ARCHIVED REPORTS_XR0009414 - Laserfiche WebLink

Home Browse Search

T r , I RISKPRO'S SESOILfor Windows User's Guide Pollutant loss via washload is computed by taking the sediment yield from the washload cycle multiplied by the adsorbed pollutant concentration in the surface Iayer While studies have been ' conducted comparing results of sediment yield with field data (Hetrick and Travis, 1988), pollutant loss via washload has not been validated in SESOIL. 3.5.8 Soil Temperature' , I The anginal SESOIL model assumed that soil temperature was equal to the user-supplied air I temperature. The model was modified by Hetrick et al. (1989) to predict soil temperature from air temperature according to the following (Toy et al 1978). r Summer Y= 16:1-15---+---0-856X, Fall: Y = 1578 + I023X, 1 Winter. Y 15 322 + 0 656X, Spring Y = 0 179 + 1 052X, where• Y = the mean monthly soil temperature (°F). ' X = the mean monthly air temperature ff) These regression -equations-are very crude and not depth dependent. However, furthe complexity is not warranted since ,soil temperature is used only in Eq. (7) and does nc significantly affect results. It should be noted that some chemical parameters and processes are dependent on temperature (for example, solubility, Henry's law constant, and rate constants for biodegradation and hydrolysis) No explicit consideration of these effects is included in ^SESOIL,,and the user should adjust the--input values for such parameters-if temperature effects T -are judged_to be.important __ _ -- - __- — ___ - - -_ _ I 1 3.5.9 Pollutant Cycle Evaluation There are several approaches used to evaluate the reliability and usefulness of an environmental ' model, such as verification, calibration, sensitivity analysis, uncertainty analysis, and validation Verification establishes that results from each of the algorithms of the model are correct Calibration is the process bf adlissting-selected_model parameters within an accepted range until the differences between model predictions and field observations are within selected cntena of performance (Doniiigan and Dean, 1985). Sensitivity analysis focuses on the relative impact each parameter or term has on the model output, in order to determine the effect of data quality on output reliability. Uncertainty analysis seeks to quantify the uncertainty in the model output as ' a function of uncertainty in both model input and model operations. Validation also compares measured with predicted results, but includes analysis of the theoretical foundations of the model, focusing on the model's performance in simulating actual behavior of the chemical in the , Page 30 '