Acoustic backscatter inversion for suspended sediment concentration and size: A new approach using statistical inverse theory

Abstract

A new method is introduced for estimating suspended sediment concentration and grain size from acoustic backscatter observations, incorporating techniques from linearized statistical inverse theory and data assimilation. A series of laboratory experiments with a sediment-laden jet were conducted for the purposes of demonstrating the method. The inversion results show improvements in stability compared to existing methods, in cases with highly-attenuating suspensions. These improvements are due to the use of statistical regularizing assumptions, which act to mitigate issues associated with poor conditioning and nonlinearity in the inverse problem. The new method is also more computationally efficient than existing methods, owing to the use of linearization. Matlab code is provided implementing the new method, see Appendix A.