Lysimeters are used in European countries as a final tier for assessing the
groundwater contamination potential of pesticides with leaching tendencies
. The effect of the artificial lysimeter bottom boundary condition on soil
water budget components was investigated with the help of the WAVE model in
comparison with data from a lysimeter study after pre-emergence applicatio
n of [C-14]methabenzthiazuron (MBT) conducted over 252 days. Simulated drai
nage fluxes in the degraded loess soil were larger using a free drainage bo
undary condition. Furthermore the effect of measured spatial variability of
the soil water retention characteristic on simulated water fluxes was inve
stigated with a monte-carlo modelling approach. The predicted variation of
leaching volume was considerably larger than observed in five lysimeter rep
licates. The effect of spatial variability of the sorption distribution coe
fficient on soil residue profiles was negligible for the herbicide methaben
zthiazuron (MBT). In contrast variable soil hydraulic properties did affect
the centre of mass of the pesticide residues. Considering the statistical
uncertainty of the degradation rate parameter estimation from laboratory da
ta resulted in huge variability of predicted total MBT residues.