Hyperspectral indices developed from multi-angular bidirectional reflectance can trace the particle size of granite

The particle size of a rock is one of the critical physical properties related to a series of geological processes. While the prevalent photometric studies performed to quantitatively assess the physical properties of minerals and rocks rely much on radiative transfer model inversions from surface reflectance, the complicated procedures required seriously limit their applications. Instead, this study has attempted an empirical-based approach to trace the particle size of granite based on multi-angular bidirectional reflected information, providing a practical method for quickly assessing the physical properties of minerals and rocks from surface reflectance. Five common hyperspectral index types together with different spectral preprocessing treatments, including the original hyperspectral reflectance, derivative analysis, continuum-removed reflectance, and apparent absorption spectra, were screened to explore the best indices for both measured and Hapke model-simulated datasets. The results suggested that the normalized difference index (ND) calculated from the apparent absorption spectra using the wavelengths of 1 325 nm and 1 800 nm is a robust index for tracing the particle size for both measured and simulated datasets.

---

Opracowanie rekordu ze środków MNiSW, umowa nr SONP/SP/546092/2022 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2024).