Evaluation of Natural Stone Weathering in Heritage Building by Infrared Thermography

: The application of non-contact diagnostic methodologies is the current challenge in the frame of the cultural heritage, referred to as preservation, monitoring and restoration. Inspired by the potential shown by infrared thermography in rock mechanics’ non-destructive applications, this paper presents the results achieved by its use for the quick survey of different weathering types affecting natural stones at historical buildings. Infrared thermography allowed recognizing and mapping the different surface temperatures arising from the presence of efﬂorescence, subﬂorescence, alveolization, black crusts and bioweathering at limestone and basalt stones. Infrared data were sided by photogrammetric three-dimensional models of surveyed spots, which provided quantitative data on the thickness of rock affected by mechanical weathering, and key correspondence between the two techniques is highlighted. Achieved results show that infrared outcomes are related to different aspects primarily involving the stone face morphology and color, as well as the environmental conditions at the surveying time. Provided interpretations were validated by ﬁeld visual inspections, which conﬁrmed the good potential of infrared thermography as a quick weathering diagnostic tool. This study can be therefore considered a starting reference for knowledge development in this scientiﬁc ﬁeld.


Introduction
Time and atmospheric agents (rain, wind, solar radiation, aggressive atmospheric pollutants, freeze-thaw cycles, crystallization of saline solutions, and growth of organisms) are the main causes of the deterioration of stone materials exposed to external conditions and employed as either construction material or decoration, e.g., [1][2][3]. Despite the various treatments aimed at preserving the appearance of stones, their weathering represents a serious problem, especially in the preservation of cultural heritage. In particular, the mineralogical composition of rocks, along with their physical properties (with specific regard to porosity), play a key control in the kinetics of degradation processes [3][4][5]. In the last decade, different non-destructive technologies have been employed to study the deterioration affecting stones at historical monuments and buildings [6][7][8]. Casula et al. [9] highlighted the presence of cracks and micro-cracks in masonry through the combination of laser and ultrasound scanning techniques, while Avdelidis and Moropoulou [10] used Infrared Thermography (IRT) to detect surface cracks, detachments, differences in material or the presence of moisture within the structures. Nevertheless, the process that leads to the analysis of the problem and to the possible solutions for the restoration and/or conservation of the stone is often long and based on a preliminary visual analysis. Moreover, visual inspections alone may suffer from a non-complete characterization of the problem due to the impossibility of detecting the weathering progress not yet visible to the naked eye. From this perspective, this paper aims to test the potential of IRT applied for the preliminary evaluation of weathering phenomena affecting natural stones, even in combination with photogrammetry and in situ inspections. The use of IRT on rocks is developed in engineering geological studies for different purposes, such as the laboratory indirect estimation of porosity and the evaluation of the pore persistence [11,12], while in the field, it is widely employed for the characterization of rock masses, e.g., [13][14][15][16]. Its use for stone characterization in cultural heritage is still under development, and this research provides a contribution to this field. To test and validate the application of this methodology, IRT surveys were carried out at a historical building located in Catania town (southern Italy). It is Palazzo delle Scienze (PdS), one of the headquarters of the University of Catania, which was built between 1935 and 1942 ( Figure 1). As for many other historical edifices built in the late Baroque in the town, its architectural style involves the use of two varieties of rocks: a light-beige limestone and a dark grey igneous rock. Both rock types were largely available at local quarries located at the Hyblean mountains (south of Catania) and Mount Etna. Representative study spots affected by different types of weathering were selected along the external walls and underwent a weathering characterization by means of in situ surveys, photogrammetric 3D models and IRT. Results were compared, aiming to understand the infrared thermal expression of selected weathering types so as to provide an innovative approach for the quick preliminary study of weathering through alternative non-contact methodologies.

Materials and Methods
The stones employed as construction material in the ancient city center of Catania during the eighteenth and nineteenth centuries are both igneous and sedimentary rock types. Their combined use gives rise to a characteristic light-dark duotone in the façades of the buildings [17]. The sedimentary ones are carbonate rocks (mainly limestones), quarried on the Hyblean mountains (south-eastern Sicily) and belonging to different geological formations, while igneous rocks belong to Mount Etna historical or prehistorical lava flows, which can be classified as trachibasalts and trachiandesites, e.g., [18,19], with a porphyritic texture and a mineral composition usually represented by plagioclase, pyroxene and olivine in a crystallized groundmass rarely vitrophyric [12]. These are, from now on, generically referred to as "basalts" and can be grouped into two categories according to their texture: massive and vesicular [20]. The methodological approach followed for this research is focused on the laboratory characterization of studied rocks and on the application of remote technologies for the rock weathering identification.

