Historic masonry buildings are an integral part of human cultural heritage, and they need to be preserved for future generations. Brick is susceptible to frost damage which is common in regions with cold and humid climates. The frost damage on the masonry walls is accumulated over the years becoming more and more critical for the integrity of the historic buildings and it is also affected by climate change (CC).
In the current research, the focus was placed on a coastal region in southern Norway with a significant number of historic masonry buildings. The frost damage risk of the masonry walls was assessed by using data from a climate reanalysis for the present conditions and from a climate model under past, present, and future conditions. Two climate-based (CB) indices accounting for the air temperature and one material response-based (MRB) index considering the temperature and moisture content inside a simulated masonry wall were used for the frost damage risk assessment. The inputs for the MRB index were calculated by heat, air, and moisture (HAM) transfer simulations. Within the HAM simulations, the indoor climate was in one case representative of an unconditioned building with air leakages and many openings, while in the second case it was representative of a small, conditioned room.
The overall impact of CC was a decrease in the frost damage risk of the masonry walls. However, an increased frost damage risk was observed from the present to the future conditions according to the MRB index for the walls of small, conditioned rooms with higher driving rain load and lower solar radiation gains. The (i) number of freeze-thaw events, (ii) periods during which freeze-thaw events occur, and (iii) CC-related trends varied based on the considered index with the most explicit risk assessment being the MRB one. Moreover, the freeze-thaw events experienced by the masonry walls of unconditioned, leaky buildings were 20 times more than the ones for the small, conditioned rooms. Significant differences were observed between the results from the climate model and the climate reanalysis which were mainly linked to the underestimation of the air temperature and the overestimation of the precipitation by the climate model. The outputs of the MRB index were translated into certain damage categories while suggestions on improving the limitations of the current research were made.
Frost shattering or brick breakdown by freezing of water present in pore spaces and joints has been considered the prevailing weathering process in cold regions (French, 2017; Washburn, 1979; Matsuoka, 1990). Frost damage of masonry is attributed to a number of mechanisms (Mensinga et al., 2010) with the prime one being the 9 % volumetric expansion that accompanies the phase change of water to ice (Ollier, 1969; French, 2017; Matsuoka, 1990; Lisø et al., 2007). This theory was also supported by Davidson and Nye (1985) who found that the ice in a 1 mm-wide slot made in lucite produced pressures up to 11 bar by volumetric expansion. Since this value is greater than the tensile strength of some porous rocks, the volumetric expansion causes frost shattering.
The
The city of Tønsberg, located on the south coast of Norway, has numerous masonry buildings of historical significance. Some examples are depicted in Figs. 1a, 2a, 3a, and 4a. Tønsberg has a cold season that lasts from November until April, during which the precipitation and the air relative humidity remain at high levels. In many cases, the wall surfaces of the historic masonry buildings flake off and crumble away due to frost damage (Figs. 1b, c, 2b, c, 3b c, 4b, c). Thus, it is important to investigate, understand, and strategize against the degradation caused by this deterioration mechanism. The assessment of past, present, and potential future conditions is also important.
Existing literature suggests a decreasing trend of the frost damage risk in southern Norway due to climate change (Sabbioni et all., 2010; Kaslegard, 2011; Leissner et al., 2015; Loli and Bertolin, 2018). However, the severity of the calculated degradation may vary significantly based on the modelling approach that is used for the frost damage risk assessment (Van Aarle et al., 2015; Sahyoun, et al., 2019; Vandemeulebroucke et al., 2019, 2021a, b). A well-accepted method to assess the frost damage risk of masonry buildings is by counting the freeze-thaw cycles that their building materials experience. The freeze-thaw events can be calculated by considering climate-based indices (Viles, 2002; Grossi et al., 2007; Brimblecombe, 2010; Sabbioni et al., 2010; Leissner et al., 2015; Loli and Bertolin, 2018) or material response-based indices (Sedlbauer and Kunzel, 2000; Straube and Burnett, 2005; Straube and Schumacher, 2006; Mantha and Arena, 2012; De Rose et al., 2014; Van Aarle et al., 2015; Vandemeulebroucke et al., 2019, 2020, 2021a, b; Zhou et al., 2020; Hao et al., 2020; Sahyoun et al., 2020).
