Spectral Wavelength Reconnaissance for Investigating Soil Silt Content

Recent advances in hyperspectral remote sensing have enabled detailed characterization of soil properties due to their high spectral resolution. This study investigates the spectral behavior of the soil silt fraction using laboratory reflectance spectroscopy. Soil samples were collected from the upper ۲۰ cm of the soil profile using a stratified random sampling approach based on auxiliary information layers in Mazandaran Province, northern Iran. Hyperspectral preprocessing techniques-including spectral averaging, Savitzky-Golay smoothing, and first-derivative transformations-were applied to enhance spectral features. Influential spectral domains and wavelength ranges associated with soil silt content were identified through correlation analysis. The highest correlation coefficients between silt content and spectral bands were observed at ۳۹۰ nm (R = ۰.۴۴), ۶۸۰ nm (R = ۰.۵۴), ۹۷۰-۹۹۰ nm (R = ۰.۵۹), ۱۴۰۰-۱۴۱۰ nm (R = ۰.۶۴), ۱۹۱۰-۱۹۳۰ nm (R = ۰.۶۷), ۲۲۰۰-۲۲۱۰ nm (R = ۰.۶۹), and ۲۳۴۰-۲۳۵۰ nm (R = ۰.۷۱). These wavebands exhibited strong potential as independent predictor variables for soil silt modeling. The identified diagnostic spectral ranges provide a foundation for future upscaling and digital soil mapping applications using airborne and spaceborne hyperspectral data.

کلیدواژه ها:

DSM ، PSS ، Soil Silt ، Hyperspectral Spectroscopy ، Upscaling