COMPARISON OF PRZM COMPUTER-MODEL PREDICTIONS WITH FIELD LYSIMETER DATA FOR DICHLORPROP AND BENTAZON LEACHING

The objective of this project was to optimize the agreement between th e Pesticide Root Zone Model (PRZM) simulations for two experiments and then use those input parameters to predict herbicide leaching in a th ird experiment. The actual numbers and the simulated results all perta in to lysimeter data which has previously been described. The agreemen t between hydrology simulation results obtained using PRZM and actual field data was optimized by (1) altering the initial soil-water conten t in the soil profile and (2) adjusting the water holding capacity for each soil horizon by changing the field capacity water content to per manent wilting point water content relationship. Once hydrology parame ters were adjusted for best fit, pesticide degradation rates and adsor ption coefficients were adjusted to maximize agreement between observe d dichlorprop and bentazon leaching. Good agreement was obtained in hy drology simulations and bentazon movement, and poor agreement was obta ined with dichlorprop movement. A ''blind simulation'' was then perfor med on bentazon movement in Lanna clay and Mellby sand soils, which in dicated greater water discharge from the Lanna soil, and minimal benta zon loss (less than 0.015 mg/m(2)) from each soil.