A widely used climate index for the frost damage risk assessment accounts
for the number of incidents within which the air temperature drops below 0
In reality, the water inside the material's pores does not freeze sharply
below 0
A climate-based index was used by Nelson and Outcalt (1987) as an indicator
of the severity of the frost damage on the material of interest.
Specifically, they calculated the number of degree days below 0
The
The outputs of the climate-based indices can be used as an indicator of the frost damage risk of the material of interest. However, there are no certain threshold values above which the building material is considered to be damaged or has a certain degree of degradation. Thus, the climate-based indices can only be used for comparative analysis among data from different periods, areas, and data sources.
A more computationally demanding process to assess the frost damage risk of
the masonry walls is to use material response-based indices. According to
this approach, certain climate parameters and material properties serve as
inputs in hygrothermal simulations (Delgado et al., 2010). The transient
hygrothermal conditions, i.e., the temperature, relative humidity, and
moisture content of the material serve as inputs in the material
response-based indices to assess the frost damage risk of the material of
interest. Sedlbauer and Kunzel (2000) and Van Aarle et al. (2015) used 0
Moreover, it is clear that the material needs to be wet enough in order to
get degraded after a freezing event. The material wetness is widely
expressed by the degree of saturation (
One main benefit of the material response-based indices that take into
account
A slightly different approach from the one that takes into account the
In the current research, two climate-based indices, accounting for the outdoor air temperature (Grossi et al., 2007; Sabbioni et al., 2010), and one material response-based index, accounting for the temperature and moisture content of a 5 mm layer in the exterior side of a simulated historic masonry wall (Sedlbauer and Kunzel, 2000; Vandemeule-broucke et al., 2019), were utilized. Past, present, and potential future conditions were assessed by using data from a climate model derived from the EURO-CORDEX project data archive (Jacob et al., 2014). Data from ERA5 (Hersbach et al., 2020) were used to assess the accuracy of the climate model results. The frost damage risk according to the material response-based index was calculated for eight different orientations, i.e., the cardinal and the intermediate ones, and two different indoor climate cases corresponding to a worst-case and a best-case scenario. According to the worst-case scenario, the indoor temperature and relative humidity were considered equal to the outdoor ones, while in the best-case scenario, the indoor temperature and relative humidity were calculated according to EN 15026.
The
The current research accounts for climate change by using data produced with
the latest hydrostatic version of the REgional MOdel REMO, version REMO2015
(Jacob and Podzun, 1997), driven by the low-resolution version of the Max
Planck-Institute Earth System Model, MPI-ESM-LR (Giorgetta et al., 2013).
The model data were acquired from the EURO-CORDEX project (Jacob et al.,
2014) data archive. They were synthesized into three 10-year periods,
1960–1969 referred to as past, 2010–2019 referred to as present and
2060–2069 referred to as future conditions. The Representative Concentration
Pathway (RCP) 8.5 was taken into account for the present and future
conditions. The RCP 8.5 represents a high-end emission scenario leading to
accelerated global warming during the ongoing century (Riahi et al., 2011).
The climate model data have a spatial resolution of 0.11
In order to assess how much the modelled climate deviates from the actual
one, a fourth climate file was prepared to contain data from the
fifth-generation European Centre for Medium-Range Weather Forecasts (ECMWF)
climate reanalysis, ERA5 (Hersbach, 2020) for the period 2010–2019 (present
conditions). Climate reanalyses combine past observations with models to
generate consistent time series of multiple climate variables. The spatial
and temporal resolution of the ERA5 data is 0.25
The climate parameters considered in the current research are the following:
Air temperature ( Air relative humidity (%); Precipitation (mm); Wind speed (m s Wind direction ( Cloud cover (oktas); Atmospheric long-wave counter-radiation incident on a horizontal surface
(W m Global short-wave radiation incident on a horizontal surface (W m Diffusive short-wave radiation incident on a horizontal surface (W m Direct short-wave radiation incident on a horizontal surface (W m
Some of the aforementioned parameters were downloaded directly from the
respective online archive, while others needed some further processing in
order to be calculated. The exact same data were used and described in
detail by Choidis et al. (2021) in a study of the impact assessment of
climate change on the biological deterioration of two timber historic
buildings in Tønsberg.
