Sullivan

Sullivan Engineering Fumigant Emissions Modeling System (FEMS) Enhances the Realism of Modeling Airborne Exposures to Agricultural Fumigants

Sullivan Environmental developed the Fumigant Emissions Modeling System (FEMS), a Monte Carlo-based software to enhance the realism of modeling airborne exposures to agricultural fumigants.  The modeling system treats the initiation of an application, calculates emission rates over time, and incorporates meteorological parameters as Monte Carlo variables.  This software was developed to directly integrate the EPA regulatory models ISCST3, CALPUFF5, and CALPUFF6 (sub-hourly capability) to support fumigation applications. FEMS was reviewed and accepted by the EPA Science Advisory Panel. The integral function in regard to the development of the software programs for the Fumigant Emissions Model System (FEMS) was to create a user-friendly operating system that uses ISCST3, CALPUFF5, and the sub-hourly CALPUFF6 model to allow modelers to accurately predict near-field impacts for a variety of pesticide applications in agricultural areas nationwide.  The distributional output provides a basis to estimate buffer zones based on the modeled concentration fields relative to regulatory, health-based thresholds.


FEMS was d esigned to model 200-100,000 simulated years of operation per model run in order to properly account for the uncertainty in both the emissions and meteorological input data. Distributions of emissions units were developed using the land method to estimate confidence intervals for natural log-normalized medians. The final emission rates for each simulated year were selected by Monte Carlo sampling from the emissions distributions. For the meteorological data stored for a number of geographical locations representing agricultural areas, uncertainty in wind direction, wind speed, and atmospheric stability were similarly developed based on the emissions distribution procedure and the model options chosen by the user. In each case, a probability density function was parametrically defined based on expert elicitation methods, which are commonly used when direct data to define the probability distribution are scarce or unavailable.


The model user has the option to establish seasonal probabilities of when the application of concern is expected to occur during the calendar year. The application probability function can account for how many times per year and the number of years that are expected between applications. In addition, the model user can account for a fraction of the maximum application rate when modeling reduced application rates per acre. The user has a choice of application type, including common rectangular, as well as circular applications.  This is all detailed in the FEMS Users Guide.   Contact Sullivan Environmental for a copy of the Fumigant Emissions Modeling System.