Laboratory Physical and Mechanical Rock Characterization
Representative specimens of both rock varieties were tested in laboratory with the aim of estimating their main engineering geological properties. Specimens were 70 mm (±5) cubes, and physical-mechanical tests were carried out according to the EN1926 and EN1936 specifications. In particular, imbibition coefficient (ic), bulk density (γ b ) and total porosity (n) were calculated based on the rock weights in dry (m dry ), saturated (m sat ) and hydrostatic (m w ) conditions, according to Equations (1)-(3) (where ρ is the water density and γ r is the rock real density calculated according to the pycnometer method).
Uniaxial Compressive Strength (UCS) and elastic modulus (E) were estimated through compression tests performed by using an automatic 3000 kN hydraulic press machine connected to an E Modulus console for the determination of the elastic modulus.

IRT Survey
The IRT technique is based on the principle that every form of matter with temperature above absolute zero emits thermal radiation, whose wavelength mainly falls within the infrared band of the electromagnetic spectrum [21]. This emitted radiation is proportional to the temperature of the emitting body, according to the Stefan-Boltzmann law. Therefore, the surface temperature of an object can be estimated by specific devices called thermal cameras and operating within the range of wavelengths as long as 13-14 µm. Due to the fact that thermal radiation is mainly invisible to the naked eye, IRT represents a useful tool to extend human perception even toward the detection of otherwise invisible phenomena. Thanks to its potential, IRT is currently used in numerous activities and scientific branches also involving the characterization of rocks. In particular, it is applied to the study of rock fracturing in rock masses, e.g., [16,22,23], or for the characterization of intact rock properties and related variations under specific conditions, e.g., [12,[24][25][26][27]. In the frame of cultural heritage preservation, IRT has already proved useful in detecting biocolonization on natural stones [28], studying sulfated crusts [29] and, more in general, for investigation of chemical degradation processes in historical buildings, e.g., [30][31][32].
In this paper, IRT images were acquired at selected spots of the PdS building with the aim of framing different weathering types and of shedding light on the corresponding thermal aspect. Surveys were carried out in the absence of direct solar radiation, according to literature experiences [16], during summer and under no rain conditions, by using a high-sensitivity thermal camera with a 320 × 240 pixels infrared resolution and a temperature range between −20 and 650 • C (with ±2 • C accuracy).
Emissivity value was set according to literature data referring to the same rock types [33], namely 0.94 and 0.95 for basalts and limestones, respectively. Thermal images were post processed to highlight specific features according to the different observed weathering types, and results were validated by field inspections.

Photogrammetric Survey
A photogrammetric digital reconstruction of the building elements surveyed by IRT was carried out with the aim of supporting the analysis presented herein. In particular, 3D models are widely used in the field of cultural heritage [34][35][36] to generate dense cloud for the analysis of digital models of structural elements. For this study, the acquisition was carried out through photogrammetry, which already showed good results in this field, e.g., [37][38][39]. Digital images were acquired by a Nikon Z62 digital mirrored camera with a CMOS (Complementary Metal Oxide Semiconductor) sensor of a size (35.9 mm × 24.9 mm) and a resolution of 6048 × 4024 px, equipped with a Nikon 24-70 mm lens (Nikon, Tokyo, Japan). Dense point clouds were then postprocessed by a specific tool implemented in the CloudCompare computer program, with the aim of achieving a Digital Elevation Model (DEM) of the analyzed structural elements. This was specifically useful for the study of the stone morphological condition under a specific weathering type (alveolization), with specific reference to the quantitative evaluation of the weathering depth.

Hyblean Limestones
Fine-grained limestone quarried at Hyblean Mountains (Figure 2a,b) shows a porous texture given by millimetric voids. Their total porosity, generally related to primary processes, is, on average, around 37.65%, and most of it is represented by interconnected voids, which confer the rock an average effective porosity of around 27%. Their bulk density ranges from 1501 to 1850 kg/m 3 , and the degree of compactness, expressed by the ratio between real and bulk densities, is between 1.43 and 1.79 (Table 1). In this case, real density was calculated according to the pycnometer method ( Figure 2c). Uniaxial compression tests ( Figure 2d) highlighted a wide range of strength values (Table 1). In particular, the most porous specimens are characterized by lower UCS and E values, suggesting a control played by porosity on the mechanical behavior of these rocks. Such properties are statistically well correlated, with specific reference to UCS, E and porosity. In particular, by plotting UCS against porosity (Figure 2e), a negative linear statistical correlation with a 0.81 determination coefficient (r 2 ) is outlined. It proves the strong dependence between the rock strength and porosity, according to several literature experiences, e.g., [40,41]. On the other hand, UCS and E are statistically correlated by a positive linear relation (r 2 0.75) (Figure 2f), suggesting that the highest strength corresponds to lower deformability of intact rock, similarly to other published experiences, e.g., [42,43].  Figure 2. (a,b) Limestone blocks at the sampling quarry; (c,d) a photo taken during the pycnometer analysis for real density estimation process; (e) representative basalt specimen in the hydraulic press before uniaxial compression test; (f) statistical regression between UCS and total porosity; (g) statistical regression between UCS and E.

