The quality of modern life is intrinsically tied to the development and maintenance of infrastructure systems. Modern energy and electricity infrastructure systems have high-reliability requirements, with people expecting power at the flip of a switch. The complex market structure and public-private partnerships at multiple levels in power generation and transmission systems make ensuring high reliability even more difficult. The 21st century brings with it multiple challenges and opportunities within these sectors. A large portion of the infrastructure fleet, like dams and fossil fuel generation plants, is old and needs replacement. Further, the decarbonization of the power sector is poised to result in the inclusion of large amounts of variable renewable energy sources, thereby introducing stochasticity in supply. The research presented in this dissertation seeks to assess and improve energy infrastructure resilience against regional spatiotemporal climate risk in the face of the upcoming decarbonization of the power sector. This dissertation seeks to develop our understanding of climate risk to energy infrastructure systems at a regional level. The analysis will be focused on the identification of organized modes of climate variability that lead to space-time clustering of risk.These investigations are accompanied by specific case studies in the contiguous United States and are applicable to electricity grids and river basins. Overall, I will focus on the ability to simulate and predict extreme climate events which pose reliability and failure risks to energy infrastructure systems. Since such events are rare, I propose methods that establish event excedance probabilities accounting for their underlying uncertainties. In chapter I, I present a novel statistical simulation model that can produce realistic, synthetic realizations of hydroclimatic fields across a large region. This k-nearest neighbor-based space-time simulator can be applied to single or multiple hydroclimatic fields across a large domain. The algorithm facilitates the estimation of the probability of extreme events that are not well represented in relatively short observational records. I apply this algorithm to wind and solar fields across the Texas Interconnection. Many regions plan to integrate more wind and solar generation into the energy grid, increasing power supply variability that can pose risks of under-supply. This simulation tool facilitates the estimation of the probability of regional wind and solar energy “droughts” and hence allows for the estimation of the storage needed to achieve desired supply-side reliability. In chapter II, I present a clustering based variant of the simulator developed in chapter I. I show how the algorithm developed in chapter I is a special case of a general class of algorithms. In Chapter II, I generalize the algorithm by introducing clustering on the neighbor likelihoods, thereby allowing for the identification of sub-regions with different state-space evolution characteristics. This allows for the application of the generalized algorithm to cases with greater heterogeneity, for example, increased temporal resolution. The clustering based k-nearest neighbor space-time simulator was developed to generate synthetic simulations of wind-solar data at an hourly timescale. I present an application of this algorithm to hourly wind-solar data across the Texas Interconnection. The application of this algorithm to estimate the underlying uncertainty and risk faced by power producers in entering short-term power supply contracts is demonstrated. In chapter III, I present a retrospective analysis of inferred energy demand trends across the contiguous United States. Future net zero scenarios generally require replacing all fossil-fuel heating with electric heat, thereby precipitating higher electricity peak loads during winter. Assuming 100% penetration of efficient electric space-heating and cooling, this chapter carries out a spatially explicit trend analysis of temperature-based proxies of electricity demand over the past 70 years. As expected, annual mean heating and cooling demand decreases and increases over most of the contiguous US, respectively. Peak thermal load is generally dominated by heating, showing large inter-annual and decadal variability, thus far not displaying statistically significant decreasing trends. In the south, the peak cooling demand has started to dominate the peak demand, but the possibility of an occasional high peak heating demand can not be discounted. Conversely, in the north, the average thermal loads are declining while the peak thermal loads are not. This points to the need for an improved pre-season forecast of peak winter heating loads. In chapter IV, I present a method for the diagnosis of low-frequency climate variability from multi-site data, which leads to spatiotemporal clustering of flooding risk at a regional level. Disruptions to energy infrastructure systems are often caused due to flooding, and the characterization of climate risk to energy infrastructure due to flooding is explored in this context. The approach is demonstrated using the Ohio River Basin as a case study. I show that the dominant timescales of flood risk within the Ohio River Basin are in the interannual (6-7 years), decadal (11-12 years), and long-term (secular) scales, with different sub-regions responding to different climate forcings. These leading modes are associated with El-Nino Southern Oscillation and secular trends. Further, the secular trend points to an east-to-west shift in flood incidence and changes in the storm track, which are consistent with certain climate change projections. Overall, the results point to the presence of compound climate risk inherent at regional levels, with the low-frequency climate variability translating into periods of increased and decreased flood risk, which all the stakeholders should consider.

