EGU24-11004, updated on 08 Mar 2024
EGU General Assembly 2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

ERODEM: Exploring Carbonate Rock Recession through Data Fusion of Extensive Experimental Data via Machine Learning

Claudio Mazzoli1, Chiara Coletti1, Luigi Germinario1, Lara Maritan1, Riccardo Pozzobon1, Nereo Preto1, and Dimitrios Kraniotis2
Claudio Mazzoli et al.
  • 1Department of Geosciences, University of Padova, Padova, Italy (
  • 2Department of Built Environment, Oslo Metropolitan University, Oslo, Norway

Pollution and climate change raise increasing concerns about the vulnerability of cultural heritage. In carbonate rocks, the primary concern is surface recession, which severely impacts the readability of details, preventing us from transmitting our legacy to future generations. Recession equations available in the literature are highly inadequate due to the complex relationship between climate conditions, hygrothermal (HT) behaviour, and stone textures. This is related to the limited set of parameters used by different authors and to the basic statistical approach used in their fitting processes.

Through this research, we aim to develop a robust and reliable model to predict stone recession, employing Machine Learning algorithms supported by Multivariate Statistical Analysis. We will examine a large database of surface recession measurements obtained from different types of carbonate rocks, which differ in their textural features (e.g. grain size, porosity) and HT behaviour (e.g. water vapour permeability), and of the relative micro-climate conditions under which they have been exposed during outdoor experiments. Additional recession data will be derived from laboratory experiments using an autoclave, allowing precise regulation of pH and temperature of water in contact with stone samples. Validation of the predictive model will involve comparing the recession predictions based on the time series of climate data and material characteristics with the observed recession obtained through the meticulous comparison of historical plaster replicas with the original monuments. This comprehensive analysis aims to ensure the model accuracy in capturing the real-world complexities of carbonate rock surface recession under varying environmental conditions.



ERODEM project was funded by the Department of Geosciences through the “Progetto Premiale” call. This initiative is part of the larger project "Le Geoscienze per lo Sviluppo Sostenibile," funded by the Italian Ministry of University and Research (MUR) within the frame of the “Progetti di Eccellenza 2023-2027”.

How to cite: Mazzoli, C., Coletti, C., Germinario, L., Maritan, L., Pozzobon, R., Preto, N., and Kraniotis, D.: ERODEM: Exploring Carbonate Rock Recession through Data Fusion of Extensive Experimental Data via Machine Learning, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11004,, 2024.