Characterizing the Variability of Bedrock Surface Using an Efficient Constraint Seed Method

Bedrock surfaceis inherently uncertain due to spatial variability and the limited site-investigation data. Although random field theory is effective in evaluating spatial characteristics of soil and rock properties, it often overestimates site uncertainty when borehole data within the site are not incorporated. This paper introduces a novel Constraint Seed Method (CSM) for characterizing bedrock surface, effectively capturing spatial variability while integrating site investigation data. A construction site in Hong Kong, is used as an example to illustrate the effectiveness of proposed method, focusing on the determination of the Grade III rock surface depth. Site investigation data are utilized to continuously update and refine the bedrock surface. This method ensures that the updated Grade III rock surface depth can align with measured data at borehole locations and significantly reduces uncertainty in surrounding areas due to spatial correlation. By integrating data from these successive drilling stages, the method achieves a dynamic and increasingly accurate characterization of the bedrock surface. The proposed method offers a more precise bedrock surface, essential for geotechnical design and construction.

Keywords: Site investigation, Geology, Uncertainty, Analytical method, Updating characterization.