CHARACTERIZING AQUIFER DISCHARGE VARIABILITY USING ANALYTICAL AND STATISTICAL TOOLS

During critical periods of river low flows, much of the discharge is contributed by baseflow. We used analytical methods and statistical tools to understand the nature of the variability in aquifer discharge. We combined the theoretical understanding of groundwater flow in an unconfined aquifer obtained from analytical methods with a statistical analysis to evaluate the nature of the variability of aquifer discharge, a crucial step in forecasting baseflow. We used existing and new analytical tools to determine how spatial location and temporal distribution of recharge impact aquifer discharge at a point or along a line across an aquifer by calculating lag time and the amplitude dampening of aquifer recharge and discharge events as they eventually appear as river gains or losses. Predictions obtained from analytical models were compared to results of statistical analyses, which were used to determine which recharge and discharge variables contribute to observed variability in aquifer discharge.

We use the eastern Snake River basin in southern Idaho to facilitate the presentation of the analytical and statistical models. The eastern Snake River aquifer (ESRA) is the sole source of drinking water for over 300,000 people and is the location of more irrigated land than any other area in the United States other than California’s Central Valley. The ESRA has a series of springs through which most discharge occurs, and the USGS has consistently collected discharge data from the spring complex since the 1950s, providing a unique opportunity to evaluate the impacts of various sources or sinks to aquifer discharge variability. Results from the analytical tools indicate recharge sources or sinks to the ESRA occurring in areas further than about 100 km from the spring complex insignificantly contribute to the annual, seasonal, and decadal-scale variation in ESRA discharge. Statistical analyses, including the use of the Akaike Information Criterion to rank models, suggest sources or sinks occurring within 100 km of the spring complex make up the majority of the annual, seasonal, and decadal-scale variation in ESRA discharge.