When we consider complex, multi-step processes such as crop growth or the progress of a disease then simple mathematical functions are inadequate to describe them and we generally use some kind of mathematical model. The commonest form in use is one that we call a simulation model although there are other forms of model such as Rule-based, (e.g. Gu et al. AI Appl 10:13-24, 1996), Bayesian (Gold, Plant Disease Epidemiol 4:84-122, 1989) and 'fuzzy' (Burrough, J Soil Sci 40:477-492, 1989), depending upon the application. Models may sometimes be combined into packages that we call decision support systems. This paper will consider simulation modelling and also the combination of complementary models. Mathematical models of the potato crop have been devised, in a range of sophistication, over a long period of years and a quite proper question is: "Where next? What are the developments that are sought or, more importantly, that are needed?" Keywords Decision support system . Potato . Scale . Simulation model . Solanum tuberosum . Sources of error . Weather data "All models are wrong but some are more useful than others".
(Box 1979)
Uses of ModelsSimulation modelling can have several purposes. One rather obvious one is as a means to formalize our knowledge and to connect sets of knowledge to one another. Then they may be used for further scientific study such as the integrative study of crop physiology or to aid in genotype evaluations, developments in systems that are