RESUMO
Two hundred and thirty-four Lebanese olive samples were collected from different regions and the corresponding oils were analysed by (1)H NMR spectroscopy. The variables obtained, related to fatty acids and minor components, were used as inputs in univariate and multivariate analyses aiming to characterize and classify the oils according to geographical, morphological, and temporal factors. Samples were sorted according to the colour, size, and shape of olives, which allowed statistically significant classifications to be achieved. A sequential strategy was developed to discriminate among samples from different altitudes and latitudes. Following this strategy, obvious trends and classifications were obtained at subregional level. Furthermore, the shift in the harvest date within a range of three weeks was considered and its effect on the classification models was investigated. Likewise, the harvest year effect was evaluated; the precipitation level in April and May had a significant impact on the characteristics of the oils.