Modeling water and nutrient discharges from coastal plain subwatersheds of Chesapeake Bay

Background/Question/Methods: The interaction between climate change and urban development will affect future nutrient loading to coastal waters.  Such changes are important because many estuaries are already degraded by excess nutrient inputs.  We used a watershed model to explore how climate and land use changes might affect future nutrient inputs to Chesapeake Bay.  We first used multi-objective optimization and multi-site averaging to calibrate and validate the General Watershed Loading Function (GWLF) model for five subwatersheds draining to the Rhode River, a tributary subestuary of the Chesapeake Bay. Values of 16 hydrological and nutrient parameters were estimated in the calibration process. The calibrated model was then applied to predict material discharges from Inner Coastal Plain watersheds on the western side of the Bay and from Central Coastal Plain watersheds on the of eastern side of the Bay. The future climate scenario was based on projections from the Hadley climate model for the 2007-2099 period. The future land cover scenario was based on projections from the SLEUTH model as implemented by the Chesapeake Bay Program. We used the same land cover projection in both prediction periods of 2030 and 2095.

Results/Conclusions: In comparison to multi-objective optimization method at a single site, the multi-site weighted average approach reduced the relative mean absolute error by 3.9-7.6% for monthly stream flow, 1-6.5% for monthly total nitrogen (TN), and 1.1-7.5% for monthly total phosphorus (TP), respectively. The GWLF model predicted a slight decrease in average annual stream flow and slight increases in average annual TN and TP discharges in the year 2030 compared with a baseline of 2000. By 2095, all stream flow and nutrient discharges were predicted to increase. The relative annual increase rate for TN was greater than for TP; and the relative annual rates of increase for both TN and TP were greater than the relative annual rate of stream flow increase.  By 2095, annual discharges of water, TN, and TP were predicted to increase by 18%, 36%, 33%, respectively, compared to the 2000 baseline. In our simulations, the influence of climate change on discharges was much greater than that of land cover change; but only small changes in land cover were predicted for the particular watersheds we modeled (0.6-4.0% increases in urban development in 2030). Our results do indicate that the influence of climate change must be considered in management efforts to reducing nutrient loading to the Chesapeake Bay.