Investigation of rare earth elements (REE) and trace elements in intrusive rocks via remote sensing (RS) and geografic information system (GIS) methods: A case study of Sivrihisar region

Rare Earth Elements (REEs)—comprising lanthanides, scandium, and yttrium—are crucial for technologies like renewable energy systems, electric vehicles, and defence applications. Notwithstanding their relative abundance in the crust, rare earth elements are rarely located in economically sufficient quantities due to complex separation processes. China dominates global rare earth element (REE) production, mostly attributable to well-defined REE-rich granitoids that exhibit an inverse correlation between REE content and silica (SiO2) concentration. This study presents a scalable, non-invasive exploration model that integrates geochemical data, Geographic Information Systems (GIS), and remote sensing (RS), focussing on the Sivrihisar and Tekören intrusive rocks in Turkey's Tavşanlı Zone. Although previous studies have acknowledged the enrichment of rare earth elements and thorium, especially in Kızılcaören, studies predominantly depended on field mapping and petrography, with minimal spatial or spectral analysis. This study addresses the gap by integrating georeferenced geochemical data (La, Ce, Nd, Th, SiO2) with Sentinel-2 multispectral and EnMAP hyperspectral imaging to delineate REE-hosting lithologies. The application of Inverse Distance Weighting (IDW) in QGIS demonstrated a distinct inverse association between REE concentrations and SiO₂, matching with trends noted in Chinese REE granites and suggesting magmatic differentiation and monazite crystallisation. GIS-based mapping revealed rare earth element enrichment in low-silica, mafic-rich granitoids. Sentinel-2 classifications (supervised, non-PCA) closely corresponded with geological maps, whereas EnMAP facilitated the extraction of unique spectral signatures over 224 bands, enhancing lithological differentiation of granitoids, marbles, alluvium, and metasandstones. Geologically, both intrusive bodies exhibit post-collisional characteristics, including magma mixing and hydrothermal alteration along structural zones. The combined geographic and spectral methodology provides an effective alternative for traditional exploration techniques. Subsequent investigations may refine this model by integrating machine learning, further field validation, and geophysical data to improve predictive precision in similar geological settings.