STATISTICAL ESTIMATION OF SOIL TYPE USING CROSS-PLOTS OF S-WAVE VELOCITY AND RESISTIVITY IN JAPANESE LEVEES

Soil type of levee body and foundation is statistically estimated using cross-plots of S-wave velocity and resistivity in Japanese levees. S-wave velocity and resistivity are collected from surface wave methods and resistivity methods. Total survey line length of the geophysical methods is about 600km on 37 rivers in Japan. Relationship between S-wave velocity, resistivity, blow counts (N-value) and soil type is collected and stored in a database. The blow counts and soil types are collected from about 400 boring logs carried out on geophysical survey lines. S-wave velocity and resistivity at the depth of the blow counts were extracted from geophysical sections. The total number of extracted data is about 4000. Soil type is classified as clay, sand and gravel for the sake of simplicity. The data is grouped in levee body and foundation. A polynomial approximation was used to estimate soil type from S-wave velocity and resistivity. In the approximation, soil type is represented by discontinuous numbers one (clay), two (sand) and three (gravel). Polynomial equations are functions of S-wave velocity and resistivity and yield a continuous number between one and three. Constants of equations are optimized by a least squares method so that the residual between calculated value (from one to three) and actual soil type (one, two and three) is to be minimum. A soil type section can be estimated from S-wave velocity and resistivity sections using the polynomial approximations. Accuracy of estimation can be statistically evaluated by comparing estimated and actual soil types. In this paper, outline of geophysical methods, collected data, polynomial approximations and the accuracy of estimation will be discussed.