Moreover, the driving rain and the incident solar radiation on vertical
components facing toward the cardinal and intermediate orientations was
calculated by using the hygrothermal software Wärme und Feuchte
Instationär, WUFI® Pro (Kunzel, 1995;
Karagiozis et al., 2001). In more detail, the driving rain was calculated
using Eq. (1) (Karagiozis et al., 2001):
Frost damage on bricks is the result of mechanical stresses, mainly
attributed to the volumetric expansion during the transformation of water
into ice in the material pores (Ollier, 1969; French, 2017; Davidson and
Nye, 1985; Matsuoka, 1990, 2001). In the current research, two
climate-based indices were used to assess the frost damage risk. The first
one is the number of 0
The number of freeze-thaw cycles that occur during the examined periods can be used as an indicator of the frost damage risk. There is not a certain number of freeze-thaw events above which the building material is considered to be degraded. Thus, only a comparative analysis among data from different periods and between the data from the climate model and ERA5 is possible. The higher the number of freeze-thaw events, the greater the frost damage risk.
A more detailed and computationally demanding process to assess the frost damage risk of the masonry walls is to use a material response-based index. In the current study, a commercial heat, air, and moisture (HAM) transport software was used to calculate the temperature and the moisture content of the material of interest, given proper climate data and material properties as inputs. Proper threshold values for the material temperature and moisture content were selected in order to account for the freeze and thaw events.
The software WUFI® Pro (Kunzel, 1995; Karagiozis et al., 2001) was used to simulate the hygrothermal performance of the masonry walls. WUFI® Pro is a solver of two coupled heat and moisture transport equations. The software accounts for heat transport via (i) thermal conduction, (ii) enthalpy flow through moisture movement with phase change, (iii) short-wave solar radiation, (iv) night-time long-wave radiation cooling. Moreover, it accounts for vapor transport through (i) vapor diffusion and ii) solution diffusion. Finally, the liquid transport mechanisms included in WUFI® Pro are (i) capillary conduction and (ii) surface diffusion.
The outdoor climate parameters considered within the hygrothermal
simulations are the ones mentioned in Sect. 2.1. For the indoor climate two
different cases were examined for each outdoor climate excitation. According
to the first one the indoor air temperature and relative humidity were
considered equal to the outdoor ones. This worst-case scenario could be
representative of a wall that is exposed to the outdoor climate on one side
and on the other side, it is shaded (Choidis et al., 2020). It could also
represent an unconditioned building with air leakages, and many openings. In
the second case, the indoor climate conditions were computed based on the
outdoor air temperature, as prescribed in EN 15026. The indoor temperature
ranged between 20 and 25
In practice, the indoor climate of historic masonry buildings can be in between the selected worst- and best-case scenarios of the current research. It may vary significantly based on the building volume, area of windows, wall dimensions and layers, moisture buffering performance, and presence of active climate control (Leissner et al., 2015; Loli and Bertolin, 2018; Choidis et al., 2020; Califano et al., 2022). All these parameters were taken into account by Bertolin and Camuffo (2014), Leissner et al. (2015), Loli and Bertolin (2018) in order to calculate the indoor climate in different building categories and further assess their energy performance and the deterioration of artifacts that are hosted indoors. In the current research, a more straightforward approach was followed with focus on a generic cross-section, a worst-case and a best-case scenario for the indoor climate and investigation of the hygrothermal performance and frost damage risk of a critical area at the exterior side of the masonry wall which according to observations (Figs. 1, 2, 3, 4) and literature (Straube and Schumacher, 2006; Ueno et al., 2013a, b, c; Van Straaten, 2016; Vandemeulebroucke et al., 2019, 2020, 2021a, b; Sahyoun, 2020) is at higher risk.
The two different considered cases for the indoor
The measured dimensions of the bricks in Figs. 1, 2, 3, 4, were
Simulated masonry wall and its material properties, derived from the WUFI® material database.
Using the material temperature instead of the air temperature in order to
account for the freeze-thaw events can more accurately describe the
performance of the wall assembly. In literature, various values have been
considered critical for freezing and thawing events. In the current
research, the most conservative approach was adopted, which considers 0
Below a critical degree of saturation (
The selected critical material temperature and moisture content were used to assess the freeze-thaw cycles of the exterior 5 mm of the masonry wall. As mentioned above, even a small number of freeze-thaw cycles can cause significant degradation of the brick wall. Mensinga et al. (2010), Ueno et al. (2013a, c) noticed a degradation in the tested specimens after 6 cycles, while in many cases it took around 20 to 30 cycles. Prepens (1991) and Van Aarle et al. (2015) suggested that damage occurs after the brick is exposed to 25–35 freeze-thaw cycles. Thus, in the current research, the risk categorization of Table 2 was adopted.
