Please cite this article in press as: Y. Asscher, et al., Combining multispectral images with X-ray fluorescence to quantify the distribution of pigments in the frigidarium of the Sarno Baths, Pompeii, Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.04.014 ARTICLE IN PRESS G Model CULHER-3587; No. of Pages 7 Journal of Cultural Heritage xxx (2019) xxx–xxx Available online at ScienceDirect www.sciencedirect.com Pompeii special issue Combining multispectral images with X-ray fluorescence to quantify the distribution of pigments in the frigidarium of the Sarno Baths, Pompeii Yotam Asscher a,∗ , Ivana Angelini b , Michele Secco c,d , Matteo Parisatto e , Antonina Chaban b , Rita Deiana b , Gilberto Artioli c,d a Israel Antiquities Authority, Hamarpe 5, Jerusalem 9777405, Israel b Department of Cultural Heritage, University of Padova, Piazza Capitaniato 7, Padova 35139, Italy c Center for the Investigation of Cement Materials (CIRCe), University of Padova, Via Gradenigo 6, Padova 35131, Italy d Department of Geosciences, University of Padova, Via Gradenigo 6, Padova 35131, Italy e GeoMEB S.r.l.s., Montereale Valcellina, Pordenone 33086, Italy a r t i c l e i n f o Article history: Received 30 July 2018 Accepted 19 April 2019 Available online xxx Keywords: Multispectral imaging Portable X-ray fluorescence (pXRF) Fiber-optic reflected spectroscopy (FORS) Pompeii Wall paintings a b s t r a c t Multispectral imaging is used to identify and semi-quantify the distribution of pigments in wall paintings based on mineral-specific band ratios. The western wall in the frigidarium of the Sarno baths in Pompeii was imaged using VIS-IR band-pass filters, stacked as multispectral data, and different pigments were measured using portable X-ray fluorescence (pXRF) and fiber optic reflected spectroscopy (FORS) to determine their chemical composition and spectral signature. Cuprorivaite and hematite were found to be the main minerals in blue and red pigments, with higher amounts of copper and iron respectively. Proportional relations were found between cuprorivaite specific band ratios from multispectral and FORS data and the intensity of copper in pXRF spectra, allowing to map the mineral content on a large scale and determine their relative concentrations. These results were confirmed by X-ray powder diffraction (XRD) and thin section analysis on painted fragments that were detached from the wall paintings and found below. This contribute indicates that on-site portable instrumentation could identify specific pigments and determine their 2D distributions based on non-invasive diagnostics of chemical composition and spectral reflectance in wall paintings. © 2019 Elsevier Masson SAS. All rights reserved. 1. Introduction The development of efficient non-invasive methods for the char- acterization of wall paintings in archaeological sites is of crucial importance for the conservation of valuable degraded frescoes, such as those found in Pompeii. In particular, when applied to wall paintings in Pompeii, any type of micro-invasive analysis is very limited due to their good state of conservation, and approaches for non-invasive characterization are in need. Recent developments for non-invasive multi-analytical characterization of paintings enable identification of pigments, and mapping their distribution, based on high-resolution hyperspectral reflectance spectroscopy in com- bination with elemental distributions. These techniques, however, are based on delicate non-portable instrumentation which is lim- ited to the museum environment [1–3]. In the archaeological site, the use of on-site mapping techniques may be often limited to qual- ∗ Corresponding author. E-mail address: yotama@israntique.org.il (Y. Asscher). itative study, including the evaluation of the wall decoration and the identification of areas of interest for subsequent micro-invasive analysis [4–6]. A quantitative approach for the identification of pigments, materials and their mixtures, using non-invasive multi- analytical techniques, will allow conservation efforts to manage and restore properly wall paintings on-site, in light of the increased degradation outside the museum environment. Quantification of non-invasive measurements is extremely difficult, due to the sur- face properties that include heterogeneity, mixing of pigments and degradation, as well as the field conditions, which are far from ideal for any systematic study. In this paper, we present a new multi-analytical non-invasive approach to identify, map the distribution, and determine the rel- ative concentrations of pigments in wall paintings on-site. Our strategy is to correlate well-established measurements on painted fragments that were studied using X-ray diffraction (XRD) and thin sections analysis, with our non-invasive techniques that include multispectral imaging, portable X-ray fluorescence (pXRF) and fiber-optic reflectance spectrometer (FORS) measurements. This way we establish a semi-quantification tool-kit by correlating well- https://doi.org/10.1016/j.culher.2019.04.014 1296-2074/© 2019 Elsevier Masson SAS. All rights reserved.