The prediction model BPS (Behaviour of Pesticides in Soils) is describ
ed in details. The model consists of three databases and four computin
g modules. The databases contain data, necessary for the model, i.e. p
esticides characteristics, soil type characteristics and climate data.
The computing modules include module for estimation of input data (ba
sed on information available in databases), module for correction of i
nput data, central computing module and the module for data processing
and graphical and numerical output. The central computing module is b
ased on the mathematical model CALF. The module for results processing
and graphical and numerical output presents calculated results in a f
orm of data sets or in a graphical form, which is very instructive and
user-friendly. Interpretation of results could be also performed in a
form of specialised maps. This interpretation requires the GIS system
, including digitised form of soil map. The BPS model and the soil GIS
of the Czech Republic are connected by means of the soil database, wh
ich is common for both systems. At present interpretation is possible
in soil map 1:500000. The model was tested under laboratory and field
conditions and examples of results obtained for atrazine are shown. Th
e results indicated, that BPS model, even in first version, could be c
onsidered as a useful tool for prediction of pesticide behaviour in so
il.