| During critical periods of stream low flows, much of the discharge is often contributed by baseflow. To understand the nature of the variability in natural aquifer discharge, we used analytical models, statistical tools, and numerical methods. For the analytical and statistical methods, we combined the theoretical understanding of groundwater flow in an unconfined aquifer obtained from solutions to the linearized Boussinesq equation with a statistical analysis to evaluate the nature of the variability of natural aquifer discharge, a crucial step in forecasting baseflow. We then developed a monthly baseflow forecast using an existing numerical model developed by the State of Idaho, and demonstrate important features of the forecast procedures by presenting an application to the Eastern Snake Plain Aquifer (ESPA) in Idaho.;Analytical results suggest that the diffusive aquifer time unit (that is, the square of aquifer length divided by aquifer diffusivity, L 2/D) governs the relationship among lag, attenuation, and distance between aquifer stresses and discharge from the aquifer. This ratio is the characteristic time for the dimensionless time known as the Fourier number (Fo = tD/L2). Most recharge events are much shorter than the ESPA aquifer time scale, which is 440 years, resulting in rapid attenuation of the aquifer stress signal with distance from the discharge point.;Statistical analyses, including a priori model selection with Akaike's information criterion (AIC) followed by model-averaging, can be used to forecast the timing and magnitude of ESPA discharge four months ahead of the peak water demand, which occurs annually in July.;A monthly aquifer discharge forecast can be developed in January, six months ahead of peak water use on the ESPA. The forecast is made using only three components, including: 1) water supply projected effects on irrigation recharge at two large irrigation entities, 2) aquifer pumping, and 3) the effects of aquifer heads (initial conditions) at the time the forecast is generated. One of the products of the numerical forecasting effort is a spreadsheet tool that generates monthly projections of aquifer discharge using only these three input parameters. Results suggest that the performance of the tool for future forecasting will be good (Nash Sutcliffe Efficiency of 0.81). |
