Coliseum project: Correlating climate change data with the behavior of heritage materials

TL;DR

The paper introduces the COLISEUM project, which combines climate data and heritage material monitoring using AI models across three French sites to predict deterioration under climate change scenarios.

Contribution

It presents a novel methodology for collecting multimodal data at heritage sites and developing AI-based weathering models to forecast material deterioration due to climate change.

Findings

Initial diagnostic results from Strasbourg Cathedral site.

Development of a climate monitoring methodology for heritage sites.

First weathering prediction models using collected data.

Abstract

Heritage materials are already affected by climate change, and increasing climatic variations reduces the lifespan of monuments. As weathering depends on many factors, it is also difficult to link its progression to climatic changes. To predict weathering, it is essential to gather climatic data while simultaneously monitoring the progression of deterioration. The multimodal nature of collected data (images, text{\ldots}) makes correlations difficult, particularly on different time scales. To address this issue, the COLISEUM project proposes a methodology for collecting data in three French sites to predict heritage material behaviour using artificial intelligence computer models. Over time, prediction models will allow the prediction of future material behaviours using known data from different climate change scenarios by the IPCC (Intergovernmental Panel on Climate Change). Thus, a climate monitoring methodology has been set up in three cultural sites in France: Notre-Dame cathedral in Strasbourg ( 67), Bibracte archaeological site (71), and the Saint-Pierre chapel in Villefranche-sur-Mer (06). Each site has a different climate and specific materials. In situ, microclimatic sensors continuously record variations parameters over time. The state of alteration is monitored at regular intervals by means of chemical analyses, cartographic measurements and scientific imaging campaigns. To implement weathering models, data is gathered in alteration matrix by mean of a calculated weathering index. This article presents the instrumentation methodology, the initial diagnostic and the first results with the example of Strasbourg Cathedral site.