Simulation of Evapotranspiration and Rainfall-runoff for the Stillwater River Watershed in Central Massachusetts
Paula L. Sturdevant-Rees
Preliminary studies were conducted on the application of the TOPMODEL to the 3.1.6 me Stillwater River watershed in the Wachusett reservoir basin. TOPMODEL is a physically based - semi distributed hydrological model, which simulates rainfall runoff at a daily or sub-daily time steps. TOPMODEL, version 95.02 was used for the modeling studies. This version of TOPMODEL assumes an exponential transmissivity function for the soil and a first order hyperbolic recession curve for the baseflow. TOPMODEL estimates runoff as the sum of base flow, inflltration "excess overland flow and saturation excess overland flow. Inflltration excess overland flow is modeled using the Green-Ampt model. TOPMODEL predicts saturated areas based on a topographic index of similarity, represented as In(a/tan~), where 'In' is the natural logarithm, a is the upslope area per unit contour length and tan~ is the slope gradient; 'a' reflects the tendency of water to accumulate at any point in the catchment and 'tanW reflects the tendency of gravitational forces to move that water downwards. Topographic index were calculated using a digital terrain analysis program called GRIDATB and required digital elevation information for the catchment as input data. In this study, TOPMODEL was applied on both daily and hourly time steps; thus requiring daily as well as hourly input data for precipitation, evapotranspiration (ET) and the stream discharge for the calibration of the model. Real time 15-minute precipitation and stream discharge data may be obtained from the United States Geological Society (USGS). However, records of the ET data do not exist for the Stillwater River basin or any other nearby locations. Hargreave's Equation and FAO Penman-Monteith Model were tested for the modeling of the ET data for the Stillwater River basin. Modeled daily ET data were obtained from the Northeast Regional Climatic Center for the study site. For hourly simulation of rainfall runoff, these daily data were distributed to hourly values using a simple daily sinusoidal curve. TOPMODELis physically based, in the sense that its parameters can be obtained from field measurements. However, the developers (Beven, K.J. and Kirkby, M., 1979) of the model notes that the actual parameters may differ from the optimum parameters. In this study, measured/estimated parameters were used for runoff simulations. Attempts were made to calibrate the parameters manually using a trial and error method. Use of automated calibration processes such as Monte Carlo simulation or GLUE methodology are recommended for future studies.