Resolving original signatures from a sea of overprint - The geochemistry of the Gungri Shale (Upper Permian, Spiti Valley, India)

2012 
Abstract The geochemistry of black shales can faithfully record paleoenvironmental conditions. However, the original geochemical signatures are obscured by post-depositional events such as diagenesis and weathering. Resolving the relative influence of these processes on the geochemistry of shales requires a combination of geochemical and chemometric (chemical statistics) approaches to data analysis that identify the sources of variance in major and trace element geochemistry. The Permian Gungri Shale (Kuling Group) at Attargoo (Spiti Valley, Himachal Pradesh, India) is a gray to black shale capped, unconformably by an iron-rich pebble-sand layer (“ferruginous layer”) marking, ostensibly, the Permian–Triassic boundary. High resolution analysis of the Gungri at Attargoo (~ 3 cm intervals) reveals a geochemical record complicated by low-grade metamorphism/diagenesis and modern weathering. Using multivariate chemometric techniques (e.g., Principal Component Analysis) we identified variance in the geochemical record associated with the original detrital and depositional conditions. We also identified 4 beds within the Gungri that mark “depositional events” which may correlate to those identified in other PT sections (e.g., Guryul Ravine, Kashmir). Identification of sources of variance allowed us to identify geochemical signatures that can be used for paleooceanographic reconstruction. The Attargoo section records a sequence of discrete events including transient iron-enrichments attributed to euxinic depositional conditions and late diagenetic formation of siderite and regression condensation surfaces associated with pulsed transgression–regression towards the uppermost Permian. Our results show that chemometric analysis of geochemical data enables the resolution of paleoenvironmental conditions in shales impacted by post-depositional alteration.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    86
    References
    19
    Citations
    NaN
    KQI
    []