Mapping a lithologically complex terrain using Sentinel-2A data: a case study of Suriyan area, southwestern Iran

Abstract

ABSTRACT The Suriyan area, southwestern Iran, possesses a variety of sedimentary, volcanic, and metamorphic rocks. This paper investigates the potential of Sentinel-2A multispectral data for lithological mapping of this geologically complex terrain using the principal component analysis (PCA), band ratio, and spectral information divergence (SID) methods. Spectral features of lithological units were analysed by surveying the laboratory spectra of hand samples, the resampled spectra of rock types to spectra from Sentinel-2A wavelength ranges, and the spectra extracted directly from applied imagery. The absorption features dominantly attributed to Fe2+, Fe3+, and Al-OH electronic and vibrational modes were observable in Sentinel-2A bands; however, the main absorption feature of Mg-OH and C-O near 2.33 µm is not collected in a distinct band by this satellite. Data processing results revealed that PCA was more successful in discriminating vegetation, greenschist, and shale, whereas it could not differentiate various types of limestone. The proposed Band Ratios, (46)/(78), (2 + 11 + 6)/(3 + 4), 12/8, and 2/5, enhanced sandstone, shale, basaltic dyke, greenschist, mica schist, meta-tuff, and various types of limestone. However, the SID algorithm discriminated few lithological units including greenschist, mica schist, shale, and Fe-bearing limestone. Considering output results, it was suggested that proposed band ratios as well as the PCA method were more successful in discriminating various lithological units, while the SID algorithm was advantageous in enhancing the metamorphic units. Field observations and microscopic studies verified the occurrences of rock units and their spatial compatibility with differentiated lithologies. It was concluded that Sentinel-2A data were successful in enhancement and discrimination of various rock types in a lithologically complex terrain.