Etna Basalts
On the other hand, massive basalt samples were taken at a quarry located on Mount Etna (Figure 3a) and show a macroscopic massive structure with no visible voids (Figure 3b). Their laboratory characterization (Figure 3c,d) returned good mechanical properties, with UCS mean values around 210 MPa. Literature reports a wide statistical range of this parameter in basalts, which is often influenced by specific rock features, such as the presence of cracks [12,20]. The total porosity of tested rocks is low (on average 4.7%), and voids are represented by both microvoids, often isolated, and microcracks. Unlike limestones, the degree of compactness of selected basalts is 1.05 (Table 2) due to the massive structure of the rock. Statistical regression analysis of selected parameters returned satisfactory correlations. For example, UCS and porosity are linked by a negative linear law, although tested rocks' porosity is comprised within a narrow statistical range (Figure 3e). A similar consideration can be reported for the relation between UCA and E, affected by a positive linear trend (r 2 0.64) (Figure 3f).

Etna Basalts
On the other hand, massive basalt samples were taken at a quarry located on Mount Etna (Figure 3a) and show a macroscopic massive structure with no visible voids ( Figure  3b). Their laboratory characterization (Figure 3c,d) returned good mechanical properties, with UCS mean values around 210 MPa. Literature reports a wide statistical range of this parameter in basalts, which is often influenced by specific rock features, such as the presence of cracks [12,20]. The total porosity of tested rocks is low (on average 4.7%), and voids are represented by both microvoids, often isolated, and microcracks. Unlike limestones, the degree of compactness of selected basalts is 1.05 (Table 2) due to the massive structure of the rock. Statistical regression analysis of selected parameters returned satisfactory correlations. For example, UCS and porosity are linked by a negative linear law, although tested rocks' porosity is comprised within a narrow statistical range (Figure 3e). A similar consideration can be reported for the relation between UCA and E, affected by a positive linear trend (r 2 0.64) (Figure 3f).

Observed Weathering Types
In this section, results related to the observed forms of weathering along the two lithotypes at PdS are presented. The description provided in this section, differentiated by weathering type, is based on in situ surveys aimed at describing the main features occurring along the selected spots. In particular, four types of weathering phenomena have been observed: (1) efflorescence and subflorescence, (2) alveolization, (3) black crust and (4) bioweathering. The first type was surveyed at both lithotypes, while alveolization, black crusts and bioweathering were reported at limestones.

Efflorescence and Subflorescence
The chemical weathering of rocks is a spontaneous and irreversible process leading the affected material towards a more stable state under a given temperature and pressure conditions [44]. Among the chemical weathering processes, efflorescence is caused by deposited salts crystallizing. After that, a solution migrates through the rock and then evaporates. These minerals are generally sulfates and carbonates of sodium, potassium, or calcium, as well as nitrates and chlorides. In particular, the role of salts in rock weathering has been recognized during the last decades, both in natural areas and in monuments [45][46][47]. The salts act not only as a physical force but also as saline solutions, which facilitates the dissolution rates of silicate minerals [45]. Since this process arises from the migration of water through the rock, its occurrence suggests water infiltration phenomena and related consequences. In fact, saline efflorescence may cause the haloclastic rupture of affected materials, being a serious issue not only in construction and decorative stones, e.g., [48], but also for masonry, underground construction, e.g., [49], and natural environments, e.g., [45]. With reference to natural stones employed as construction material, the overpressure generated by salt crystals causes the exfoliation of the rock through a progressive detachment of the shallowest rock portions. More specifically, the results of salt crystallization and precipitation under the surface of a material causing spalling, sugaring, flaking and/or pitting is acknowledged as subflorescence. Whereas efflorescence is relatively easy to remove, the subflorescence is difficult as it is formed below the rock surface [50].
Among the PdS elements chosen for this study, this weathering form was found at both the igneous rock, used as cladding along the lower part of the façade and as column stereobate along the south-facing prospect of the building, and at limestones occurring along the façade and used as column stylobates. With respect to the igneous rock type, we found some differences according to the building façade exposure: the most affected sector faces north, where the igneous rock spalling and flaking is deeply developed (Figure 4a,b). Evidence of efflorescence was found both at the bared inner rock portion and at the most external weathered and loose rock sheet ( Figure 4a); in this case, the rock thickness affected by mechanical and chemical weathering is maximum. Along the west-facing façade, evident signs of efflorescence are visible at the igneous rock strip also bordering the stairway of a secondary entrance to the building, from now on referred to as STW. This structural element protrudes with respect to the PdS main structure, being more exposed to the atmospheric weathering agents (Figure 4c). In this spot, we could also evaluate a sort of evolution of the mechanical rock flaking, testified by different layers of flaked rock sheets that we numbered according to progressive orders starting from the unflaked rock face (Figure 4d). This testifies to the subflorescence progression within the inner rock through different flaking stages.   The evidence of efflorescence and subflorescence is also reported along the colonnade, with specific reference to the basaltic stereobates (Figure 4e,f), where there is an advanced state of deterioration (flaking order II). Signs of efflorescence occur within the studied element, both at the exfoliated rock portions and at unflaked spots. In this latter case, salt deposits are also found aligned along borders, representing the potential detachment margin of the ongoing rock flaking.
Efflorescence phenomena also affect the limestone rock type, although in this case, it is combined with other weathering types, such as alveolization, which is discussed in the following sections ( Figure 5).
Heritage 2022, 5, FOR PEER REVIEW 11 Efflorescence phenomena also affect the limestone rock type, although in this case, it is combined with other weathering types, such as alveolization, which is discussed in the following sections ( Figure 5).

