ARCHIVED REPORTS_XR0009414 - Laserfiche WebLink

Home Browse Search

IRISKPRO'S SESOIL for Windows User's Guide environment under study (Note that the term validation has often been broadly used to mean a variety of things, including all five of the techniques mentioned above ) ' A number of calibration, validation, and sensitivity studies have been performed on the SESOIL model The model has been verified by extensive testing using extreme ranges of input data. Studies of the hydrologic and washload cycles have already been discussed above (see Sections ' 3.3 and 3 4). The following discusses the kinds of evaluations that have been performed on the pollutant cycle of the SESOIL model Note that model validation is a continuing process, no model is ever completely validated To assess SESOIL's predictive capabilities for pollutant movement, a pollutant transport and validation study was performed by Arthur D Little, Inc. under contract to EPA (Bonazountas et al , 1982). The application/validation study was conducted on two field sites, one in Kansas and one in Montana. SESOIL results were compared to data for the metals chromium, copper, I nickel, and sodium at the Kansas site and the organics naphthalene and anthracene at the Montana site Results showed reasonable agreement between predictions and measurements, although the concentrations of the metals were consistently underestimated at both sides, the rate I of metal movement at the Kansas site was consistently overestimated, while the concentrations of the organics were overestimated at the Montana site Bonazountas et al. (1982) state that the overestimations for the organics were probably due to the fact that biodegradation was not considered in the simulations Note that this study was done with the original SESOIL model, not the modified model that is described herein Hetrick et al. (1989) compared predictions of the improved version of SESOIL with empincal data from a laboratory study involving six organic chemicals (Melancon et al , 1986) and from three different field studies involving the application of aldicarb to two field plots (Hornsby et I al., 1983, R L Jones, 1986, Jones et al , 1983, 1985) and atrazine to a single-field watershed (Smith et al , 1978) Results for several measures of pollutant transport were compared including the location of chemical peak vs time, the tame-dependent amount of pollutant leached to groundwater, the depth distnbution of the pollutant at vanous times, the mass of the chemical degraded, and the amount of pollutant in surface runoff This study showed that SESOIL predictions were in good agreement with observed data for both the laboratory study and the field studies. SESOIL does a good job of predicting the leading edge of the chemical profile (Hetrick et al , I1989), due mainly to the improvement of the pollutant depth algorithm to include the chemical sorption characteristics (see Section 3 5 2 above) Also, when a split-sample calibration/validation procedure was used on three years of data from the single-field watershed, SESOIL did a good job of predicting the amount of chemical in the runoff The model was less effective in predicting actual concentration profiles; the simulated concentrations near the soil 6 surface underestimated the measurements in most cases One explanation is that SESOIL does r IPage 31 I