Inferring bed topography and stickiness from surface data on ice streams

At high sliding velocities bed topography and stickiness strongly affects the surface velocity and the surface topography over a certain range of horizontal scales. This opens up the possibility of estimating bedrock topography and stickiness from surface data on fast flowing ice streams. Formal inverse methods based on singular value decomposition are used to explore the degree of non uniqueness in the estimation. The forward model is based on a regular perturbation theory describing the effects of small-amplitude basal disturbances on a linearly viscous medium. The potential of this inverse procedure is illustrated using synthetic data and examples from ice streams.