Characterizing the uncertainty of pesticide leaching in agricultural soils

A Monte-Carlo numerical simulation procedure for making regional assessments of pesticide leaching has been developed. This procedure uses probability density functions for organic matter, field capacity, and wilting point developed from information on approximately 3000 soils. Variations in climatic conditions were incorporated by random selection of yearly rainfall data. The procedure was demonstrated for aldicarb applied to corn grown in Ohio. A total of 2000 parameter sets were evaluated using the unsaturated zone model PRZM. The simulation results indicated that convergence of the 90th, 95th, and 99th percentiles for movement past 0.3, 0.6, 0.9, and 1.5 m was obtained after 500 simulations. The relative uncertainty associated with these percentiles was approximately 50% after 1500 simulations. The downward movement of aldicarb residues was most sensitive to changes in field capacity. These simulations, coupled with additional soil-specific simulations, indicated no significant movement of aldicarb residues beyond 1.8 m for applications to Ohio corn.