Risk for mechanical damage of the masonry wall based on the number of freeze-thaw events.
Climate data for three different decades, i.e., 1960–1969 (referred to as past), 2010–2019 (referred to as present), and 2060–2069 (referred to as future), derived from the MPI-ES-LR_REMO2015 model were used for the examination of the climatic changes occurring throughout the years. A fourth climate file with data derived from the ERA5 for the period 2010–2019 (present) was used in order to examine the accuracy of the climate model data. The climate parameters that mostly affect the hygrothermal performance and the frost damage risk of the building components are presented in Figs. 6 and 7.
The signal of climate change in terms of the air temperature (Fig. 6a) is
an average increase of 1.6
The air temperatures are slightly underestimated in the model data, showing
an average difference of 0.3
According to the climate model data, the air relative humidity remains at the same levels under past, present, and potential future conditions, with an average value of approximately 85 % (Fig. 6b). Within the cold months of the year, the air relative humidity is higher and has average values even above 90 % during January, February, and December. The air relative humidity is overestimated significantly by the climate model data since, according to the ERA5 dataset, its average value is 78 %. According to the ERA5, the monthly average relative humidity remains below 86 %.
Comparison among
In some months, the precipitation levels seem to decrease slightly from past to present conditions and to increase from present to future conditions (Fig. 6c). However, precipitation levels vary considerably among different years in the region so, based on 10-year periods, it is hard to consistently detect any trends and the results might be different based on a continuous data series. Nevertheless, from December to April, the highest precipitation level is projected for the latest period 2060–2069. This agrees with other climate model simulations, indicating increasing wintertime precipitation in southern Norway (Lehtonen et al., 2014; Räisänen and Ylhäisi, 2015). The climate model data significantly overestimate the precipitation, especially during the cold months of the year.
The driving rain constitutes a major source of moisture for the brick and, according to the material response-based index, a source of frost damage risk thereof. Thus, the driving rain on vertical masonry walls facing toward the cardinal and intermediate orientations was calculated and depicted in Fig. 7a. The total driving rain load for each ten-year period was computed and divided by ten to get the annual average (Fig. 7a). The orientations with the highest driving rain load in descending order are the south, southeast, east, southwest, northeast, north, west, and northwest. The driving rain load, as well as the impact of climate change on it, are not significant for the west, northwest, and north orientations, and consequently, the focus was placed on the part of Fig. 7a that refers to the northeast, east, southeast, south, and southwest orientations. For the examined five orientations the driving rain load decreases from the past to the present conditions and increases from the present to the future conditions. The most significant changes are observed for the southeast, south, and southwest orientations. The distribution of the driving rain load to the different orientations based on the climate model data is in accordance with the results derived from the ERA5. In the case of the south and southeast orientations, which are the ones with the greatest exposure to wind-driven rain, the climate model underestimates the driving rain load compared to the ERA5. On the contrary, the driving rain load is overestimated in the case of the east orientation compared to the ERA5. For the rest orientations, minor differences were observed among the data from the climate model and ERA5.
Another important parameter that affects the material temperature and drying is the incident solar radiation. In Fig. 7b the solar radiation incident on vertical walls facing toward the cardinal and intermediate orientations is depicted. The total incident solar radiation for each ten-year period was calculated and divided by ten to get the annual average (Fig. 7b). According to the data derived from the climate model, the orientations with the highest solar radiation load in descending order are the south, southwest, southeast, west, east, northwest, northeast, and north. At this point, it is worth mentioning that in Tønsberg during the cold months of the year, the incident solar radiation is significantly lower than the average values of Fig. 7b. However, it can still contribute to the increase of the material temperature above the thawing threshold value during days with air temperatures below the freezing threshold. The incident solar radiation shows a slight decrease through the years for all examined orientations. Moreover, it is underestimated by the climate model data compared to the ERA5.
