Pesticide leaching models are being used to assist in the regulation a
nd management of pesticides by indicating their potential for leaching
to groundwater. Uncertainty in model input data is not, regrettably,
included in most pesticide leaching assessments. In the work described
here, we use logarithmic transformations of the attenuation factor (A
F), a simple process-based index model, to represent uncertainty in a
pesticide leaching assessment. Characterization of a wide range of pes
ticides as 'leachers' or 'non-leachers' for a specific Hawaii hydrogeo
logical setting is facilitated by comparing the log-transformed AF, de
signated AFR, for each chemical with two reference chemicals for which
leaching behavior in Hawaii is known. Defining a mean and uncertainty
interval for the AFR index of each chemical being ranked provides a p
ractical method of incorporating data uncertainty into a regulatory pr
otocol. (C) 1998 Elsevier Science B.V.