The performance of nine deterministic, one-dimensional, dynamic pesticide l
eaching models with different complexity was evaluated using a field experi
ment with bentazone and ethoprophos on a humic sandy soil with a shallow gr
oundwater table. AU modelers received an extensive description of the exper
imental data. Despite this fact, the interpretation of the experimental dat
a was ambiguous, leading to tremendous user-dependent variability of select
ed model inputs. Together with the fact that most modelers calibrated at le
ast part of their model, the possibility for evaluating model concepts was
limited. In the case of bentazone, most model predictions were within the 9
5% confidence intervals of the observations. In the case of ethoprophos, mo
del performance was often poor due to the ignorance of volatilization, kine
tic sorption and adaptation of the microbial population. Most models were c
alibrated using on-site measured data, Limiting the possibility for extrapo
lation for policy-oriented applications. (C) 2000 Elsevier Science B.V. All
rights reserved.