This one-day course is for all geoscientists who want to extract hidden information from substantial amounts of chemical and physical data using multivariate statistical (chemometric) tools. The course emphasizes applications rather than the mathematics of various chemometric methods and will include a demo version of Pirouette 5.0 chemometric software. Case studies focus on the following topics of immediate interest to geoscientists:

• Hierarchical cluster analysis (HCA) and principal component analysis (PCA) of biomarker and stable isotope data for oil-oil and oil-source rock correlation to establish petroleum systems.
• Quantitative regression analysis of chromatographic peaks by alternate least squares (ALS) to de-convolute mixed oils derived from two or more sources. In exploration, ALS identifies mixed oils (e.g., pre-salt and post-salt oils in the Middle East and Southern Atlantic). In production, ALS allows allocation of mixtures originating from multiple reservoir zones.
• Prediction of physical properties by partial least squares (PLS) of data obtained by micro-analytical techniques. PLS allows investigators to predict API gravity, sulfur, and viscosity for reservoir zones where only small samples of cuttings from storage are available for analysis.