Site-specific outlier analysis of borehole datasets in Seoul, South Korea considering geostatistical optimization based on GIS

Subsurface information is commonly interpreted from borehole data that represent soil and rock profiles along boring logs with three-dimensional spatial coordinate information, layer information, and standard penetration test results. When possible a more reliable spatial distribution of the target physical properties can be guaranteed with additional testing of these locations to appropriately identify outlying observations of geotechnical uncertainties. Irregular measurements typically differ greatly from other measurements at the same site or significantly deviate from nearby observations. In this study, a framework for geostatistical optimization method for 3D geotechnical spatial grid, applying optimum outlier thresholds with clustering method, was proposed to incorporate site-specific uncertainties of the geo-layers to identify geotechnical characteristic value. And optimized geostatistical information for geotechnical boundary was derived and verified according to the sequential procedure for representative Seoul areas.