Seafloor spatial representation and bathymetry data quality

Multibeam echosounders are sophisticated systems which are more and more used by the geologist community for they allow the survey of large swaths providing simultaneously bathymetry and sonar imagery. The performance of such systems is often described in terms of swath width and depth accuracy. However the knowledge of the spatial representation that can be achieved with such multibeam systems has to be analyzed in detail. This should enable the geologists to better understand what are the smallest morphological features that can be actually detected. The size of these features depends in fact on the slant angle and distance as well as on the bottom detection scheme and the beam size and spacing. The analysis presented also puts emphasis on the effect of uncertainties in geometric parameters such as position and attitude of the vessel together with bottom slopes. A complementary aspect to the spatial resolution is the evaluation of bathymetry coherence between data set by quality estimation algorithms. In fact, the filtering of spikes in a data set does improve the quality but this one has yet to be estimated in a relevant way. For this purpose, there are versatile methods which can be applied at a later stage of the data processing, just before the data merging in view of obtaining a digital terrain model. These methods are aimed to assess the intrinsic quality of each separate data set (for example corresponding to a single profile) thus allowing the operator to select only the pertinent data sets according to the quality required by the final mapping scale. Data set merging is an important step in the whole process of mapping a given area with data gathered from one or several cruises for the data quality may significantly vary according to profile direction, vessel behavior, multibeam parameters (either fixed or selected)