The ten-year average of the total annual
First, the frost damage risk of the historic masonry walls was assessed by
counting the 0
Observing the results on a monthly basis, there is a significant increase of
the 0
In another direction, it was observed that due to the existing climate
change-from the past to the present conditions- the number of 0
The month with the greatest number of 0
Reviewing other research results in the same field it was observed that the
average annual number of the 0
Finally, as can be seen in Table 3, the climate model underestimates the
number of 0
Number of freeze-thaw events with freezing occurring at air
temperatures below 0
The number of freeze-thaw events was calculated for the four climate
excitation of the current research under the assumption that freezing occurs
at air temperatures below
According to the data derived from the climate model, there is a decreasing trend of the frost damage risk due to climate change (Table 4), which agrees with the results from Grossi et al. (2007), Sabbioni et al. (2010), Bertolin and Camuffo (2014), Leissner et al. (2015). The impact of climate change is more significant from the past to the present conditions. In contrast to the overall decreasing trend of the frost damage risk over the years, there is one month, February, during which the frost damage risk increases due to climate change (Table 4). That is because during February, under the past climate conditions, almost half of the air temperatures were below the freezing threshold, while very few of them were above the thawing threshold value (Fig. 6a). With the air temperature increase due to climate change, there are more air temperatures hovering around the critical freezing and thawing thresholds under the present and future climate excitations and, thus, higher frost damage risk. This observation is in accordance with other studies that suggest an increased frost damage risk over the years in cases of very cold climates (Viles, 2002; Brimblecombe and Camuffo, 2003; Pakkala et al., 2014).
Moreover, freeze-thaw events were observed for the months of March and April under the past climate conditions (Table 4). Considering the present climate conditions there is not any freeze-thaw event during April, while under the future climate excitation, there is no frost damage risk for both March and April. This observation is linked to the air temperature increase due to climate change. Finally, the month with the highest number of freeze-thaw events is, in all examined cases, January.
The impact assessment of climate change on the frost damage risk of cultural
heritage buildings was also investigated by Bertolin and Camuffo (2014) for
the whole of Europe based on outdoor air temperature data. Bertolin and
Camuffo (2014) accounted for the average conditions over the past
(1961–1990), near future (2021–2050), and far future (2071–2100) under two
different moderate emission scenarios, A1B from IPCC 4th assessment
report (IPCC, 2007) and RCP4.5 from the IPCC 5th assessment report
(IPCC, 2014). The number of freeze-thaw events depicted on the risk maps by
Bertolin and Camuffo (2014) is in the same order of magnitude as the results
presented in Table 4. The differences in the number of freeze-thaw events
calculated by Bertolin and Camuffo (2014) and Table 4 can be attributed to
the different threshold values for the freezing and thawing events, periods
of examination, climate models, and climate change scenarios. Moreover,
Bertolin and Camuffo (2014), Loli and Bertolin (2018) implemented
hygrothermal simulations to calculate the indoor climate in various generic
buildings and assessed the frost damage risk based on the indoor air
temperature and the
The number of freeze-thaw events, calculated on the basis that freezing
occurs at air temperatures below
Number of freeze-thaw events with freezing occurring at air
temperatures below
In the current section, the results of the material response-based index for
the case that the indoor air temperature and relative humidity are equal to
the outdoor ones are presented and discussed. In Fig. 8 (panel a) the number of
events during which the material temperature drops below 0
In Fig. 8a it can be seen that there is a very high number of events in
which the material temperature drops below 0
This is attributed to the fact that the moisture content of the layer of
interest remains at lower levels than the critical value of 77.5 kg m
According to the data derived from the climate model, the number of events
during which the material temperature drops below 0
The average moisture content of the layer of interest during periods with
material temperatures below 0
The data derived from the climate model underestimate the 0
The southeast-oriented wall which is the worst-case scenario experienced 187 freeze-thaw cycles during 2010–2019 according to the climate data from ERA5 (Figs. 8c and 9). This number of events corresponds to significant degradation risk with observable results according to Table 2. The same categorization is attributed to the northeast, east, south, and southwest-orientated walls (Fig. 8c). The simulation results refer to a 360 mm masonry wall with one side exposed to the outdoor climate (including all parameters mentioned in Sect. 2.1) and the other side exposed to the outdoor air temperature and relative humidity. The simulation results are in accordance with the on-site observations of a masonry fence (Fig. 2) in which extended damage was observed.
