1 Introduction

Limestone rock-hewn statues are significant cultural heritages in Southeast China and valuable materials for studying the secularisation of Buddhism (Su 1989). Southeast China has a humid subtropical climate with wet and scorching summers, as well as damp and cold winters. It experiences abundant precipitation, with an average annual rainfall of 800–1600 mm. The region is characterised by hilly landscapes and developed river systems, with a drainage density of 0.3–0.4 km/km2 (Zheng 2008). According to the 2021 general survey results on rock-hewn temples and statues in Zhejiang Province, China, there are 87 rock-hewn sites in this province. Most of these sites are carved from limestone cliffs (Li 2011). Preliminary survey results also indicate that many of the limestone rock-hewn sites in this area have different deterioration patterns, such as powdering, spalling, and peeling (Wei et al. 2018).

Water is a crucial weathering driver of stone heritages, such as stone buildings, grottos, and rock-hewn statues (Sass and Viles 2022). Due to their close association with geological formations and direct exposure to nature, rock-hewn statues have access to abundant water sources, including rivers, floods, rainfall, groundwater, capillary water, and condensate water (** surface moisture. By detecting temperature anomalies on the rock facade, it can indicate areas with higher water content, as these areas tend to display lower temperatures (Moropoulou et al. 2001; Martínez-Garrido et al. 2018; Barbosa et al. 2021; Pappalardo et al. 2022). Furthermore, nuclear magnetic resonance (NMR) technology has been utilised in the study of water distribution in severely damaged wall paintings and rock walls (Di Tullio et al. 2010; Camaiti et al. 2015). But, the widespread adoption of NMR has been hindered due to its high cost and complex operation.

Determining moisture in limestone rock-hewn heritage remains a complex issue that necessitates attention to mitigating weathering risks and preserving these valuable artefacts. Currently, a systematic measurement strategy for determining moisture content in such heritage is lacking. This paper aims to fill this gap by evaluating the spatial distribution of moisture on the rock facade through zonal moisture content testing, while investigating the potential sources of moisture and exploring the relationship between deterioration and moisture.

2 Study Site

The Ciyunling statues, situated east of Yuhuangshan Mountain in Hangzhou City, are an integral part of the Hangzhou West Lake Cultural Landscape, designated as a World Heritage site (see Fig. 1a, b). These statues were hewn during the Wuyue states of the Five Dynasties in 942 CE, and they are regarded as representative works of Buddhist statues from that era. There are four niches in total, containing 13 stone statues (Chang 1995). Niche 1 (See Fig. 1d), located at the site's north end, was chosen for the on-site investigation. This niche consists of three statues: the Bodhisattva Ksitigarbha in the middle, with a height of 198 cm, and two followers on either side, measuring 141 cm and 134 cm in height. And a thick rock platform connects the statues to the ground, with a height of approximately 65 cm.

Fig. 1
figure 1

Location and information of the Ciyunling statues. a the map of China displaying the West Lake Cultural Landscape of Hangzhou (World Heritage Site). b The map of the West Lake Cultural Landscape of Hangzhou including the specific location of the statues (the map was reproduced based on the official map of the inscribed property published on the UNESCO website (UNESCO WHC 2011). c The flat plan of the Ciyunling statues (reproduced from the figure published in the survey report (Chang 1995)). d Three statues in niche 1

Full size image

3 Methods

3.1 Petrological, Physical and Mechanical Properties Measurements

The mineral composition and texture properties of Ciyunling limestone that were sampled from the same bedding layer and facade as the Ciyunling statues were observed by a polarising microscope and determined by X-ray diffraction. Water absorption coefficient by capillary and splitting tensile strength were determined by standard test methods BS EN 1925:1999(1999) and ASTM c1006/c1006m-20a (2020), respectively. Before the two standard tests, the limestone samples were dried at 50℃ until reaching a constant mass.

The porosity and density of the limestone were determined by the Mercury Intrusion porosimeter (AutoPore IV 9510). The average pore diameter, pore volume and specific surface area of the rock samples were determined by Quantachrome AUTOSORB-1-C automatic physisorption analyser (the rock samples were pretreated at 300℃ under vacuum for 15 h with nitrogen used as the adsorbate). Testing was conducted in the State Key Laboratory of Chemical Engineering, Zhejiang University.

3.2 In Situ Moisture and Hardness Measurement

The field test was conducted on April 25, 2023, under cloudy weather conditions and an average temperature of 11 °C. The previous day, April 24, experienced light rain accompanied by a gentle north wind. Throughout the field test, there were no apparent signs of water flow on the statue's surface. The protective shelter surrounding the statue extends about 50 cm beyond its natural outline. Considering the statue's placement in the northeast, facing southwest, it is presumed that little direct rainfall or wind-driven rain impacted the statue's surface.

