A multivariate regression analysis for deriving engineering parameters of expansive soils from spectral reflectance

Soil expansiveness is a major problem that need due attention in geotechnical investigations. It can detrimentally influence performance and life time of lightweight engineering infrastructures. For this study, soil samples were collected from eastern part of Addis Ababa city. Specific expansive soil engineering parameters; Consistency limits (liquid limits (LL), plastic limits (PL) and plasticity indices (PI)), free swell and cation exchange capacity were measured in a soil mechanics laboratory. Reflectance spectra of each soil sample were acquired in a remote sensing laboratory using ASD fieldspec full range spectrometer. A multivariate calibration method, partial least squares regression (PLSR) analysis, was used to relate engineering parameters and absorption feature parameters calculated from the reflectance spectra of expansive soils at specific wavelengths (~1400 nm, ~1900 nm and ~2200 nm). Correlation coefficients obtained showed that a large portion of the variation in the engineering parameters (R=0.85, 0.86, 0.68, 0.83 and 0.64 for CEC, LL, PL, PI and FS respectively) could be accounted for by the spectral parameters. Results of the prediction models are high, indicating potential of spectroscopy in deriving engineering parameters of expansive soils from their respective reflectance spectra, and hence its potential applicability in geotechnical investigations of such soils.