Indicating Compound Hazards on Porous Building Materials of Greece’s Monuments ^†

Abstract

The deterioration of porous building materials in archeological monuments is often intensified by slow and cumulative compound climate events, including salt crystallization cycles. This research examines the spatial patterns of these damaging events across Greece using high-resolution climate simulations derived from ERA-Interim and ERA5 Reanalysis datasets, as well as EURO-CORDEX models. By analyzing both past conditions (1980–2004) and projected future scenarios (2025–2049) under RCP4.5 and RCP8.5, the study identifies regions at heightened risk and explores how climate change influences the occurrence and possibly alters the aggressiveness of such events. By mapping the total frequency of these events and their anticipated changes under future climate conditions, this study contributes to developing a climatology of compound events that affect porous building materials of cultural heritage importance.

1. Introduction

The impact of atmospheric conditions on salt transitions events, such as crystallization/dissolution and hydration/de-hydration, that affect the structure of monuments’ porous materials in Greece, has been the object of study presented by Markantonis et al. [1] at COMECAP2023. The current study attempts a different approach on the calculation of the annual number of salt crystallization cycles, focusing on halite and sodium sulfate. Halite (NaCl) and sodium sulfate (Na₂SO₄) crystallization cycles demand a certain amount of time to occur and subsequently pose a threat to the material by causing cracks in the pores. For this reason, the main difference in methodology is the time step applied for the characterization of one cycle. Both events studied require two consecutive days of certain temperature (T) and relative humidity (RH) conditions, as analyzed in Section 2.2. The compound events studied consist of the combination of temporal and bivariate conditions leading to salt crystallizations. The examination of high-frequency compound events areas is conducted with the use of reanalysis models (ERA5, WRF-ERAInterim) for the historic period (1980–2004). Also, to quantify the impact of climate change between the historic and the future period (2025–2049) according to the frequency of salt transitions, four regional climate models (RCMs) were used.

2. Data and Methodology

2.1. Data

Mean daily ambient temperature (TG) and mean relative humidity (RH) values are necessary for the calculation of salt crystallization cycles for both types of salts. Two reanalysis models are used due to the lack of observations in most of the country. The first model was built in the Environmental Research Laboratory (EREL) of the National Center of Scientific Research “Demokritos” (NCSRD) or WRF_ERAI and was produced by dynamically downscaling ERA-Interim, using the Weather Research Forecast (WRF) model, to 5 km × 5 km [2]. The second model is the latest available reanalysis product, ERA5, from the European center for Medium-Range Weather Forecasts (ECMWF), of spatial resolution∼ 30 km × 30 km [3].

Differences between the historic and the future period (2025–2049) for the frequency of salt crystallization cycles can be examined with the use of data from RCM simulations driven by general circulation models (GCMs). In this regard, we obtained data from five models included in the EURO-CORDEX initiative. All chosen EURO-CORDEX models with available daily data for both RCP scenarios were selected because they have the finest spatial resolution of 12 km × 12 km and have also been evaluated by Cardoso et al. [4]. Information on the regional and parent models is given in

Table A1. List EURO-CORDEX simulation models information.

Institution	Reference	Regional Model	Forcing Model
Météo-France/Center National de Recherches Météorologiques	[7]	ALADIN63	CNRM-CERFACS-CNRM-CM5
Koninklijk Nederlands	[8]	KNMI-RACMO22E	ICHEC-EC-EARTH
Swedish Meteorological and Hydrological Institute	[9]	SMHI-RCA4	MPI-M-MPI-ESM-LR
Danish Meteorological Institute	[10]	DMI-HIRHAM5	NCC-NorESM1-M

of the Appendix A.

2.2. Methodology

Salt transitions occur when the conditions of RH cross a certain equilibrium, different for each kind of salt. In the case of halite, we adopt the methodology proposed by Benavente et al. [5]. Therefore, RH must be higher than 75.3% during the first day of the crystallization cycle, while it must drop below 65.3% on the second day. TG during both days ranges between 0 and 40 °C.