The portable Testo 606-2 moisture meter was utilised to determine the water content of the statues. The Testo 606-2 is a resistance-type moisture meter that measures materials' electrical resistance (ER) and displays the material moisture directly in percentage by weight via the stored characteristic curves between ER and moisture weight ratio. This moisture meter was selected due to the efficacy of ER technology in reflecting the distribution of surface moisture, specifically offering benefits in furnishing a comprehensive spatial assessment. (Weiss and Sass 2022).

The investigation area encompassed three statues and the background rock walls (See Fig. 2). This area was subdivided into eight equal parallel layers, with different measuring areas assigned to the statues and the background rock walls in each layer. The smaller statues on the left and right were further divided into six separate measuring areas (L1–L6, R1–R6), while the middle statue was divided into eight measuring areas (C1–C8). The left and right background rock walls were divided into eight areas each (BR1–BR8, BL1–BL8), totalling 36 measuring areas. Within each area, 20 moisture content readings were randomly taken, and the average moisture content of each area was calculated. The moisture content distribution was then mapped based on these calculations.

Fig. 2
figure 2

The overall measurement area and moisture measurement spots on the niche1 façade

Full size image

Moreover, the Leeb hardness tester (BH200C) was used to measure the surface hardness of the statues. This type of hardness tester is a non-invasive device based on the rebound theory and is commonly employed to assess the preservation status of built heritage (Viles et al. 2011; Freire-Lista and Fort 2017; Wang et al. 2020). Although the impact force exerted by the small metal ball is minimal, it can leave a small round mark on the rock surface during testing (Desarnaud et al. 2019). To minimise potential damage to the statues' appearance caused by intensive measurement, only six test areas and the rock platform beneath the statues were selected for surface hardness measurement. Ten random measurements were taken in each measuring area, and the average value was determined. Finally, the Kruskal–Wallis H test, a nonparametric test, was performed using IBM SPSS statistical software to ascertain any differences among the selected measuring areas.

3.3 Laboratory Calibration for the Establishment of the Relation Between the In Situ Moisture Data and the Gravimetric Moisture Content

A laboratory calibration test was designed to establish the relationship between moisture content obtained from the resistance-type moisture meter (Testo 606–2) and gravimetric moisture content, i.e. the mass of water per mass of dry stone. A local limestone sample was taken from the same bedding layer as the Ciyunling statues (See Fig. 3). The limestone specimen was placed in a dryer and vacuumed for 2 h. Deionised water was then injected into the vacuumed dryer until the water level was 5 cm above the highest point of the specimen, followed by opening the vacuum valve to restore normal atmospheric pressure. After 24 h, the specimen was removed from the water and placed in a transparent polymethyl methacrylate box with openings on the side and bottom to reduce the interference of air disturbance.

Fig. 3
figure 3

Ciyunling limestone sample for moisture calibration tests. a The main view of the limestone sample, where the number on the rock surface represents test points for water content, totaling ten points. b Left view of the limestone sample. c Right view of the limestone sample. d Rear view of the limestone sample. e Top view of the limestone sample

Full size image

The indoor environment conditions during the laboratory test were maintained at 25 ± 0.5 ℃/60 ± 2%rh. The moisture content of the specimen was measured at regular intervals, starting from saturation to drying. Measurements were taken every 30 min for the first 1.5 h and every 1 h for the next 8 h, totalling 570 min. The gravimetric method was used to determine the actual water content of the specimen while concurrently utilising the Testo 606–2 moisture meter to gauge the moisture content on the outer surface of the specimen that had been initially exposed to the ambient atmosphere during sampling. Each measurement involved ten fixed spots (See Fig. 3a). Regression analysis was performed to establish a relationship between the actual moisture content and the detected surface moisture content. Using the regression formula, the actual moisture content of the Ciyunling statues can be calculated by substituting the in situ detected moisture value.

4 Results and Discussions

4.1 Petrological, Physical and Mechanical Properties of Ciyunling Limestone

The Ciyunling statues were constructed on lower Permian strata, and the lithology therein is characterised mainly by limestone composed of calcite [CaCO3] and high-magnesium calcite [CaMg(CO3)2]. The stone has a greyish white appearance and coarse crystalline structure (see Fig. 4).

Fig. 4
figure 4

Microphotographs of the Ciyunling samples under a. plane-polarised light (PPL) and b cross-polarised light (XPL)

Full size image

The main physical and mechanical properties are presented in Table 1. The pore structure test results reveal that the rocks are relatively dense, with low porosity and nano-scale average pore size, which induces a low water absorption capacity.

