Digital modeling of karst aquifers has had poor success for several reasons. Aquifer-wide details of karst hydraulics are poorly known. Nearly all numerical groundwater models are designed only for laminar flow, and even when turbulent-flow conduit modules are added, it is very difficult to predict well yields, flow directions, flow velocities and contaminant transport because of the local heterogeneity of karst aquifers. Nevertheless, there are several benefits from attempting to design digital karst models. By comparing the output from idealized models with actual field observations, the behavior of real aquifers can be better understood. Models of idealized conduit systems, designed...