Controlled-environment agriculture (CEA) offers notable advantages over traditional field agriculture by improving water and space use efficiency. However, high energy costs and emissions remain significant obstacles to achieving economic viability and environmental sustainability. This research addresses these challenges through multi-scale modeling and techno-economic analysis (TEA) to advance sustainable CEA crop production.We first explore geothermal heating as a sustainable alternative to fossil fuels for greenhouse operations. Despite the potential, uncertainties around economic feasibility persist, driven by high capital costs and regional variability in heat demand and geothermal resources. Our TEA framework evaluates the feasibility of geothermal systems across different regions, using New York State as a case study. Results indicate that geothermal heating can be economically competitive with natural gas under favorable conditions, while also achieving meaningful reductions in carbon emissions (0.7 to 1.1 kg CO2e per kg of lettuce). Next, we investigate energy conservation strategies for plant factories (PFs) using building energy modeling to estimate energy and water use and carbon emissions across diverse U.S. locations. Energy consumption ranges from 6.2 to 12.0 kWh per kg of fresh weight (FW) lettuce, and water use between 2.0 and 9.8 liters per kg FW, depending on design and operational factors. Carbon emissions varied from 0.6 to 3.9 kg CO2e per kg FW lettuce, influenced by the local energy grid's primary power sources. The analysis highlights how design and operational choices, along with the energy grid's composition, significantly influence resource consumption and environmental impact. Finally, we address the sensitivity of CEA system models to crop-specific parameters, such as growth and evapotranspiration rates. A sensor-based monitoring system was developed and tested in experimental trials to capture accurate and frequent growth and water use data for two different lettuce cultivars. The results underscore the necessity of cultivar-specific parameterization to improve the accuracy of crop simulation models, emphasizing the need for precision in modeling approaches to optimize CEA systems. Overall, this research provides critical insights and practical frameworks for enhancing the economic and environmental sustainability of CEA systems, offering pathways to optimize resource use and reduce emissions through targeted technological and modeling advancements.