THE LAKE DILLON WATERSHED, COLORADO: A WATERSHED-SCALE WATER-QUALITY STUDY

Onsite wastewater systems (OWS) are considered a viable method for treating wastewater, particularly in locations where connection to a centralized wastewater treatment facility is not feasible. In the Lake Dillon watershed located in Summit County, Colorado, there are many OWS in use. Concerns about the loading of nutrients, particularly phosphorus (P), to Lake Dillon from various sources within the watershed have prompted the formation of a large research project, involving several entities, to assess the cumulative effects of OWS at the site- and watershed-scales. This specific study focuses primarily on the transport and fate of phosphorus and the relative contribution of numerous sources of P within the watershed.

Numerical models are powerful tools for improving knowledge of wastewater flow and pollutant transformations. The Soil and Water Assessment Tool (SWAT) was chosen for the project. SWAT utilizes primarily physically based data to simulate chemical loadings to streams. Parameters associated with snowmelt and orographic effects were adjusted to create a rigorous match to observed streamflow by improving the magnitude and timing of streamflow hydrograph peaks along the Blue River, one of three tributaries of Lake Dillon.

To simulate the nutrient contributions from OWS in the watershed, a fertilizer procedure was defined in SWAT in areas of high OWS density because SWAT does not contain explicit algorithms for simulating OWS. Calculations were completed to determine the correct amount and chemical composition of the added fertilizer. In addition, a sensitivity study was conducted to determine the importance of numerous SWAT variables related to P transport and fate. Based on results of the study, several sensitive parameters were adjusted to create a match to limited observed P data along the Blue River.

The modeling effort revealed that SWAT could be customized to simulate hydrology in the Lake Dillon watershed. The sensitivity study indicated that uncertainty of the values of several chemical and hydrologic SWAT input parameters inhibits the assignment of relative contributions of P to specific sources in the watershed, including OWS. Additional measurement of several hydrologic parameters is essential if the distribution of P sources within the watershed is to be accurately simulated by the model.