Alveolar Weathering
Alveolar or honeycomb weathering is a kind of differential weathering phenomenon developing as a result of a number of different weathering processes involving soluble salts in most cases [51], granular disintegration and/or scaling. Its expression on natural stones is the widespread presence of visible cavities (alveoli) variable in shape and size. In particular, pressures generated by salt crystallization, arising from supersaturated solutions percolation through the rock, are a major cause of the physicochemical weathering of rock structures [52] and references therein. This phenomenon is stressed by the abrasive effect played by rock incoherent grains, taken over by turbulent air currents, which contribute to enlarging the cavity size.
In this case study, alveolar weathering is well developed in the limestone rocks used for both the upper façade and for the colonnade. Due to their porosity grade and low mechanical strength, this lithotype is a common candidate for this weathering process, as testified by literature studies on rocks belonging to the same geological setting, e.g., [53,54], and, more generally, on carbonate rock types worldwide, e.g., [51,55,56]. Nevertheless, the presence of alveoli is localized to specific portions of both the façade and columns. In the first case, they are found immediately above the basal band of basalt, with specific reference to STW (Figure 5a), while in columns, they occur at the stylobate ( Figure  5b,c). In both cases, the underlying basalt stone is more protruding, thus offering a basis for the bounce of rain water droplets, which enhances the mechanical limestone disaggregation. The evolution of such weathering process can be recognized by the morphological definition of alveoli, which tend to coalesce as the weathering develops, leading to the formation of wide deteriorated sectors (Figure 5c).

Alveolar Weathering
Alveolar or honeycomb weathering is a kind of differential weathering phenomenon developing as a result of a number of different weathering processes involving soluble salts in most cases [51], granular disintegration and/or scaling. Its expression on natural stones is the widespread presence of visible cavities (alveoli) variable in shape and size. In particular, pressures generated by salt crystallization, arising from supersaturated solutions percolation through the rock, are a major cause of the physicochemical weathering of rock structures [52] and references therein. This phenomenon is stressed by the abrasive effect played by rock incoherent grains, taken over by turbulent air currents, which contribute to enlarging the cavity size.
In this case study, alveolar weathering is well developed in the limestone rocks used for both the upper façade and for the colonnade. Due to their porosity grade and low mechanical strength, this lithotype is a common candidate for this weathering process, as testified by literature studies on rocks belonging to the same geological setting, e.g., [53,54], and, more generally, on carbonate rock types worldwide, e.g., [51,55,56]. Nevertheless, the presence of alveoli is localized to specific portions of both the façade and columns. In the first case, they are found immediately above the basal band of basalt, with specific reference to STW (Figure 5a), while in columns, they occur at the stylobate (Figure 5b,c). In both cases, the underlying basalt stone is more protruding, thus offering a basis for the bounce of rain water droplets, which enhances the mechanical limestone disaggregation. The evolution of such weathering process can be recognized by the morphological definition of alveoli, which tend to coalesce as the weathering develops, leading to the formation of wide deteriorated sectors (Figure 5c).

Weathering from Air Pollution and Biological Activity
The weathering arising from air pollution on carbonate stones is expressed by black crusts. Their formation originates from a sulfation reaction (SO 2 reacts with calcite to form gypsum), and the entrapment of particles (especially soot) causes its blackening [57][58][59]. This phenomenon is well acknowledged in the literature, as it largely affects heritage in urban environments thanks to the capability of atmospheric pollutants to form compounds that can interact with building materials. The degradation products are visible on calcareous building materials, e.g., [60,61], especially in sheltered sectors, preserved from washing action. Gypsum, indeed, is soluble in water, and it is usually washed away. Black crust evolution leads to a progressive thickening and a rock porosity reduction. Moreover, the black color favors heat absorption more than the surrounding light color rock, enhancing mechanical dilatation phenomena, which progressively weaken the mechanical attitude of the rock itself until its gradual disintegration. At PdS, black crusts were surveyed at the upper portion of STW (Figure 6a), at the hollow portion between the column toruses, and at some sheltered façade spots, such as those below the window sills (Figure 6b).

