A method for correcting systematic error in crowdsourced bathymetric data based on dual-dimension compensation

In view of the limitations of existing methods in crowdsourced error correction, a dual-dimensional system error correction method based on static calibration and dynamic monitoring of shipborne equipment is proposed. Firstly, by combining the anchored static observation method with Kriging interpolation, a spatially continuous bathymeter bias surface model is constructed to effectively compensate for the spatial non-uniformity error of the sensor. Secondly, based on structured navigation log records and the sliding window least squares method, a dynamic draft compensation model is established to track the real-time impact of ship draft changes on water depth measurements. Experimental results show that the proposed method decouples the stable and random components of system errors, reducing the root mean square error (RMSE) from 1.23 meters to 0.45 meters and the mean absolute error (MAE) from 0.89 meters to 0.30 meters, outperforming traditional methods and effectively improving the accuracy and reliability of crowdsourced bathymetric data.