Research on Multivariate Statistical Methods for the Identification of Ancient Ceramics from the Husi Kiln

The identification of ancient porcelain has primarily relied on the manual "visual inspection" by researchers in the cultural heritage sector, lacking scientific and unified criteria.This approach is highly subjective, and the results are often unconvincing.The Husi Kiln, discovered in the 1970s, is a newly found kiln site.Its production periods, vessel shapes, and firing techniques overlap significantly with those of the Hutian Kiln and the Longquan Kiln.The discovery of the Husi Kiln site has made the identification of the kiln origins of existing porcelain products increasingly complex and difficult.Based on this, this study employs X-ray fluorescence spectroscopy, a modern nuclear analysis technique, to conduct scientific sampling and measurement of Qingbai glazed porcelain from the Song Dynasty produced by different kilns.The compositional makeup of the body and glaze of porcelains from different kilns was analyzed.Using multivariate statistical analysis methods, characteristic patterns in the compositional makeup of products from different kilns were identified.Furthermore, a "fingerprint" characteristic component analysis method for porcelain kiln identification was proposed.This provides a reliable basis for the scientific and effective identification of Husi Kiln porcelain, can help determine the provenance of Qingbai glazed porcelain found domestically and overseas, and assist in finding the home for Hubei's treasures.