Weathering from Air Pollution and Biological Activity
The weathering arising from air pollution on carbonate stones is expressed by black crusts. Their formation originates from a sulfation reaction (SO2 reacts with calcite to form gypsum), and the entrapment of particles (especially soot) causes its blackening [57][58][59]. This phenomenon is well acknowledged in the literature, as it largely affects heritage in urban environments thanks to the capability of atmospheric pollutants to form compounds that can interact with building materials. The degradation products are visible on calcareous building materials, e.g., [60,61], especially in sheltered sectors, preserved from washing action. Gypsum, indeed, is soluble in water, and it is usually washed away. Black crust evolution leads to a progressive thickening and a rock porosity reduction. Moreover, the black color favors heat absorption more than the surrounding light color rock, enhancing mechanical dilatation phenomena, which progressively weaken the mechanical attitude of the rock itself until its gradual disintegration. At PdS, black crusts were surveyed at the upper portion of STW (Figure 6a), at the hollow portion between the column toruses, and at some sheltered façade spots, such as those below the window sills ( Figure  6b). Bioweathering, on the other hand, occurs when the employed stone becomes subject to colonization by microorganisms [62]. Among the biodeteriogens, the most widespread are microscopic algae and lichens, which may implant on porous or deteriorated rock surfaces, penetrating between already detached fragments. On light-color rocks, their colony occurs as a darker film, especially localized in damp sectors. Evidence of bioweathering was surveyed at STW, with specific reference both to the handrail, exposed to wetting and absorption of meteoric water, and to a localized spot along the vertical limestone wall subject to water flow (Figure 6a).

IRT Survey
In this section, results arising from IRT application to the survey of the afore-described weathering types are presented.

IRT Applied to Efflorescence and Subflorescence
The infrared images related to the basaltic rocks taken into account herein show that efflorescence phenomena can be detected by IRT due to a different surface temperature owned by affected sectors with respect to the surrounding rock portions. With reference to column stereobates, the analysis of IRT images highlighted a surface temperature varying according to specific features occurring within the framed area. In column III ( Figure  7a), for example, surface temperature ranges from 32 to 35 °C suggesting a sort of thermal Bioweathering, on the other hand, occurs when the employed stone becomes subject to colonization by microorganisms [62]. Among the biodeteriogens, the most widespread are microscopic algae and lichens, which may implant on porous or deteriorated rock surfaces, penetrating between already detached fragments. On light-color rocks, their colony occurs as a darker film, especially localized in damp sectors. Evidence of bioweathering was surveyed at STW, with specific reference both to the handrail, exposed to wetting and absorption of meteoric water, and to a localized spot along the vertical limestone wall subject to water flow (Figure 6a).

IRT Survey
In this section, results arising from IRT application to the survey of the afore-described weathering types are presented.

IRT Applied to Efflorescence and Subflorescence
The infrared images related to the basaltic rocks taken into account herein show that efflorescence phenomena can be detected by IRT due to a different surface temperature owned by affected sectors with respect to the surrounding rock portions. With reference to column stereobates, the analysis of IRT images highlighted a surface temperature varying according to specific features occurring within the framed area. In column III (Figure 7a), for example, surface temperature ranges from 32 to 35 • C suggesting a sort of thermal non-uniformity of the rock, even considering that no direct radiation was affecting the framed elements at the time of the survey. Starting from a general overview (Figure 7a), a clear distinction between warmer and colder sectors can be described. Warmer sectors are found where the igneous rock shows a lower degree of weathering, namely at those exposed rock portions that are not affected by efflorescence and which could have also been bared by flaking phenomena (Figure 7b). Contrarywise, colder temperatures are found at the loose rock portions, which have already been partially detached by subflorescence phenomena, and are represented by rock slices more exposed to environmental cooling (Figure 7c). The correspondence between such thermal outcomes and the state of the rock material was verified in the field. In fact, the loose rock slice occurring at the upper-left side of the column III stereobate is recognizable even to the naked eye. Nevertheless, by observing the thermal image (Figure 7a-c), the cold area is much wider than the visible weathered rock slice. This suggests that the rock portion labeled by the negative thermal anomaly (cold region) has already been detached by its sub-stratum. This should be regarded as a flaking expansion area in the frame of weathering evolution monitoring or forecasting. Further cold sectors are found at the upper stereobate corners, with specific evidence on the right side. Although the cold corner is also an effect of the rock cooling process, as suggested by literature studies [63], the cold anomaly labeling the right edge of the stereobate could be thought of as an index of a loss of adhesion of the rock to the substratum; this phenomenon occurs, even though with a slighter thermal anomaly, also at the opposite side. Finally, intermediate surface temperatures (Figure 7d) correspond to rock portions affected by a slight saline crystallization film, providing an indication of the progression of the considered weathering phenomenon. For column II (Figure 7e), similar considerations can be provided. In particular, a thermal contrast between the upper and lower rock sector is outlined, highlighting that weathering is heavier at the upper stereobate portion. More specifically, by filtering the image considering only the highest surface temperature range (Figure 7f), indicating the sound rock, it is possible to state that the stone degradation progresses from top to bottom. The lowest surface temperatures were found at the partially flaked rock slices, and the flaking expansion sectors can be recognized (Figure 7g). Intermediate surface temperatures (Figure 7h) show that the deposition of salt crystals is widespread, with a specific concentration at the hollow rock surface parts, which arise from the bush hammered finish of the stone. Moreover, signs of efflorescence are found at both the rock bared by occurred flaking and at the portions surrounding the stone slices still in place.
By analyzing the basaltic cover around STW (Figure 8a), the above considerations find a positive correspondence. In particular, the rock flaking process can be recognized due to the thermal contrast occurring between a sound rock (warm surface temperatures) and detached slice (cold surface temperatures) (Figure 8b). The colder region allows also bordering the flaking expansion area, which cannot be easily mapped only by a naked eye observation. Moreover, the wide efflorescence spots located along the basaltic wall hold the lowest recorded temperatures, suggesting a certain thickness of deposited salts (Figure 8c). This feature can be observed even to the naked eye, although IRT image shows that the colder regions are not related only to the presence of salts but also to the occurrence of moisture originating from the contact fissures between stone slabs and with a clear ascent path from the bottom to the top of the stone strip. Even in this case, the intermediate temperatures are related to those rock portions hosting even slight efflorescence signs. A further note is that the negative, horizontal, linear anomaly occurring between the two orders of basaltic strips is only related to the fact that the lower order is more protruding than the one above, thus providing a quicker cooling of the specific rock portion.  clear ascent path from the bottom to the top of the stone strip. Even in this case, the intermediate temperatures are related to those rock portions hosting even slight efflorescence signs. A further note is that the negative, horizontal, linear anomaly occurring between the two orders of basaltic strips is only related to the fact that the lower order is more protruding than the one above, thus providing a quicker cooling of the specific rock portion.

