Assessment of subpixel methods for determining petroleum microseepage-induced alterations using sentinel 2 data in the Gachsaran Formation, Haftkel oilfield, SW Iran

Abstract Hydrocarbon microseepage from subsurface reservoirs alters the mineralogical and geochemical properties of near-surface sediments producing detectable mineralogical, geochemical and spectral anomalies. Conventional detection methods such as XRD, XRF, and geochemical assays offer reliable analytical accuracy but are spatially limited and costly. This study assesses the capability of Sentinel-2 multispectral data integrated with subpixel analytical techniques to detect microseepage-induced alterations in the Haftkel oilfield, SW Iran. Laboratory spectroscopy and XRD analyses indicated mineralogical signatures associated with seepage-related alteration. Band ratios (B8/B11, B3/B4) enhanced iron-reduction anomalies but were insufficient to map sour gypsum. Pixel Purity Index (PPI) and subpixel algorithms such as Mixture Tuned Matched Filtering (MTMF), and Linear Spectral Unmixing (LSU) were utilized to extract and detect microseepage-induced alteration signatures. MTMF delineated major alteration zones, while LSU produced the most accurate results, including the detection of a previously unrecognized microseepage-induced altered site which was supported by field observations. The results show that integrating Sentinel-2 with LSU offers an effective approach for detecting hydrocarbon microseepage-induced alterations in evaporitic terrains.