Revista de la Asociación Geológica Argentina
versión impresa ISSN 0004-4822
DELENDATTI, G. L.. Characterization of hydrothermal altered areas using Principal Component Analysis applied to Landsat TM imagery, San Juan province. Rev. Asoc. Geol. Argent. [online]. 2003, vol.58, n.1, pp. 97-108. ISSN 0004-4822.
The Principal Component Analysis (PCA) transformation technique using Landsat TM imagery has been applied to identify hydrothermal altered zones in the eastern rim of Cordillera Frontal, San Juan province (S. 30º 45´- 31º 00´ and W. 69º 30´- 69º 38´). The AR principal component images produced by transformation of TM 5-7, 4-5-7 and 1-4-5-7 were used to enhance zones with anomalous concentrations of argilic/phyllic alteration assemblages and silica-rich zones. FE images obtained by PCA of TM 3-4, 1-3-4 and 1-3-4-5 enhanced iron-oxide rich areas. Eigenvector analysis and loadings correlation between raw and transformed bands determined the close relationships between TM 5-7 and 3-4 with the AR and FE principal component images. One Crosta RGB composite was obtained by combination of AR, AR+FE and FE bands, highlighting three main alteration classes with similar spectral signatures. Class A comprised argillic and phyllic alteration assemblages (white pixels); class B, iron oxides and argillic/phyllic assemblages mixed zones (grey zones); and class C, silica-rich zones (darkest pixels). The field control comprised mapping of altered areas and analysis of hand samples with short wave infrarred (SWIR) spectrometer. The mineralogy determined by SWIR studies confirmed the characteristics of alteration classes identified from the TM imagery. Varied concentrations of sericite/illite +/- jarosite +/- tourmaline and alunite +/- jarosite were detected in the A and B zones. In addition, silica related with acid alteration was identified in the C zone. This study showed PCA to be a valuable tool for geological reconnaissance of large poorly mapped areas as it is able to reduce the risk of studying false anomalies.
Palabras clave : Satellite imagery; Hydrothermal alteration; Spectrometry; Cordillera Frontal; San Juan.