Almost every archaeological predictive model every made has used one common dataset, regardless of methodology employed and subject area it was applied to, water. In some cases, water has been used as the only predictive feature in relation to archaeological resources. However, the creation and representation of these datasets, todate, have been very simplistic. While advances in GIS has improved upon the creation of hydro dataset, and made redundant the older method of relying on hardcopy maps, these datasets are still poor proxies for real world conditions.

This paper will explore how more advance hydro systems can improve predictive models. In some cases, these advance datasets can lead to a higher degree of precision in accuracy in predictive models. While, only in the initial testing phase of advanced hydro systems and predictive models this paper will also explore the potential these systems could have. Especially focusing on the potential part they could play in new versions of predictive models that used agent based modelling.