A tool for a fast, semi-automatic cleaning of multibeam swath bathymetric records from spikes

Here we show a tool for the auto-cleaning of Digital Terrain Models (DTM) from spikes. This tool basically fits the contiguity of well developed depth measurements in a regularly spaced DTM, that helps to individuate and remove spikes from survey records. The method consist in a recursive approach in which a bathymetric grid is generated starting from very large cell size values (e.g., about ten times the grid cell size of the final product) and then compared with the vertical distance between each measurement and the filtered DTM (residuals). Finally, a threshold value is opportunely chosen and beams with residuals that are larger than the threshold value are excluded from the cleaned record. The processing flow requires several iterations cycles, made by reducing both the cell size and the threshold value for each iteration step. The method here proposed is a self-excluding process that helps to choose between data by killing (or not) each single measurement, thus resulting in a simple removal of wrong measurement from the original bathymetric record.