Research on Vegetation Effect Removal and Hydroxyl Mineral Identification Based on Vegetation Corrected Continuum Depths Model

Hydroxyl-bearing (OH-) minerals (mica group and clay minerals) constitute important components of surface rocks and hydrothermal alteration zones. Their diagnostic absorption feature in the shortwave infrared region (around 2 200 nm) serves as a key criterion for mineral mapping using remote sensing. However, in vegetated areas, spectral interference from both green and senescent vegetation severely constrains the effective extraction of this diagnostic feature. To mitigate vegetation effects, this study introduces vegetation corrected continuum depths model based on hyperspectral imagery acquired by PRISMA, and conducts pixel-scale vegetation effect removal and hydroxyl mineral identification. The model establishes a multivariate linear regression relationship between vegetation-sensitive bands (674 nm and 2 078 nm) and the mineral diagnostic band (2 200 nm) in terms of absorption depth, thereby quantitatively reducing vegetation-induced spectral interference and restoring the true absorption depth of hydroxyl-bearing minerals. The Wuzengsayi mining area in Altay, Xinjiang, was selected as the study area, where lepidolite—a typical hydroxyl-bearing mineral closely associated with lithium mineralization—is widely developed. The results demonstrate that the hydroxyl mineral abundance anomalies extracted after correction show a high spatial consistency with the distribution of known ore-bearing pegmatite veins. Field sampling and petrographic microscopic identification confirm that samples from high-abundance zones are dominated by lepidolite, validating the effectiveness of the proposed method for identifying hydroxyl-bearing minerals in vegetated regions.