Thesis (M.S.)--Oregon State University, 2010. / Printout. Includes bibliographical references (leaves 53-55). Also available on the World Wide Web.

The very concept of sustainable development calls for a holistic understanding of socio-economic and natural systems in order to help achieve greater sustainability. The complexity characterizing such systems, however, makes it likely impossible for quantified approaches of even isolated problems to account for all relevant factors in a single, robust and deterministic representation of reality – an inherent feature which largely motivates the use of statistical models applied to empirical data. On three independent examples with significant socio-economic and environmental importance, the present dissertation illustrates how econometrical models applied to real-life environmental data can be fruitfully deployed to facilitate the identification and motivation of innovative policies to achieve greater sustainability. Specifically, the first chapter explores the extent to which large-scale irrigation affects local climate by inducing cooler temperatures in areas located downwind from irrigated land, an externality with positive economic consequences quantified in terms of improved crop yields and reduced human mortality. The second chapter illustrates the benefits offered by a family of new differencing estimators (as theoretically derived from a generalization of existing techniques found in the literature) on the example of the nonparametric estimation of error variance in streamflow measurements - a step that is critical for the accurate prediction by hydrological models of extreme flood events. The third chapter investigates the joint effect of traffic speed and acceleration on urban air quality in order to help anticipate the consequences of innovative traffic regulation on the concentration of key air pollutants with detrimental consequences on human health and the economy.