A Practical Approach to Quality Control and Quality Analysis of Depth Data

Quality analysis of a collected depth dataset is important in order to find the statistic accuracy and variance of the depth data. Any systematic errors have to be identified and if possible minimised. It is also important to establish if the data fulfils the accuracy standard stated in the survey specification. The accuracy of collected depth data depends on several parameters such as the accuracy of the sensors in the survey system, the accuracy of the sensor calibration and the accuracy of any additional reductions to the dataset. After a proper calibration of the sensors, the survey system may be tested against a ground truth object on the seafloor to verify the horizontal and vertical accuracy. Ground thruthing can also be used in order to determine the object detection capabilities of the survey system. Online control has to ensure that sensors and collected depths are within acceptable limits. Additional reductions like tide or geoidal models etc. have to be analysed to ensure acceptable accuracy. The quality analysis of the resulting depth data may include ensuring sufficient data density for object detection and standard deviation calculation, data gaps, outlier detection, standard deviation and variation, crosscheck analysis, surface comparison, etc. This paper discusses some practical methods of establishing the theoretical accuracy of the survey system, to perform online control during data collection and to perform quality analysis on the resulting depth data.