IRT Applied Alveolization
For the infrared study of alveolization, IRT analysis was sided by the 3D photogrammetric evaluation of the surface stone morphology. In particular, three-dimensional digital models of column stylobates and STW were analyzed. As previously stated, alveolar weathering affects the limestone rock portions immediately overlying a protruding element, which in this case, is represented by basaltic rock. All the column stylobates are affected by such weathering process, as testified by both visual inspection and the 3D model. In particular, the alveoli presence prevails in the central portion of the stylobate, and their maximum horizontal depth is around 1.4 cm (Figure 9). The IRT expression of hollow rock portions, when framed under no direct irradiation, is a warm surface temperature, according to literature experiences, e.g., [64][65][66]. This is because the hollow morphology prevents the rock from cooling if compared to the surrounding rock portions more exposed to the external environment. In this case, IRT images highlight that alveoli are labeled by the warmest temperatures, retracing the cavities highlighted by the 3D model, although when a salt crystallization film occurs within the alveolus, its surface temperature is lowered (Figure 9). Moreover, if subflorescence occurs enough to lead to detachment of rock slices, the rock surface temperature is even lower (Figure 9). Moisture at the upper corners of stylobates is present even in this case, as previously observed at the basaltic stereobates, and it affects the whole upper part of the structural element, thus suggesting the upper origin of water infiltration.
model. In particular, the alveoli presence prevails in the central portion of the stylobate, and their maximum horizontal depth is around 1.4 cm (Figure 9). The IRT expression of hollow rock portions, when framed under no direct irradiation, is a warm surface temperature, according to literature experiences, e.g., [64][65][66]. This is because the hollow morphology prevents the rock from cooling if compared to the surrounding rock portions more exposed to the external environment. In this case, IRT images highlight that alveoli are labeled by the warmest temperatures, retracing the cavities highlighted by the 3D model, although when a salt crystallization film occurs within the alveolus, its surface temperature is lowered (Figure 9). Moreover, if subflorescence occurs enough to lead to detachment of rock slices, the rock surface temperature is even lower (Figure 9). Moisture at the upper corners of stylobates is present even in this case, as previously observed at the basaltic stereobates, and it affects the whole upper part of the structural element, thus suggesting the upper origin of water infiltration. At STW, the heaviest weathering is represented by alveolization and efflorescence, with consequent flaking due to subflorescence (Figure 10a). In particular, alveolar weathering shows signs of an advanced process, with a heavy mechanical disaggregation of rocks. In fact, wide eroded areas, rather than a purely alveolar structure, locally occur with a developed alveolization and efflorescence, with a maximum horizontal depth of about 2 cm (Figure 10b). This affected part is located over the basaltic strip, which favors, as previously stated, the rain droplet rebound, thus enhancing the mechanical rock weathering. The thermal image of this portion (Figure 10c) shows a prevalence of low surface temperatures, with the lowest values (about 29.5 °C) found where alveolization and efflorescence coexist. On the other hand, alveoli alone are characterized by the highest surface temperature (30.5-31.0 °C) due to their hollow morphology (Figure 10b), according to At STW, the heaviest weathering is represented by alveolization and efflorescence, with consequent flaking due to subflorescence (Figure 10a). In particular, alveolar weathering shows signs of an advanced process, with a heavy mechanical disaggregation of rocks. In fact, wide eroded areas, rather than a purely alveolar structure, locally occur with a developed alveolization and efflorescence, with a maximum horizontal depth of about 2 cm (Figure 10b). This affected part is located over the basaltic strip, which favors, as previously stated, the rain droplet rebound, thus enhancing the mechanical rock weathering. The thermal image of this portion (Figure 10c) shows a prevalence of low surface temperatures, with the lowest values (about 29.5 • C) found where alveolization and efflorescence coexist. On the other hand, alveoli alone are characterized by the highest surface temperature (30.5-31.0 • C) due to their hollow morphology (Figure 10b), according to what was observed in columns. In this case, flaking portions have a slightly higher surface temperature than the surrounding area with efflorescence.
The weathering process evolves upwards into efflorescence (surface temperature between 29.7 • C and 29.9 • C), while the highest temperatures found at the upper STW portion are related to other weathering types, namely black crusts and bioweathering, which are discussed in detail in the following section.
Heritage 2022, 5, FOR PEER REVIEW 17 what was observed in columns. In this case, flaking portions have a slightly higher surface temperature than the surrounding area with efflorescence. The weathering process evolves upwards into efflorescence (surface temperature between 29.7 °C and 29.9 °C), while the highest temperatures found at the upper STW portion are related to other weathering types, namely black crusts and bioweathering, which are discussed in detail in the following section.

