Measurement and simulation of atrazine and alachlor leaching into two field soils

Two small farmland plots were treated with atrazine and alachlor at a rate of 1.4 and 1.6 kg ha−1, respectively. Subsurface drains were placed in each plot to allow sampling of the leachate from February 14 to May 26, 1989. The early leaching phase was high for both herbicides. Simulations were also made with a deterministic, physically-based model (HYDRUS). Different approaches were used to estimate the values of both sorption (Kd) and degradation (μ) parameters, namely, conventional soil-transfer functions, best-fitting to measured field data, and mass-balance and time-step procedures. The Kd-values obtained by the best-fitting method were the lowest, those calculated by conventional pedotransfer functions the highest. Measured and simulated leaching data were then compared. Qualitative (graphics) and quantitative (statistics) evaluations indicated that the herbicide leaching predictions depended greatly on the sorption parameter input. Computer simulations using Kd-values derived from the best-fitting and time-step approaches gave the best results in predicting the herbicide leaching under field soil conditions.