τ→μγ and μ→eγ as probes of neutrino mass models

We discuss the possibility of discriminating between different supersymmetric see-saw models by improving the experimental sensitivity to charged lepton flavour violating processes. Assuming a hierarchical neutrino mass spectrum, we classify see-saw models according to how the hierarchy Δm²Δm²_atm is generated, and study the predictions of each class for the branching ratios of τ→μγ and μ→eγ. The process τ→μγ is found to be a particularly promising tool to probe the fundamental see-saw parameters, and especially to identify the origin of the large atmospheric mixing angle. Predictions for μ→eγ are more model-dependent. We point out that, even with an improvement of the experimental sensitivities by three orders of magnitude, both τ→μγ and μ→eγ could escape detection in models where Δm²_atm is determined by one of the lightest right-handed neutrinos.