A more detailed analysis of the orientation with the highest frost damage risk is provided in Table 5. January is the month with the highest frost damage risk under all four climate excitations. The decreasing trend of the frost damage risk is more intense from the past to the present conditions than from the present to the future ones. There is no frost damage risk during the months of April and October, while according to the signal of climate change, the frost damage risk during the month of March will be very small in the future. An increase in the frost damage risk for the months of January and November is predicted for the future which is mainly attributed to the projected precipitation increase.
Hourly data of temperature at the control point 2.5 mm in depth
from the outer side of the simulated wall and moisture content at the 5 mm layer at the outer side of the wall. The critical temperature, 0
Number of freeze-thaw events of a 5 mm layer at the exterior side
of the southeast-oriented masonry wall. Freezing occurs at a material
temperature below 0
In the following section, the results of the material response-based index
for the case that the indoor air temperature and relative humidity were
calculated according to EN 15026 are presented and discussed. In Fig. 10
(panel a) the number of events during which the material temperature drops below 0
A very significant number of material temperature drops below 0
It is also observed that the average moisture content during periods with
below 0
According to the data derived from the climate model, the number of events
during which the material temperature drops below 0
The data derived from the climate model underestimate the 0
The southeast-oriented wall was selected as in Sect. 3.4.1 in order to be
examined in more detail. As can be seen in Figs. 10c and 11 it experiences
8 freeze-thaw cycles during 2010–2019 according to the climate data from
ERA5, which is significantly less than the 187 freeze-thaw cycles of Fig. 9. The calculated number of freeze-thaw events corresponds to “damage
likely” according to Table 2. It is worth mentioning that the frost damage
risk level for all the examined orientations ranges from “no damage” to
“damage likely” (Fig. 10c, Table 2). The simulation results refer to a 360 mm masonry wall with one side exposed to the outdoor climate (including all
parameters mentioned in Sect. 2.1) and the other side exposed to an indoor
climate calculated according to EN 15026. The indoor air temperature ranges
from 20 to 25
A month-by-month analysis of the southeast orientation according the ERA5 data is provided in Table 6. January is the month with the highest frost damage risk under all four climate excitations. There is a decrease of the frost damage risk from the past to the present conditions and an increase from the present to the future ones which is linked with the respective changes in the driving rain load. There is no frost damage risk during the months of April, October, November, and December. Based on the signal of climate change it is possible that in the future February and March will also be months with no frost damage risk, while an increase in the number of freeze-thaw cycles during December is also possible.
Hourly data of temperature at the control point 2.5 mm in depth
from the outer side of the wall with southeast orientation and moisture
content at the 5 mm layer at the outer side of the same wall. The critical
temperature, 0
Number of freeze-thaw events of a 5 mm layer at the exterior side
of the southeast-oriented masonry wall. Freezing occurs at a material
temperature below 0
Brick is a building material vulnerable to frost damage. As a result of the action of this deterioration mechanism the material surface flakes off and crumbles away. The main cause is the pressure exerted when the water inside the material pores freezes to ice and increases its volume. Tønsberg is a city in southern Norway that has a significant number of historic masonry buildings which are affected by this deterioration mechanism. The warmer and more humid climate during the cold months of the year, which is attributed to climate change, is expected to affect the action of this deterioration mechanism.
In the current research, climate data from a climate model under past,
present, and future conditions were used in order to account for the signal
of climate change. An additional climate file for the present conditions
derived from the ERA5 was used to assess the accuracy of the climate model
results. Two climate-based indices and one material response-based index
were used for the assessment of the frost damage risk. The first climate
index accounts for the number of 0
The material temperature and moisture content were calculated by simulating the hygrothermal performance of a 360 mm masonry wall under the four mentioned outdoor climate excitations. The indoor climate was in one case considered equal to the outdoor one – worst-case scenario –, while in the second case it was calculated according to EN 15026 – best-case scenario –.
Significant differences were observed in (i) the number of freeze-thaw events, (ii) the periods during which freeze-thaw events occur and (iii) the climate change-related trends for the different indices.
The overall impact of climate change is a decrease in the frost damage risk of the masonry walls. The only exception was the east, northeast, and southeast-oriented walls of the simulated small, conditioned room – with indoor climate calculated according to EN 15026 – for which an increased frost damage risk was predicted from the present to the future conditions. At this point, it is worth highlighting that the degradation by this deterioration mechanism is cumulative and despite the overall decreasing trend of the number of freeze-thaw events, the actual damage on the masonry walls increases through the years.
