Tracking Energy and Resource Consumption for Sustainable and Resilient Development

The main objective of this dissertation is to analyze trends in energy and resource consumption to measure sustainable and resilient development with specific applications to world countries, US states power grid, and US urban public transportation systems. Using human (HC), natural (NC), and produced (PC) capital from Inclusive Wealth as representatives of the triple bottom line of sustainability and inspired by network science, I first introduce a Network-based Frequency Analysis (NFA) method to track sustainable development in world countries from 1990 to 2014. The method compares every country with every other and links them when values are close. The country with the most links becomes the main trend, and the performance of every other country is assessed based on its ‘orbital’ distance from the main trend. Orbital speeds are then calculated to evaluate country-specific dynamic trends. Overall, I find an optimistic trend for HC only, indicating positive impacts of global initiatives aiming towards socio-economic development. However, I also find that the relative performance of most countries has not changed significantly in this period, regardless of their gradual development. Furthermore, I develop a technique to cluster countries and project the results to 2050 and find a significant decrease in NC for nearly all countries, suggesting an alarming depletion of natural resources worldwide. Then, Fisher information (FI), originally developed by Ronald Fisher as a means of measuring the amount of information about an unknown parameter that is present in any observable data, was used to assess stability in multivariate systems. An open source Python script to calculate FI was developed and subsequently applied to analyse stability in the performance of Public transportation systems (PTS) in the 372 US urbanized areas (UZA) reported by the National Transit Database. Finally, by looking at the 50 US states, Shannon entropy was adapted to measure how state-wide electrical energy mixes have evolved in every US states, and it was then used to assess how robust current energy mixes are to any disruption. I notably observe changes for 26 states between the years 1968 and 1980. From simulating several types of disruptions, I then detect three different classes of states: vulnerable (10 states), moderately robust (17 states), and robust (23 states). Expectedly, some states are particularly vulnerable as they depend predominantly on a single energy source (e.g., West Virginia with 95% coal usage). In contrast, I find seven states (i.e., South Dakota, Illinois, Vermont, Connecticut, Maine, New York, and New Jersey) that have particularly robust energy mixes, all with fossil fuel shares below 50% in 2015. The main technical contributions of this thesis are the development of the NFA method and the adaptation of Shannon entropy to evaluate robustness of a system, both of which were applied to extract meaningful information to track sustainable and resilient development. Additionally, the development of an open source Python script to compute FI offers another significant contribution of this dissertation. Overall, the methods developed in this dissertation can be applied to any data, irrespective of the scale, to extract meaningful information and gain new insights about a system. The current proliferation of data creates the opportunity to understand our surrounding systems (e.g. urban engineering sytems) in a much better way, which is critical to ensure the sustainability and resiliency of any system. However, a precise interpretation of those data is significantly challenging and the methods presented in this dissertation possess the potential to tackle those challenges, which is fundamental if we, as a society, aspire to become more sustainable and resilient.