Identification of pegmatite dikes based on remote sensing spectral library and WorldView-3 Data: A Case study in the central Altyn region of Xinjiang

Abstract

The central Altyn region in Xinjiang is a prospective region for the mineralization of rare metals, such as lithium and beryllium, and is the target area for the exploration of rare metal deposits of the granitic pegmatite type in this study. However, harsh natural conditions complicate the effective identification of pegmatite dikes via conventional techniques, thereby limiting exploration. The narrowness of pegmatite dikes and aeolian sand cover presents challenge, yet remote sensing technology holds significant potential in this area. High spatial resolution data are essential, and a systematic remote sensing identification method has yet to be established. In this study, we established a spectral library of rocks and minerals in the research area and applied image enhancement techniques such as PCA, ICA, MNF, and band ratios to WorldView-3 imagery. By optimizing the display with RGB combinations, 764 pegmatite dikes were successfully identified, and five pegmatite dike group areas were delineated. A “dike-centered limited buffer” method for identifying mineralized pegmatite dikes was proposed. Using measured spectra combined with the spectral angle mapper (SAM) method, lithium-beryllium-bearing dikes were differentiated from non-lithium-beryllium-bearing dikes, and ultimately, 58 lithium-beryllium-bearing pegmatite dikes were identified. Field validation in Areas III, IV, and V indicates that the identification results align well with on-site observations. In this study, we significantly improved the accuracy of pegmatite dike identification in the central Altyn region, providing an effective remote sensing technique and theoretical support for rare metal exploration in this area.