Moreover, a decrease in the number of months with freeze-thaw events is observed due to climate change. This observation is linked to the climate change-induced air temperature increase.
In contrast with the overall trend, there are certain months of the year for which the frost damage risk increases. Based on the results of the climate-based indices these are the coldest months of the year, like February, which due to the climate change-induced temperature increase experience more incidents with air temperatures hovering around the freezing and thawing thresholds. According to the material response-based index, an increased frost damage risk was observed from the present to the future conditions for the months and orientations with an increased driving rain load due to climate change. Such an example is January for the southeast-oriented masonry walls. At this point, it is also worth highlighting that, firstly, the driving rain and, secondly, the solar radiation gains are the parameters that define the orientation with the maximum frost damage risk. The frost damage risk of the masonry walls varies significantly among the different orientations and, as described above, so does the signal of climate change.
In the case of the material response-based index, the indoor climate has a very significant role in the frost damage risk assessment. The freeze-thaw cycles calculated for the worst-case scenario were even 20 times more than the ones for the best-case scenario. This is mainly attributed – according to the best-case scenario – to the greater capacity of the indoor air to store moisture from the material. In this way, during freezing temperatures, there is not enough moisture in the material pores to freeze, swell, and cause damage.
In another direction, it was observed that the climate model underestimates the frost damage risk compared to the ERA5 according to the climate-based indices. This is attributed to the underestimation of the air temperature by the climate model data compared to the ERA5.
On the other hand, the climate model in most of the examined cases overestimated the frost damage risk according to the material response-based index. This is mainly linked to the overestimation of the precipitation and, consequently, the driving rain load and material moisture content by the climate model compared to the ERA5. The only exception to this observation was the south-oriented wall of a room with an indoor climate calculated according to EN 15026. It was, also, observed that the differences in the calculated frost damage risk between the data from the climate model and ERA5 were more significant for the orientations with high driving rain load and low solar radiation gains.
In addition, of great importance is the fact that the outputs of the material response-based index were translated into certain damage categories. This allows the validation of the simulation results, the prediction of the point at which a critical level of damage will be reached, and also makes the results understandable to a wider audience. Additional experimental data would help to define the critical values more accurately for the various risk levels.
Finally, future research could focus on improving certain limitations of the current study and, specifically, (i) include data from more climate models and climate change scenarios and bias correct the data (Gaur et al., 2019; Gaur and Lacasse, 2022), (ii) use climate data until the far future, (iii) collect information about the geometry, materials, climate control, and moisture loads of case studies, (iv) take into account the depression of the freezing point due to the pore size of the material and the presence of salts, (v) leverage experimental data for the critical number of freeze-thaw events (Germinario et al., 2022; Salvini et al., 2022), (vi) account for more deterioration mechanism, like salt crystallization. In this way, it will be possible to investigate and quantify even more accurately the degradation of cultural heritage buildings and, finally, organize proper strategies to improve their preservation and resilience.
The climate data that support the findings of this study are derived from
the online databases of the Copernicus climate change service (
Conceptualization, PC; data curation, PC; formal analysis, PC; investigation, PC; methodology, PC; project administration, PC; resources, PC; software, PC; validation, PC; visualization, PC; writing – original draft preparation, PC; writing – review & editing, PC, GBAC and DK; funding acquisition, PC and DK. All authors have read and agreed to the published version of the manuscript.
The contact author has declared that neither of the authors has any competing interests.
Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This article is part of the special issue “European Geosciences Union General Assembly 2022, EGU Division Energy, Resources & Environment (ERE)”. It is a result of the EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022.
This work is a part of the HYPERION project. HYPERION has received funding from the European Union's Framework Program for Research and Innovation (Horizon 2020). The content of this publication is the sole responsibility of Oslo Metropolitan University and does not necessarily reflect the opinion of the European Union. The authors would also like to acknowledge Ilari Lehtonen (Finish Meteorological Institute) for providing information and guidelines for the use of the climate data of the current research, Jørgen Solstad (Vestfold and Telemark County Council) for providing information about the historic masonry buildings in the city of Tønsberg, and Claudio Mazzoli (Department of Geosciences, University of Padova) for providing information on the frost damage of porous materials.
This research has been supported by the Horizon 2020 (grant no. HYPERION (821054)).
This paper was edited by Michael Kühn and reviewed by three anonymous referees.