Leveraging robotic total stations and multi-sensor adjustment for accurate multibeam bathymetry

Abstract

In our study we geo-reference a multibeam echosounder (MBES) data set with a robotic total station (RTS). We use this data to perform a multi-sensor least-squares adjustment that leverages planar correspondences within the MBES data to estimate the trajectory. We compare this to GNSS-based processing and conventional direct geo-referencing. Our results show that the adjustment improves the precision from an interquartile range of 8 cm to 4 cm. Also, the height residuals of the trajectory adjustment can be reduced to 1 mm with RTS-based positioning versus 12 mm with GNSS. When the RTS-based MBES dataset is compared with bathymetric LiDAR, the standard deviation is 4 cm, and the median difference is -1 cm. Our approach offers improved accuracy and precision compared to GNSS-based systems and can also be used in GNSS-challenged environments such as mountainous regions or next to quay walls.