IRT Applied to Black Crust and Bioweathering
The post-processing of infrared images acquired at STW and at a selected spot of the PdS façade allowed carrying out some consideration even with respect to the presence of black crust and bioweathering. In particular, the occurrence of black crust can be identified by IRT thanks to the chromatic difference existing between the crust and the washed limestone, leading to dissimilar heat absorption. This can be well observed by IRT during the rock cooling when images are acquired under no direct solar radiation (such as after sunset or in shadow conditions). In this case, black crusts occurring at the stairway ( Figure  11a) wall are labeled by the highest surface temperatures (30.7-31.0 °C), while the washed, light-colored rock keeps a lower surface temperature (30.0-30.2 °C). Consistent results were achieved by analyzing and comparing the thermal outputs of an exposed façade portion and a close part below a window sill, sheltered from rain washing (Figure 11b). In this case, the warmest temperatures were found where the rock was affected by a black crust cover, while intermediate surface temperatures were found at washed rock and the lowest values at rock parts affected by efflorescence, in accordance with afore provided results. Just a note on Figure 11b: the colder region occurring right below the window sill likely originated from the shadow effect played by the column capital, which prevented this part from direct insolation before the survey.

IRT Applied to Black Crust and Bioweathering
The post-processing of infrared images acquired at STW and at a selected spot of the PdS façade allowed carrying out some consideration even with respect to the presence of black crust and bioweathering. In particular, the occurrence of black crust can be identified by IRT thanks to the chromatic difference existing between the crust and the washed limestone, leading to dissimilar heat absorption. This can be well observed by IRT during the rock cooling when images are acquired under no direct solar radiation (such as after sunset or in shadow conditions). In this case, black crusts occurring at the stairway (Figure 11a Consistent results were achieved by analyzing and comparing the thermal outputs of an exposed façade portion and a close part below a window sill, sheltered from rain washing (Figure 11b). In this case, the warmest temperatures were found where the rock was affected by a black crust cover, while intermediate surface temperatures were found at washed rock and the lowest values at rock parts affected by efflorescence, in accordance with afore provided results. Just a note on Figure 11b: the colder region occurring right below the window sill likely originated from the shadow effect played by the column capital, which prevented this part from direct insolation before the survey. Similarly, the bioweathering dark crust occurring along the wall, specifically located where signs of water flow were surveyed (Figure 11a), shows high surface temperatures (about 31 °C). This is due to the similar thermal behavior arising from the similar color owned by these two types of weathering. In this case, a differentiation based only on IRT could represent a challenge, although it must be underlined that this survey was carried out during the dry season, and no rainfall had occurred in the weeks preceding the survey.

Discussion and Conclusions
Results arising from the surveying methodological approach presented in this paper show that Infrared Thermography (IRT) represents a useful non-contact tool to be employed in the quick preliminary evaluation of the weathering state of natural stones. From Similarly, the bioweathering dark crust occurring along the wall, specifically located where signs of water flow were surveyed (Figure 11a), shows high surface temperatures (about 31 • C). This is due to the similar thermal behavior arising from the similar color owned by these two types of weathering. In this case, a differentiation based only on IRT could represent a challenge, although it must be underlined that this survey was carried out during the dry season, and no rainfall had occurred in the weeks preceding the survey.

Discussion and Conclusions
Results arising from the surveying methodological approach presented in this paper show that Infrared Thermography (IRT) represents a useful non-contact tool to be employed in the quick preliminary evaluation of the weathering state of natural stones. From the analysis of different types of weathering occurring at two stone types employed for the construction of a heritage building, results can be summarized and discussed as follows ( Figure 12). Figure 12. Schematic summary of the IRT qualitative ranges of temperature related to each weathering type observed under no solar radiation conditions and divided by rock type.

