A replication study in dendrochronology-revisiting the panels of two portraits of Rembrandt.

In the Replicating a Rembrandt Study project, which revisited the attribution of two portraits of Rembrandt while exploring the strengths and limitations of replication studies in art history, dendrochronological research was carried out to reproduce and replicate research conducted in the 1990s. One of the portraits, at the Germanisches Nationalmuseum in Nuremberg, Germany, had then been attributed to Rembrandt, whereas the other one, at the collections of the Mauritshuis in The Hague, The Netherlands, was considered a copy. In this study, the reproduction involved reassessing the results obtained for the two panels in the 1990s by comparing the tree-ring measurements produced then with reference oak chronologies from the source area (Poland and the eastern Baltic), and with a third panel from the Rijksmuseum that had matched the wood of the Nuremberg portrait, and seemingly originated from the same oak tree. The replication entailed remeasuring the tree rings in the panels through digital photography and using modern software to compare them to new reference chronologies and the Rijksmuseum panel. Both approaches confirmed the original results, including the southern Baltic provenance of the wood and a same-tree match between the Nuremberg and Rijksmuseum panels. However, the reproduction identified measurement errors in the initial study, while the replication corrected these errors. Furthermore, thanks to the improved reference datasets currently available, the replication provided more accurate interpretations about the felling dates of the trees. This research demonstrates that dendrochronology is a reliable science that should yield consistent results if applied rigorously, regardless of the software or ring-width acquisition method employed. However, for reproducibility, detailed reporting, including reference datasets used and statistical values obtained, is required. Long-term storage of dendrochronological data and digital images from the tree-ring sequences allows for verification and reanalysis without the need to re-examine the artworks. Therefore, it is advised that museums and art collectors commissioning dendrochronological research request dendrochronological reports that contain detailed graphs and information, as well as the shared stewardship of the tree-ring datasets and digital images produced by dendrochronologists.