In this dissertation, mathematical programming models and statistical analysis tools have been formulated and designed to study the strategic and optimal solutions to allocate the resources and manage the risk for the renewable energy and precision agriculture. The dissertation, which consists of four papers, lies at the interface of optimization, simulation, and statistical analysis, with a focus on decision making under uncertainty for biofuel process design, renewable energy supply chain management and precision agriculture.

Bio-oil gasification which integrates fast pyrolysis and gasification processes is a relative new conversion technology and this integrated biofuel production pathway has been promoted to take advantage of economies of scale and logistic efficiency. The design of the supply chain networks, especially under uncertainties, is one of the most important decisions faced by the biofuel industry. In the first paper, we proposed a two-stage stochastic programming framework for the biofuel supply chain optimization problem considering uncertainties, including biomass supply availability, technology advancement, and biofuel market price. The results show that the stochastic factors have significant impacts on the decision on fast pyrolysis plant locations, especially when there is insufficient biomass. Also, farmers' participation can have a significant impact on the profitability and robustness of this supply chain design.

Another major challenge faced by the cellulosic biofuel industry is that investors are hesitant to take the risk to construct commercial scale production facilities. Techno- economic analysis (TEA) has been widely adopted to overcome this challenge. The optimal facility locations and capacities as well as the logistic flow decisions for biomass supply and biofuel distribution should be incorporated into techno-economic analysis as well. In the second paper, the author aims to provide a new method that integrated the supply chain design into the techno-economic analysis as well by evaluating the economic feasibility of an integrated pathway on biomass pyrolysis and bio-oil gasification. The results indicate that hybrid fast pyrolysis and bio-oil gasification pathway is more suitable for a decentralized supply chain structure while biomass gasification pathway is more suitable for a single centralized facility supply chain structure.

Feeding millions of people throughout the world who face hunger every day is a formidable challenge. Precision agriculture has attracted increasing attention in the community of farmland management. Farmland management involves a sequence of planning and decision-making processes, including seed selection and irrigation schedule. In the third paper, a mixed integer programming optimization model is proposed to provide decision support on seed selection and irrigation water allocation for customized precision farmland management. The results show that significant increase of farmers’ annual profit can be achieved by carefully choosing irrigation schedule and type of seed. The proposed model can also serve as a risk analysis tool for farmers facing seasonal irrigation water limits as well as a quantitative tool to explore the impact of precision agriculture.

The effect of limited water on corn grain yield is significant and management decisions are essential to optimize farmers’ profits, particularly under stochastic environment. The fourth paper takes uncertainties such as crop price, irrigation water availability and precipitation amount into consideration. A multi-stage stochastic programming is formulated to evaluate the effects of structure of decision making process on farmers’ income. The case study results indicate multi-stage stochastic programming is a promising way for farmland management under uncertainties and can increase farmers’ income significantly.

In order to enhance the data utilization and results interpretation, statistical methods such as Monte-Carlo simulation considering parameter interactions, linear regression analysis, and moment matching method for scenario generation are also applied. The overarching goals of this dissertation is to quantify and manage the uncertainties along the modeling process and provide proper mechanisms that lead to optimal decisions. The outcomes of the research have the potential to accelerate the commercialization of second generation of biofuel and lead to sustainable utilization of water resources. The insights derived from the research contributed to the decision making process under uncertainties.