Optimal Spatial Interpolation Method Considering Topography and Density of Geotechnical Information

Abstract

Kriging is widely used in many fields to interpolate spatial data. Kriging is a geo-statistical method that predicts property values at unknown points using a weighted linear combination through known values of given points. Since geotechnical information can be irregularly distributed in the target area due to topographic and artificial factors, it is more efficient to apply spatial interpolation in the target area. However, when kriging is generally applied, there is a problem that the influence of the topographic classification and the topography characteristics is not considered. Therefore, it should consider the topographic classification, density of geotechnical investigation data, and grid size for dividing the area to perform reliable spatial interpolation.

The aim of this research is to propose a reasonable spatial interpolation method with high reliability when considering the effects of topographical characteristics (average slope, adjacent slope), grid size and density of borehole data. The geotechnical investigation data input form was standardized by DB schema. And the elevation of borehole data was corrected through the design guideline. To perform the topographic classification, the slope of the DEM was calculated by neighborhood algorithm, which is generally used to analyze slope. The target area was divided into 0.25km × 0.25km, 0.5km × 0.5km, and 1km × 1km, and spatial interpolation was performed by average ground condition (0.25km × 0.25km) and simple kriging for each layer boundary elevation and thickness. The optimal spatial interpolation method was proposed by using kriging and average ground condition.