Infrastructure systems around the world face immediate crises and smoldering long-term challenges. Consequently, system owners and managers must balance the need to repair and replace the aging and deteriorating systems already in place against the need for transformative investments in deep decarbonization, climate adaptation, and transportation that will enable long-term competitiveness. Complicating these decisions are deep uncertainties, finite resources, and competing objectives. These challenges motivate the integration of “hard” investments in physical infrastructure with “soft” instruments like insurance, land use policy, and ecosystem restoration that can improve service, shrink costs, scale up or down as future needs require, and reduce vulnerability to population loss and economic contraction. A critical advantage of soft instruments is that they enable planners to adjust, expand, or reduce them at regular intervals, unlike hard instruments which are difficult to modify once in place. As a result, soft instruments can be precisely tailored to meet near-term needs and conditions, including projections of the quasi-oscillatory, regime-like climate processes that dominate seasonal to decadal hydro-climate variability, thereby reducing the need to guess the needs and hazards of the distant future. The objective of this dissertation is to demonstrate how potentially predictable modes of structured climate variability can inform the design of soft instruments and the formulation of adaptive infrastructure system plans. Using climate information for sequential adaptation requires developing credible projections of climate variables at relevant time scales. PartI considers the drivers of river floods in large river basins, which is used throughout this dissertation as an example of a high-impact hydroclimate extreme. First, chapter 2 opens by exploring the strengths and limitations of existing methodologies, and by developing a statistical-dynamical causal chain framework within which to consider flood risk on interannual to secular time scales. Next, chapter 3 describes the physical mechanisms responsible for heavy rainfall (90th percentile exceedance)and flooding in the Lower Paraguay River Basin (LPRB), focusing on a November-February(NDJF) 2015-16 flood event that displaced over 170 000 people. This chapter shows that: 1. persistent large-scale conditions over the South American continent during NDJF 2015-16 strengthened the South American Low-Level Jet (SALLJ), bringing warm air and moisture to South East South America (SESA), and steered the jet towards the LPRB, leading to repeated heavy rainfall events and large-scale flooding; 2. while the observed El Niño event contributed to a stronger SALLJ, the Madden-JulienOscillation (MJO) and Atlantic ocean steered the jet over the LPRB; and 3. while numerical sub-seasonal to seasonal (S2S) and seasonal models projected an elevated risk of flooding consistent with the observed El Niño event, they had limited skill at lead times greater than two weeks, suggesting that improved representation of MJO and Atlantic teleconnections could improve regional forecast skill. Finally, chapter 4 shows how mechanistic understanding of the physical causal chain that leads to a particular hazard of interest – in this case heavy rainfall over a large area in the Ohio River Basin (ORB) – can inform future risks. Taking the GFDL coupled model, version 3 (CM3) as a representative general circulation model (GCM), this chapter shows that 1. the GCM simulates too many regional extreme precipitation (REP) events but under-simulates the occurrence of back to back REP days; 2. REP days show consistent large-scale climate anomalies leading up to the event; 3. indices describing these large-scale anomalies are well simulated by the GCM; and 4. a statistical model describing this causal chain and exploiting simulated large-scale in-dices from the GCM can be used to inform the future occurrence of REP days. Even the best climate projections must confront epistemic uncertainties. Part II of this dissertation explores how intrinsically flawed projections should inform sequential adaptation.First, chapter5reviews approaches for planning under uncertainty, considering the role of classical decision theory, optimization, probability, and non probabilistic approaches. Next, chapter 6 considers how different physical mechanisms impart predictability at different timescales and the implications of secular, low-frequency cyclical, and high-frequency cyclical variability for selection between instruments with long and short planning periods. In particular, this chapter builds from three assertions regarding the nature of climate risk: 1. different climate risk mitigation instruments have different project lifespans; 2. climate risk varies on many scales; and 3. the processes which dominate this risk over the planning period depend on the planning period itself. Defining M as the nominal design life of a structural or financial instrument and N as the length of the observational record (a proxy for total informational uncertainty), chapter 7 presents a series of stylized computational experiments to probe the implications of these premises. Key findings are that: 1. quasi-periodic and secular climate signals, with different identifiability and predictability, control future uncertainty and risk; 2. adaptation strategies need to consider how uncertainties in risk projections influence