Analytic comparison of multibeam echo soundings

Abstract: This study makes a numeric and visual comparison of two different multibeam echo sounding (MBES) surveys performed at the exact same location, in an effort to establish how reliable and comparable the vertical depth measurements are. The horizontal position uncertainty is not issued in this study. The analysis was performed on data delivered by two survey contractors, The Norwegian Coastal Administration (NCA) and the Norwegian Mapping Authority Hydrographic Service (NMAHS). The area of interest is located north west of the Norwegian city of Ålesund, and contains a dredged channel with eight distinctively Norwegian seamarks, perhaps best described in English as High Speed Vessel Seamark with Indirect Lights (HIB) making it safer to navigate the area. Several MBES surveys were performed both before and after the dredging of the area. The present study deals with NCA, NMAHS survey data and self-elaborated survey data from an old method with a rope and a sinker known as Weighted Sounding Line (WSL) to measure the depts. Comparative analyses were set up, using both visual methods and practical empirical methods. The findings are presented in maps, raster images and tables, illustrating results from each analysis. Two of the analyses were evaluated against the International Hydrographic Organization, IHO S-44 standard, setting preferences for hydrographic measurements. The analysis indicates some artifacts at especially the NCA data produced before the dredging. The first analysis method in the present study compares the two operator’s datasets before dredging within three test zones, and even though it is based on only a few test zones, the results still gives an indication that some differences occur. But perhaps just as important, two out of three measurements indicate a very good match. A second method makes a comparison between one of the datasets after dredging, and manually measured reference depths indicating an average uncertainty of 0.19m, well within the ±0.26m defined by the IHO S-44 standard for operating on depths of 10m. A third set of analysis were also performed on a cell by cell basis between the two surveys datasets collected before dredging, both as a visual analysis and a numeric analysis in tables. The visual comparison revealed that only a few number of cells had differences involving the values at the high end of the scale, as 3.5m to 5.3m. A pattern also emerged from the visual analysis, revealing what seemed to be artifacts along the heading direction of the survey. A numeric analysis was set up to find the value of the highest appearance of difference, the frequency the differences appear at. The difference of -0.13m occurs 18771 times, as the one with the highest frequency illustrated in a difference graph. Analysis against the IHO S-44 standard shows that 89% of the 410055 cells tested are within the IHO S-44 standard. Hillshading reveals the appearance of artifacts with a ridge pattern in the NCA data. The errors were described as an effect of incorrect velocity values through the water column on a multibeam swath. A second hillshade analysis performed on the NMAHS data revealed a missing HIB-West1, affecting the results of the analysis using WSL. The analysis also established that the NCA survey tends to measure the depths lower than the NMAHS survey. 91% of the cells are located lower than the NMAHS cell depth values. This could be a consequence from the fact that NCA calculates the middle value of the soundings as the valid cell value, while NMAHS selects the most shallow depth value indicating the minimum depth within the cell. Some artifacts also appeared as holes in the NMAHS dataset before dredging probably caused by variations of the vessels speed, mostly located at the north and south ends. The artifacts taken into account, the depth differences and the uncertainties generally still are within the IHO S-44 standard where this standard is applied. The NCA survey data before dredging did not cover the entire area of the dredged channel, and thereby fails in meeting the IHO S-44 criteria of a Full Seafloor Search for Special order surveys where under keel clearance is critical. 3D analysis contributes with a visual view of the area and a better understanding of the calculated differences.