SELECTION CRITERIA OF XENOBIOTIC LEACHING MODELS IN SOIL

Selection criteria of xenobiotic dissipation models in soil are discus sed. Model potentialities and user requirements are diverse; it is par amount to evaluate each model in view of its application. Three main t opics are evaluated: user/programme interactions, model purposes, and real world/model relationships. 1) A programme (or a paper-written mod el) can be easy or hard to use, require few or many input data, make u se of little or much computer or user time, demand input data from the user or search for them in connected data bases, be user-friendly or not. 2) The purpose of the model can be defined by the spatial and tem poral scales of operations, by the variety and quantity of outputs, by the facilities offered to the user for data presentation (in tabular or graphical formats), by basic processes (elementary sub-models) whic h are simulated, by the equations used, and by the types of algorithms used in the implementation. 3) The relationships with the real world can be studied by means of sensitivity analysis, of error propagation tests (sometimes called uncertainty analysis), of stochastic validatio ns, and of validations with data sets obtained in the field.