Grid search inversion of teleseismic receiver functions

Reliable inversion of teleseismic receiver functions (RFs) for crustal and upper-mantle structure is difficult due to the non-linearity and non-uniqueness of the problem. We report here on an approach, based on the degrees of freedom in the signal, that produces models with similar complexity to the data. The approach mitigates the non-uniqueness and overcomes the difficulties of non-linear minimization. The method begins with an initial forward modelling exercise to determine layer numbers and inter-layer properties that are input to the grid search method. The grid search uses heuristics to reduce the search space size and yields a suite of acceptable models for the crustal and upper-mantle structure, based on a misfit measure derived from the pointwise uncertainties of the RF estimate. We compare our grid search method with a standard iterative inversion method, using data from the IRIS network station SACV, on Santiago Island in Cape Verde, to demonstrate the wider range of models our method can produce.