Forest is an essential component of a bioeconomy. Forest supplies timber and raw materials for various products, and provides essential ecosystem services (e.g., clean water and air, flood control, and carbon storage). Systems analysis tools such as life cycle assessment (LCA) have been widely used to quantify the potential environmental implications of wood products. However, previous studies have rarely considered the dynamic responses of forest ecosystems to the changing market and climate in the future, given the limited capability of LCA in modeling dynamic systems. The talk will present dynamic, multi-scale systems modeling frameworks to address these research challenges. The framework integrates consequential LCA with engineering simulations, machine learning, systems dynamics, and economic-ecological modeling. This talk will present two case studies for framework demonstration. The first case study investigates the future impacts of large-scale mass timber adoption in the building industry on global forest ecosystems. The results reveal the role of different socioeconomic pathways and changing climate and market conditions in determining the net carbon consequences of mass timber adoption. The second case study quantitatively explores the impacts of ecosystem dynamics, especially temporal changes of soil organic carbon (SOC), on the life cycle environmental footprints of biorefineries co-producing biofuels and biochar. This case study will present to what extent SOC matters to LCA and the roles of various time- and location-dependent factors. Finally, the talk will discuss how the frameworks can support climate-beneficial, resource-efficient, and environmentally friendly forest resource utilization in a changing climate and environment.

Machine learning (ML) is increasingly used in environmental research to process large datasets and decipher complex relationships between system variables. However, due to the lack of familiarity and methodological rigor, inadequate ML studies may lead to spurious conclusions. In this regard, explicitly summarizing important but easily overlooked pitfalls and providing corresponding good practices would be useful for the environmental community to better understand current situations and possible improvements. In this review, we screened more than 2000 peer-reviewed articles on supervised ML from 10 highly impact environmental journals and identified more than 30 key items and provided evident-based data analysis based on selected 148 highly cited ML research articles. We provided a tutorial-like compilation that includes misconceptions of terminologies, needed sample size and feature size, data enrichment and feature selection, randomness assessment, data leakage management, data splitting, method selection and comparison, model optimization and evaluation, and model explainability and causality, to support the fact that there are clear margins to be improved for majority of studies. By pin-pointing the main pitfalls, illustrating good examples, reference model optimization pathways, and reference modeling paradigms, we sincerely hope these materials help current and prospective environmental researchers to adopt more rigorous data preprocessing and model development standards for facilitating more accurate, robust, and practicable model uses in environmental research and applications. In this presentation, we will also briefly illustrate good examples and pitfalls to avoid when using the emerging ML/NLG (natural language generation) tool, ChatGPT, for environmental research.

The pursuit of sustainability has catalyzed broad investment in the research, development, and deployment (RD&D) of innovative technologies, but navigating and securing funding for the expanding landscape of technology development has become increasingly challenging for technology developers and decision-makers. Quantitative sustainable design (QSD)—a methodology that combines process design and sustainability analysis under uncertainty—is a valuable tool in accelerating technology RD&D and assisting decision-making through simulation-based design and analysis tailored to desired applications. This presentation will illustrate the utility of QSD with three distinct examples: (i) characterizing the viability of large-scale, centralized thermochemical systems for treatment of emerging contaminants with simultaneous valorization of wet organic wastes; (ii) prioritizing experimentation for a small-scale hydrogen and methane-producing anaerobic treatment system for energy recovery from high-strength industrial wastewater; and (iii) prioritizing deployment locations for non-sewered systems for safe and affordable sanitation in resource-limited communities. In all examples, we highlight how insight generated from QSD has helped in identifying system performance drivers, and how the discovery of such drivers has contributed to the changes in research agenda and decision-making. By focusing on actions that would have not been taken place without QSD, we hope to show the significance and synergies of parallel experimental and modeling activities, and the implications of this approach in addressing the complex challenges in a rapidly changing world.

The aging wastewater infrastructure and changing climate call for a paradigm transformation from traditional energy-intensive wastewater treatment plants (WWTPs) into more efficient, net energy-positive water resource recovery facilities (WRRFs) with minimized environmental footprints. The recent New York State (NYS) Food Donation and Food Scraps Recycling Law has opened up opportunities to re-envision the role of traditional WWTPs in the food-water-energy nexus by integrating food waste co-digestion with other innovative waste energy and resource recovery technologies, such as biosolids upcycling, effluent thermal energy recovery (ETER), and nutrient recovery. By leveraging the spatial distribution of thousands of food waste producers and hundreds of WWTPs, retrofitting investment and community savings on energy products can be designed to prioritize disadvantaged communities (DACs) in NYS. In this study, we aim to develop a spatially dependent decision tool to strategically facilitate the selection of the optimal wastewater retrofitting design for each unique WWTP in NYS, and the prioritized WWTPs whose retrofits will best address food waste challenges and make beneficial reuse of recovered energy and resources with minimized environmental footprints. With techno-economic analysis (TEA), life cycle assessment (LCA), and spatially specific DAC impact mapping, the economic, environmental, and social benefits of different retrofitting configurations are evaluated, and the trade-offs among these benefits are identified. This study will not only contribute to local municipalities and communities, but also advance the environmental justice (EJ) and food waste reduction goals in NYS.

As traditional water resources dwindle, alternative water augmentation strategies for potable water production—like water reuse or seawater desalination—are becoming more mainstream. However, the decision to pursue one of these alternatives depends on a wide range of disparate factors that make these options challenging to compare. Analytic Hierarchy Process (AHP) provides a framework to study how technical, environmental, and social factors play into the decision for a municipality to pursue water reuse or seawater desalination. Using ASPEN forprocess modeling, the technical factors are be evaluated through a techno-economic analysis (TEA) to determine the levelized cost of water. Using SimaPro for a life cycle assessment,environmental factors are evaluated by translating assoicated greenhouse gas emissions into a social cost of carbon. And, lastly, using historical data on technology adoption, the social factors are evaluated to determine the cost of achieving technological legitimacy. Drawing on a wide range of public databases to determine how these variables vary across different contexts, this work is intended to explore why water reuse—a water production strategy that has a lower cost and environmental footprint—has a significantly lower adoption rate than seawaterdesalination around the world, and how increasing water demands are likley to impact water reuse's implementation into the future.