Geotechnical and hyperspectral dataset for gold tailings

This dataset contains hyperspectral and geotechnical data collected from 300 artificially prepared gold tailings samples. The test specimens were prepared in petri dishes under controlled laboratory conditions with unique particle size distributions, moisture conditions, and densities. The geotechnical parameters of each test specimen were measured using applicable ASTM standard methods or calculated, and include percent sand, percent silt, percent clay, fines content, solids content, gravimetric moisture content, volumetric moisture content, degree of saturation, void ratio, porosity, total density, and dry density. Hyperspectral data were obtained using an ASD TerraSpec Halo spectrometer, which measures reflectance across a wavelength range of 350 to 2500 nm. The spectrometer includes three hyperspectral detector arrays, including a visible and near-infrared (VNIR) detector with a resolution of 3 nm, a short-wave infrared (SWIR) detector with a resolution of 9.8 nm, and a second SWIR detector with a resolution of 8.1 nm. For each test specimen, two scans were taken at different locations and averaged to produce a single reported hyperspectral reflectance. The dataset includes interpolated reflectance data at 1 nm intervals. The data are presented in a csv file where the first row contains column titles that describe the sample identification number, geotechnical parameters, and hyperspectral data. Each subsequent row in the dataset contains information about these fields for each of the 300 samples.

Methods
For a full description of the methodology used to collect these data, please refer to the associated publication titled Ex Situ Hyperspectral Sensing and Machine Learning for Tailings Characterization (https://doi.org/10.1139/cgj-2024-0656).