the success of decision pathways; and 3. stylized experiments reveal how bias and variance of climate risk projections influencerisk mitigation over a finite planning period. Chapter 7 elaborates these findings through a didactic case study of levee heightening in the Netherlands. Integrating a conceptual model of low-frequency variability with credible projections of sea level rise, chapter 7 uses dynamic programming to co-optimize hard (levee increase) and soft (insurance) instruments. Key findings are that 1. large but distant and uncertain changes (e.g., sea level rise) do not necessarily motivate immediate investment in structural risk protection; 2. soft adaptation strategies are robust to different model structures and assumptions while hard instruments perform poorly under conditions for which they were not de-signed; and 3. increasing the hypothetical predictability of near-term climate extremes significantly lowers long-term adaptation costs. Finally, part III seeks to unpack the conceptual experiments of parts I and II to inform policy and future research. Chapter 8 describes how constructive narratives about climate change can discourage climate fatalism. Instead, chapter 8 emphasizes that while climate change is and will be a critical stressor of infrastructure systems, individuals, communities, and regions have agency and can mitigate its consequences. Finally, chapter9concludes by discussing the key findings of this dissertation and exploring how future work on decision under uncertainty, technology, and earth systems science can aid the design and management of effective infrastructure services.

The rising temperature of the earth due to climate change has shown to alter the variations of hydro-climate variables, including their intensities, frequencies and durations. Extreme events such as floods are, in particular, susceptible to any disturbances in climate cycles. As such it is important to provide policymakers with sufficient knowledge about the probable impacts of climate change on hydrologic extremes and most importantly on floods, which have the highest impacts on the societies. For this reason analysis of hydro-climate extremes is commonly performed using data at each site (or grid cell), however due to the limited number of extreme events, these analyses are not robust. Current methods, such as the regional frequency analysis, which combine data from different locations are incapable of incorporating the spatial structure of the data as well as other explanatory variables, and do not explicitly, assess the uncertainties. In this thesis the spatial hierarchical Bayesian model is proposed for hydro-climate extreme analyses using data recorded at each site or grid. This method combines limited number of data from different locations, estimates the uncertainties in different stages of the hierarchy, incorporates additional explanatory variables (covariates), and can be used to estimate extreme events at un-gaged sites. The first project develops a spatial hierarchical Bayesian method to model the extreme runoffs over two spatial domains in the Columbia River Basin, U.S. The model is also employed to estimate floods with different return levels within time slices of fifteen years in order to detect possible trends in runoff extremes. Continuing on the extreme analysis, the impact of climate change on runoff extremes is investigated over the whole Pacific Northwest (PNW). This study aims to address the question of how the runoff extremes will change in the future compared to the historical time period, investigate the different behaviors of the regional climate models (RCMs) regarding the runoff extremes, and assess the seasonal variations of runoff extremes. Given the increasing number of climate model simulations the goal of the third project is to provide a multi-model ensemble average of hydro-climate extremes and characterize the inherent uncertainties. Outputs from several regional climate models provided by NARCCAP are considered for the analysis in all seasons. Three combination scenarios are defined and compared for multi-modeling of extreme runoffs. The biases of each scenario are calculated and the scenario with the least bias is selected for projecting seasonal runoff extremes. The aim of the fourth project is to quantify and compare the uncertainties regarding global climate models to the ones from the hydrologic model structures in climate change impact studies. Various methods have been proposed to downscale the coarse resolution General Circulation Model (GCM) climatological variables to the fine scale regional variables; however fewer studies have been focused on the selection of GCM predictors. Additionally, the results obtained from one downscaling technique may not be robust and the uncertainties related to the downscaling scheme are not realized. To address these issues, in the fifth study we employed Independent Component Analysis (ICA) for predictor selection which determines spatially independent GCM variables (as discussed in Appendix A). Cross validation of the independent components is employed to find the predictor combination that describes the regional precipitation over the upper Willamette basin with minimum error. These climate variables along with the observed precipitation are used to calibrate three downscaling models: Multi Linear Regression (MLR), Support Vector Machine (SVM) and Adaptive-Network-Based Fuzzy Inference System (ANFIS).