Statistically Evaluating Water Consumption Historically and Across Multiple Users in Virginia

This study explores key aspects of water usage in Virginia via a broad-scale analysis of multiple water users through thirty years of time-series records from the Virginia Water Use Data System. A full spectrum of users is considered, including water used for energy, industrial, agricultural and municipal applications. The extent of the relationship between the volume of water used and drivers like economic and climatic conditions are not well defined in humid environments like Virginia. Mann-Kendall testing is applied to identify water use trends through time both statewide and at the county level. A panel regression is employed to identify relationships between water use and explanatory variables of climatic and economic conditions, both spatially and temporally. Key trends include that industrial and energy sector water withdrawals per facility are significantly decreasing over time. Water used for agricultural applications was found to increase on warmer than average years and decrease on wetter than average years, indicating the panel regression methodology successfully demonstrated and quantified intuitive trends. Interestingly, municipal and industrial water usage had a statistically significant relationship with the Gini coefficient, a measure of inequality in rainfall distribution, indicating intraseasonal variability may play an important role in water use trends that is not apparent using seasonal averages alone. Overall, this work contributes to the understanding of water use trends at the state level for Virginia, and better characterizes long-term trends and short-term variability in water withdrawal.