At sub-continuum scales, pore network geometry strongly influences flow and solute transport. Fractures spaced ~0.1 to ~10m impart continuum scales to fractured aquifers which may exceed the size of many practical problems. Sub-continuum flow models often use statistical simulation of explicit fracture networks, leading to serious problems of system identification an uniqueness. This studentship will address this by determining whether tracer breakthrough curves may be used to characterise fracture topography at different scales, either as the sole evidence or in combination with bulk permeability. The main method will be a systematic exploration of the influence of fracture network topologies and scale on tracer breakthrough, using simulations on 2-D and 3-D models which represent the void space explicitly and create dispersion solely as a product of inhomogeneity in flow rate. Results will be compared with existing tracer data and used to infer the characteristics of the fracture networks involved. The outcome hoped for will be a methodology for using tracer tests to constrain fracture geometry in unknown cases.