Simulating herbicide volatilization from bare soil affected by atmospheric conditions and limited solubility in water.

A numerical model that simulates pesticide fate was developed to predictthe behavior of triallate after application to a field soil. The model has options that allow water and/ or heat transport and can limit simulated aqueous-phase concentrations to triallate solubility in water. Several methods for describing the volatilization boundary condition were tested to assess the accuracy in predicting the volatilization rate, including an approach that requires no atmospheric information and an approach that couples soil and atmospheric processes. Four scenarios were constructed and simulated, to compare with measured volatilization rates. The peak measured volatilization rate (168 g ha(-1) h(-1)) was most accurately predicted with the scenario that included the most complex model (100 g ha(-1) h(-1)). The simplest model overpredicted the peak rate (251 g ha(-1) h(-1)), and the others underpredicted the peak rate (16-67 g ha(-1) h(-1)). The simulations that limited aqueous solubility provided relatively similar values for the total emissions (21-37% of applied triallate), indicating that simplified models may compare well with measurements (31% of applied). A prospective simulation over a period of 100 days showed that applying triallate to the soil surface would ultimately lead to atmospheric emissions of 80% of the applied material with 6% remaining in soil. Incorporating triallate to a depth of 10 cm would reduce emissions to less than 5% and lead to 41% remaining in soil.