Automatic Statistical Processing of Multibeam Echosounder Data

This paper presents the CUBE (Combined Uncertainty and Bathymetry Estimator) algorithm. Our aim is to take advantage of statistical redundancy in dense Multibeam Echosounder data to identify outliers while tracking the uncertainty associated with the estimates that we make of the true depth in the survey area. We recognise that a completely automatic system is improbable, but propose that significant benefits can still be had if we can automatically process good quality data, and highlight areas that probably need further attention. We outline CUBE and its associated support structures, and apply it to a dataset from Woods Hole, MA, USA. We illustrate CUBE\u27S output surfaces, show that the algorithm faithfully maintains significant bathymetric detail, and how the algorithm\u27s auxiliary outputs can be used in the decision-making process. Comparison with a selected sounding set shows that CUBE’S outputs agree very well with traditional approaches