Mosaic floors of Roman Villa del Casale:Principal component analysis on spectrophotometric and colorimetric data.

Abstract Spectrophotometric and colorimetric data obtained during a measurement campaign aimed at supporting the Roman "Villa del Casale" (Piazza Armerina, Sicily, Italy) conservation activities, are presented. Special attention was paid to the possible variation of the chromatic coordinates, possibly due to the interventions of cleaning, consolidation, and protection. Data have been analyzed by the Principal Component Analysis (PCA) statistical technique, with the attempt to investigate its role in data variability reduction and verify its effectiveness in interpreting the phenomena occurring on the mosaic surface of the Villa, through grouping the observations into homogenous clusters. Effectiveness in the use of the information provided by the spectrophotometric and colorimetric analyses is strongly related to the immediacy and ease of data reading by the restoration operators for whom the issues concerning the color measurement and its representation are often unfamiliar. This paper analyses data of different mosaic tesserae before/after the cleaning intervention and presents data clustering with PCA. This statistical technique has provided a synoptic scheme capable of improving data interpretation concerning the chromatic behavior of the materials. Moreover, the cluster distribution highlighted by the multivariate analysis made it possible to identify, more clearly, the parameters that mostly contribute to the chromatic shift and to monitor the behavior of variously colored tesserae. © 2013.