Abstract: In recent years, evaluation of site geological conditions attracts more attention as more offshore projects are developed. Based on side-scan sonar data, this study calculates the horizontal scale of fluctuation using an improved space average method and an autocorrelation function method, and obtains a map of soil thickness distribution on the seabed over an offshore area by applying the ordinary Kriging method and the data of horizontal scale calculations. Results show that it is feasible to analyze the spatial autocorrelation and variability of soil thickness using the side-scan sonar data, and our scale calculation method is more stable and simpler. It is found that this scale falls in the range of 400-600 m for the soil layers in the study area. Generally, the accuracy of soil thickness distribution calculated using a data interpolation scheme that is determined by the calculated scales is satisfactory. However, evaluation errors may be quite big for those locations with small correlation distance, indicating a significant special variation in soil properties. To reduce such errors, the sweeping path of the side-scan sonar and the number of boreholes near these locations should be replanned to meet project needs.

Key words: side-scan sonar, horizontal scale of fluctuation, regional evaluation, soil thickness, geophysical data, the ordinary Kriging interpolation method