structures under Event 1 can online estimate structural damage due to subsequent Events 2-3 from the measured structural responses, making them valuable tool for rapid warning systems. Moreover, the digital twins derived for these two structures under Event 1 can successfully predict structural responses and damage under Events 2-3, which can be integrated with the incremental dynamic analysis (IDA) method to assess structural collapse and its financial risks. Furthermore, it enables multi-step IDA to evaluate earthquake series' impacts on structures. Overall, this thesis develops an efficient method for providing reliable information on earthquake-affected structures' current and future status during or immediately after an earthquake, considerably guaranteeing safety. Significant validation is implemented against both experimental and real data of RC structures, which thus clearly indicate the accurate predictive performance of this HLA-based method.
