California State University, Monterey Bay | April 30th, 2010
Summary
The Carmel River supplies fresh water to the residents of the Monterey Peninsula within the water district served by the California American Water company (Cal-Am). The S
The Carmel River supplies fresh water to the residents of the Monterey Peninsula within the water district served by the California American Water company (Cal-Am). The State Water Resources Control Board, in 1995, ordered that Cal-Am reduce their annual diversions from the Carmel River Watershed to within legal right (3,376 acre-feet per year). An alternative water source to the Carmel River has yet to be determined, and as a result, Cal-Am must continue to divert more water than their legal entitlement. The current magnitude of diversions has a negative impact on the spawning and migrating habitat of Steelhead (Oncorhynchus mykiss). Existing software in the Tarsier Environmental Modeling Framework was used to model the spatial distribution of surface water along the Carmel River. The model simulated the flow of water downstream from catchment area to the Pacific Ocean. The river channel was represented by a network data set comprised of links, representing individual reaches of the river, connected by nodes. Prior to my work, systematic error in the model was thought to be partially the result of the model lacking a simulation of the interactions between the surface water and the underlying aquifer. A groundwater sub-model was developed to correct for the systematic error. The groundwater sub-model simulated the movement of water between the river channel and the aquifer. Stock variables representing a shallow and deep aquifer were added to each link of the network data set. Simulated water in the surface water stock of each link percolates to these aquifer stocks until the groundwater reaches aquifer capacity, allowing surface water to continue flowing downstream. The model also allows the lateral flow of groundwater according to Darcy’s Law. Quantitative and qualitative analysis of model output compared to observed data showed an increase in model accuracy. Quantitatively, a Nash- Sutcliff Coefficient was improved from 0.88 to 0.97 with the addition of the groundwater model. Qualitatively, a visualization of the longitudinal profile of the river system showed the simulated aquifer controlling the surface flow. A hypothetical application of this model is presented where reducing the pumping rate from the aquifer allowed the wetted river channel to increase by 2.5 km. Future work on the model should include a reservoir sub-model and accounting for spatial variability in the precipitation throughout the catchment area. With these improvements it will be possible to improve predictions of the spatial distribution of surface water along the Carmel River given hypothetical scenarios such as the rate and spatial distribution of pumping from the Carmel Aquifer. Using these simulations to inform decisions of river management could benefit all stakeholders of the river.