Sodium sulfate’s crystallization equilibrium is both TG and RH dependent. Three different pathways may lead to the transition of the sodium sulfate system, as is depicted in Figure 1. Each transition may cause an impact on the pores of the material. The arrows in the RH–TG diagram (Figure 1) denote the transition of the Mirabilite–Thenardite system equilibrium when the RH–TG equilibrium in the monuments’ microclimate changes. In the diagram, in Phase 2, TG is limited between 0 and 32.2 °C, and RH is described by Equations (1) and (2) below, that have been proven experimentally by Steiger and Asmussen [6]:

RH1 = −0.0122*(TG²) + 0.1218*TG + 98.2847

(1)

RH2 = 0.73*TG + 60.58

(2)

Phase 1 has no upper limit in TG and the RH must be higher than RH1 when TG is lower than 32.2 °C and higher than 85% when TG is higher than 32.2 °C. Finally, the upper limit of RH in Phase 3 is RH2 when TG is lower than 32.2 °C and almost constant at 85% for higher temperatures. In conclusion, damages on the porous materials may occur when sodium sulfate shifts (i) from Phase 1 to Phase 2, or (ii) from Phase 3 to Phase 2, or (iii) from Phase 1 to Phase 3, from the first day to the following.

To calculate the frequency of events for the EURO-CORDEX models, we calculate their ensemble which is the average number of the four different models. Each cycle, as described, is counted as one event.

3. Results

For both types of events, firstly, we present the mean annual frequency of the historic period of salt crystallization cycles for both reanalysis models and the ensemble of the EURO-CORDEX RCMs. Secondly, we present the difference between the historical period and the two RCP scenarios (RCP4.5 and RCP8.5).

3.1. Halite

Comparing the two reanalysis models, we can see that WRF_ERAI (Figure 2a) yields higher values than ERA5 (Figure 2b) in most of the country. ERA5 yields higher values (up to 12) in Epirus, Macedonia and Thrace, while WRF_ERAI reaches 16 events per year in Corfu and some mountain regions of Evia and Crete. The ensemble of RCMs (Figure 2c) exceeds 16 events per year in an extended region of Western Greece and mountainous regions of Thrace and Eastern Greece.

Figure 3 presents the differences between the past period of 1980–2004 and the two RCP scenarios. RCP8.5 (Figure 3b) yields the greatest changes with most areas of the country depicting negative change. That reduction reaches two events per year in the Evros region. RCP4.5 (Figure 3a) yields smaller changes than RCP8.5 at some areas in mountainous mainland Greece.

3.2. Sodium Sulfate

Similar to halite, WRF_ERAI is again the reanalysis that yields the highest number of annual sodium sulfate cycles. Most mountainous regions exceed 41 events per year, and a few regions reach 65 events per year (Figure 4a). On the contrary, ERA5 (Figure 4b) yields the highest values in the Ionian and Aegean islands. The geographical distribution of events for the ensemble of RCMs is similar to that of ERA5 (Figure 4c). The annual frequency of events nowhere exceeds 41 events per year.

In the case of sodium sulfate, positive trends can be observed for both scenarios in the mountainous areas of the country and especially for RCP4.5 (Figure 5a). RCP8.5 yields negative trends in most of the country and especially in the coastal regions in Southern Greece that reaches six events per year (Figure 5b).

4. Discussion and Conclusions

The examination of the mean annual frequency of salt crystallization cycles using two different reanalysis models showcased the importance of high spatial resolution results. Since both events are sensitive to relative humidity changes, RH may vary significantly in a few kilometers due to the complex orography of Greece. ERA5, even as a newer reanalysis model, is still very coarse to be safely used for the study of salt crystallization cycles. WRF_ERAI offers a more detailed study of the events, pointing to the need for downscaling the ERA5 dataset to the finest possible resolution.

Compared to the previous study [1], the new methodology reduces the number of events counted in the historic period, as it demands a longer time for the crystallization of salts. This method offers greater confidence in the counting of the events since the salt transitions in environmental conditions need this time to occur.

The use of the RCMs ensemble is necessary for the study of the impact of climate change on salt crystallization cycles. The methodology applied in this study is limited by the assumption that the mechanism that leads to the crystallization of salt remains stable and unaffected by climate change. Also, the study neglects that a monument may not be affected homogenously by the damage caused by weather conditions. Nevertheless, in this work, a more robust methodology is followed by setting new limits to the characterization of salt crystallization cycles. Thus, the results of the current study can be a useful tool for the identification of the vulnerability of stone-built structures in Greece.