EER REVIEW 19
(1) Efflorescence affects both studied rock types, and its occurrence leads to the formation of a light-color salt crystallization film on the stone surface. When the salt crystallization occurs within the rock structure (subflorescence), it leads to the progressive flaking of the rock surface, producing thin slices of rock material and progressively detaching from the rock. IRT allowed the detection of these weathering types thanks to key thermal contrasts. In particular, at basalts, the highest surface temperatures were found at the sound rock, where weathering did not occur, or at the bared rock portions, where flaking phenomena exposed the rock underneath the weathered surface. Intermediate surface temperatures were found where the salt crystallization film occurred, with the lowest values measured at its maximum thickness. Lowest surface temperatures were highlighted at the rock slices partly detached from the substratum due to flaking phenomena. Moreover, the cold regions affecting such rock portions can be regarded as flaking expansion areas, thus providing interesting information otherwise undetectable by the naked eye. The same considerations can be carried out for limestones, where efflorescence is often coupled with other weathering types and where flaking showed a slightly higher surface temperature than the surrounding area with efflorescence, unlike what was observed at the basaltic stereobates. This may be due to the presence of a heavier efflorescence around the limestone rock slice, which enhances the moisture absorption also due to the highest porosity of the rock type if compared to basalts. (2) Alveolar weathering affects only limestones, which represent the weaker rock type analyzed herein. Its occurrence was found localized in those rock regions overlaying the basaltic, more protruding strip, which favors the rain droplet rebound and their consequent impact and erosion on the limestone surface. Both stylobates of columns and the exposed wall of the western staircase are affected by this weathering type, (1) Efflorescence affects both studied rock types, and its occurrence leads to the formation of a light-color salt crystallization film on the stone surface. When the salt crystallization occurs within the rock structure (subflorescence), it leads to the progressive flaking of the rock surface, producing thin slices of rock material and progressively detaching from the rock. IRT allowed the detection of these weathering types thanks to key thermal contrasts. In particular, at basalts, the highest surface temperatures were found at the sound rock, where weathering did not occur, or at the bared rock portions, where flaking phenomena exposed the rock underneath the weathered surface. Intermediate surface temperatures were found where the salt crystallization film occurred, with the lowest values measured at its maximum thickness. Lowest surface temperatures were highlighted at the rock slices partly detached from the substratum due to flaking phenomena. Moreover, the cold regions affecting such rock portions can be regarded as flaking expansion areas, thus providing interesting information otherwise undetectable by the naked eye. The same considerations can be carried out for limestones, where efflorescence is often coupled with other weathering types and where flaking showed a slightly higher surface temperature than the surrounding area with efflorescence, unlike what was observed at the basaltic stereobates. This may be due to the presence of a heavier efflorescence around the limestone rock slice, which enhances the moisture absorption also due to the highest porosity of the rock type if compared to basalts. (2) Alveolar weathering affects only limestones, which represent the weaker rock type analyzed herein. Its occurrence was found localized in those rock regions overlaying the basaltic, more protruding strip, which favors the rain droplet rebound and their consequent impact and erosion on the limestone surface. Both stylobates of columns and the exposed wall of the western staircase are affected by this weathering type, which often occurs together with efflorescence and related flaking evidence. Their horizontal depth was estimated through the use of 3D photogrammetric models, which allowed comparing the morphology of the alveolar structure to the IRT outcomes. In particular, when the alveolar structure is well defined, alveoli keep a high surface temperature due to the hollow morphology, while when efflorescence and/or flaking occurs within the cavity, its surface temperature is lowered. (3) Black crust and bioweathering were surveyed at limestones. The first one occurs in the most sheltered portions of the façade or where rain water did not play a washing effect on the rock surface. Bioweathering occurs in specific regions, especially in correspondence with signs of water flow. Due to the fact that surveys were performed during the dry season, the IRT aspect of these two weathering types is given by warm surface temperatures, arising from the darker color owned by the weathered rock portion. Therefore, in this specific case, the chromatic difference between the washed/unweathered rock and the darker films allows differentiating these elements by IRT.
It is underlined that the IRT results reported herein refer to surveys carried out under no direct solar radiation, which is considered the best time for IRT surveys for rock mechanics [16] and the absence of rain. Therefore, such results should be considered preliminary and more studies should be further performed to investigate the applicability of IRT to this specific field. Anyhow, IRT proved a reliable tool for a preliminary zonation and mapping of the weathering processes affecting natural stones. This gains relevance when quick studies are required, especially for preliminary assessments. IRT can be, indeed, employed to define restoration strategies and/or to prioritize interventions, as it allows a quick location of the problem, especially if located at high structural elements/building portion. The useful combination with 3D photogrammetry has a high potential for the quantitative estimation of weathering development in terms of the affected thickness of the stone. This would be useful also for monitoring purposes, both to study the weathering progression on heritage buildings and to verify the efficiency of restoration works already carried out.