Overcoming the temporal resolution bottleneck in daily climate records for high-frequency environmental stressors on cultural heritage
Authors:
- Vassilios DallasVDVassilios Dallas
- Diamando VlachogiannisDiamando Vlachogiannis
- Athanasios SfetsosAthanasios Sfetsos
- Ioannis KaratasiosIKIoannis Karatasios
Preprint
- 8 September 2026
- preprint
- Published by Elsevier in SSRN Electronic Journal
Abstract
The physical degradation of cultural heritage materials, driven by mechanisms such as thermal fatigue, freeze-thaw cycles, and salt crystallisation, is primarily governed by sub-daily environmental fluctuations. However, state-of-the-art climate projections, such as those from the CMIP6, are restricted to a daily temporal resolution, creating a significant technical bottleneck for high-resolution impact modelling. This study addresses this resolution gap by implementing a rigorous validation framework for the temporal downscaling of temperature, relative humidity, and wind speed. Using the ERA5 atmospheric reanalysis as a high-frequency ground-truth benchmark across five climatically diverse pilot sites, deterministic downscaling algorithms were evaluated for their ability to reconstruct sub-daily variability from daily statistics. Qualitative assessment utilising probability density functions reveals that the downscaling methodology effectively recovers the distribution tails and rare extreme events typically suppressed by daily averaging. Quantitative results substantiate this performance, yielding mean absolute errors for temperature near or below $1.0^\circ$C and median absolute errors for relative humidity below $5\%$, a level of precision comparable to the operational uncertainty of high-quality meteorological sensors. While wind speed reconstruction proves more challenging in maritime environments due to high-frequency gustiness, the methodology maintains high statistical fidelity and physical consistency for the bulk of the data. The practical utility of the framework is further demonstrated through the calculation of threshold-based indices, which reveal that reconstructed hourly data successfully identify chronic and accelerated exposure risks for porous concrete, traditional renders, and stone masonry that are systematically underestimated by daily aggregates. By providing a validated and transferable workflow, this research establishes the methodological foundation necessary to unlock the use of daily climate projections for robust, high-resolution risk assessments and the adaptive preservation of cultural heritage in a changing environment.