Optimizing alteration mineral detection: A fusion of multispectral and hyperspectral remote sensing techniques in the Sar-e-Chah-e Shur, Iran

• Published in: Remote sensing applications 2024-08, Vol.35, p.101249, Article 101249
• Main Authors: Habashi, Jabar, Mohammady Oskouei, Majid, Jamshid Moghadam, Hadi, Beiranvand Pour, Amin
• Format: Article
• Language: English
• Subjects: Advanced argillic alteration; ASTER; Data fusion; HyMap; Hyperion; Landsat 8; Mineral detection; Normalized score; Porphyry copper; Sentinel-2
• ISSN: 2352-9385; 2352-9385
• DOI: 10.1016/j.rsase.2024.101249

The Sar-e-Châh-e-Shur region in the Lut Block of eastern Iran has significant potential for diverse ore mineral deposits, particularly copper and iron mineralization. This study used multispectral remote sensing data, including ASTER, Sentinel 2, and Landsat 8 for detecting alteration zones associated with ore mineralization in the study area. Spectral ratios associated with porphyry copper-iron alterations and advanced argillic-iron alterations were utilized to establish training data, supplemented by hyperspectral data processing derived from Hyperion (spaceborne) and HyMap (airborne) sensors. Several classification algorithms were implemented, and evaluation metrics such as the Kappa coefficient and overall accuracy were calculated using a confusion matrix to assess classification performance. Field validation was conducted through thin mineral section analysis and geochemical stream sediment data, yielding a Normalized Score (NS) for conformity estimation. Notably, the Landsat 8, ASTER, and Sentinel 2 - classes category 2 -artificial neural networks-thresholds 2 (LAS-2-ANN2) class (representing advanced argillic alterations) was achieved a high score of 3.5, corresponding to an 87.5% match to the geological evidence. In addition, hyperspectral data processing using artificial neural network algorithms was yielded an overall score of 2.43, indicating a 61% match. In conclusion, this investigation highlighted the efficacy of fusing multispectral and hyperspectral data for accurate mineral alteration mapping in high potential regions, offering valuable insights for future exploration campaigns. •We employed hyper and multispectral imaging to directly map Copper–Iron.•Fusion of multispectral (ASTER, Sentinel 2, Landsat 8) data and hyperspectral (Hyperion, HyMap) data to identify alteration zones.•The normal score offers straightforward validation without requiring intricate calculations.