<p dir="ltr">Evidence for climate change-related impacts and risks is already widespread globally, affecting not only the ecosystems but also the economy and health of our communities. Data-driven predictive modeling approaches such as machine learning and deep learning have emerged to be powerful tools for interpreting large and complex non-linear datasets such as meteorological variables from weather stations or the distribution of infectious droplets produced in a cough. However, the strength of these data-driven models can be further optimized by complementing them with foundational knowledge of the physical processes they represent. By understanding the core physics, one can enhance the reliability and accuracy of predictive outcomes. The effectiveness of these combined approaches becomes particularly feasible and robust with the recent advancements in the High-Performance Computing field. With improved processing speed, algorithm design, and storage capabilities, modern computers allow for a deeper and more precise examination of the data. Such advancements equip us to address the diverse challenges presented by climate change more effectively.</p><p dir="ltr">In particular, this document advances research in mitigating and preventing the consequences of global warming by implementing data-driven predictive models based on statistical, machine learning, and deep learning methods via two phases. In the first phase, this dissertation proposes frameworks consisting of machine and deep learning algorithms to increase the resilience of small-scale renewable energy systems, which are essential for reducing greenhouse gas emissions in the ecosystems. The second phase focuses on using data from physics-based models, i.e., computational fluid dynamics (CFD), in data-driven predictive models for improving the design of air cleaning technologies, which are crucial to reducing the transmission of infectious diseases in indoor environments. </p><p dir="ltr">Specifically, this work is an article-based collection of published (or will be published) research articles. The articles are reformatted to fit the thesis's structure. The contents of the original articles are self-contained. </p>

<p dir="ltr">The urgent global shift towards decarbonization necessitates the development of robust frameworks to navigate the complex technological, financial, and regulatory challenges emerging in the clean energy transition. Furthermore, the increased adoption of renewable energy sources (RES) is correlated to the exponential growth in weather data research over the last few years. This circular relationship, where big data drives renewable growth, which feeds back the data pipeline, serves as the primary focus of this study: the development of computational tools across diverse spatial and temporal scales for the optimal design and operation of renewable energy-based systems. Two scales are considered, differentiated by their primary objectives and techniques used. </p><p dir="ltr"> In the first one, the integration of probabilistic forecasts into the operations of RES microgrids (MGs) is studied in detail. It is revealed that longer scheduling horizons can reduce dispatch costs but at the expense of forecast accuracy due to increased prediction accuracy decay (PAD). To address this, a novel method that determines how to split the time horizon into timeblocks to minimize dispatch costs and maximize forecast accuracy is proposed. This forms the basis of an optimal rolling horizon strategy (ORoHS) which schedules distributed energy resources over varying prediction/execution horizons. Results offer Pareto-optimal fronts, showing the trade-offs between cost and accuracy at varying confidence levels. Solar power proved more cost-effective than wind power due to lower variability, despite wind’s higher energy output. The ORoHS strategy outperformed common scheduling methods. In the case study, it achieved a cost of \$4.68 compared to \$9.89 (greedy policy) and \$9.37 (two-hour RoHS). The second study proposes the Caribbean Energy Corridor (CEC) project, a novel, ambitious initiative that aims to achieve total grid connectivity between the Caribbean islands. The analysis makes use of thorough data procedures and optimization methods for the resource assessment and design tasks needed to build such an infrastructure. Renewable energy potentials are quantified under different temporal and spatial coverages to maximize usage. Prioritizing offshore wind development, the CEC’s could significantly surpass anticipated growth in energy demand, with an estimated installed capacity of 34 GW of clean energy upon completion. The corridor is modeled as an HVDC grid with 32 nodes and 31 links. Underwater transmission is optimized with a Submarine-Cable-Dynamic-Programming (SCDP) algorithm that determines the best routes across the bathymetry of the region. It is found that the levelized cost of electricity remains on the low end at \$0.11/kWh, despite high initial capital investments. Projected savings reach \$ 100 billion when compared with ”business-as-usual” scenarios and the current social cost of carbon. Furthermore, this infrastructure has the potential to create around 50,000 jobs in construction, policy, and research within the coming decades, while simultaneously establishing a robust and sustainable energy-water nexus in the region. Finally, the broader implications of these works are explored, highlighting their potential to address global challenges such as energy accessibility, prosperity in conflict zones, and sharing these discoveries with the upcoming generations.</p>