The effect of bathymetric measurement uncertainties on multibeam echosounder sediment classification

Abstract

In the context of the Dutch Natura2000 monitoring program, the Cleaver Bank, located in the Dutch part of the North Sea, has been surveyed extensively in 2013, 2014 and 2015. The backscatter data acquired by a 300 kHz multibeam echosounder have been used for classifying the sediments. In addition to the backscatter, information is also contained in the bathymetric measurements. A parameter often used is the so-called depth residuals parameter. This, in theory, provides a measure for the roughness. However, the imperfectness of the bathymetric measurements also contributes to the depth residuals (as determined from the measured bathymetry). In this contribution the AMUST prediction tool that provides estimates for the uncertainty in the bathymetric measurements, resulting in variations in the measured bathymetry even for a perfectly flat seafloor, is described. The tool is used to assess the expected uncertainty in the MBES measurements taken at the Cleaver Bank. When comparing the predicted uncertainties and those derived from the measurements, a good agreement is found. It is hypothesized that deviations are, at least partly, due to the seafloor morphology. This information will be used as an additional measure to further improve the classification performance.