Table 1 Lithologic character, porosity and physical/mechanical properties of the limestone sample
Full size table

4.2 Moisture Content and distribution of the Facade of Ciyuling Statues

The laboratory calibration test established a correlation between the detected moisture content (ωc) of the Ciyuling limestone using the Testo 606-2 moisture meter and the actual moisture content (ωo) obtained by the gravimetric method. The saturated water content of the limestone is 0.56wt.%. As shown in Fig. 5, there is a linear correlation between ωc and ωo (R2 = 0.9123). The linear equation is as follows:

$${\omega_o} = \, 0.04713 \cdot {\omega_c}\;- \, 0.3174.$$
(1)
Fig. 5
figure 5

The scatter plot of the moisture content and linear regression result for the Ciyunling limestone sample

Full size image

It can be seen that Testo 606-2 moisture meter can provide a solid response for the water content of the Ciyunling limestone, and the linear equation can be used to calculate the actual water content of the statues. Figure 6a depicts the detected moisture content using the Testo 606-2 moisture meter, while Fig. 6b represents the actual moisture content calculated using Eq. (1).

Fig. 6
figure 6

The moisture distribution on the surface of Ciyunling statues. (a) Moisture distribution based on the electrical resistance values from the Testo 606-2. (b) Actual water content derived from the linear regression equation

Full size image

Figure 6a shows that the three statues' moisture content, indicated by the red and white zone, is higher than that of the background rock wall between the statues (blue zone). The wettest area (red zone) is concentrated on the middle statue's waist and below, and on the chest and leg of the two smaller statues on the left and right sides. This indicates that water mainly accumulates in the parts of the statues below 1.6 m. Previous studies have shown that water transport in materials with a predominant micropore structure (< 100 nm) is primarily due to capillary condensation and surface diffusion (Siegesmund and Durrast 2011). Meanwhile, no trace of water flow was observed on the surface of the Ciyuling statues during the field testing. Therefore, it is likely that the moisture on the surface of the statues, dominated by nano-pores, mainly comes from gaseous water in the air that accumulates in the pores through capillary condensation.

Capillary condensation refers to the process of gaseous water condensing into liquid water in materials with a confined micro-structure, such as porous rock. This phenomenon can significantly impact the characteristics of adsorption, lubrication, friction, and corrosion at the solid–liquid interface (Huber 2015). The occurrence of capillary condensation is influenced by ambient humidity and pore size. Studies shown that capillary condensation in nano-scale pores follows the Kelvin equation, and the smaller the pore size, the lower the ambient humidity required for capillary condensation (Yang et al. 2020). Considering the location of the Ciyuling statues in a humid subtropical area with low-lying terrain and a dense river network, these climate and geo-hydrological characteristics would facilitate capillary condensation on the statues.

On the other hand, the statues have a three-dimensional form and a greater surface area than the flat background rock wall. Therefore, the statues theoretically contain more pores and a larger pore volume, allowing them to absorb and retain more moisture. This increased moisture content can lead to the dissolution of atmospheric acidic ions such as \({CO}_{3}^{2-},{SO}_{4}^{2-},{NO}_{X}^{-}\), thus accelerating the weathering process of the statues (Li et al. 2001). Additionally, the acid-insoluble matter content in the limestone of this region is generally low (< 5%), making the shrinkage of rock volume caused by non-uniform weathering more significant (Sun et al. 2002). Thus, water accumulation on the statue's surface indicates a higher likelihood of surface weathering and deterioration than the background rock wall between statues.

4.3 Surface Hardness, Surface Moisture Content and Weathering Degree of the Ciyunling Statues

As shown in Table 2, the surface hardness and moisture vary across the three areas of the statues, namely the facade root, leg and chest/shoulder. Based on the outcomes of the Kruskal–Wallis H test, there are statistically significant differences in surface hardness between the chest/shoulder area and the facade root. Similarly, a significant difference in moisture content was identified between the chest/shoulder area and the facade root. This suggests that, in general, the surface hardness of the limestone diminishes as the height decreases, while, reversely, the moisture content decreases as the height increases. Consequently, it can be presumed that the degree of weathering at the facade root is likely greater than at the higher parts of the statues, such as the chest/shoulder area, as a decline in surface hardness typically signifies a higher degree of weathering (Wilhelm et al. 2016; Wang et al. 2021). This circumstance could not only jeopardise the authenticity of the artefact but also compromise the stability of the overall structural niche.

Table 2 Median value，25% and 75% quartiles of surface hardness and moisture content, Kruskal–Wallis H test results for surface hardness and moisture content obtained by Testo 606-2 on the Ciyunling statues façade
Full size table

Furthermore, it should be noted that the distribution of surface hardness values is discrete. This discreteness can be attributed to both the inherent heterogeneity of the rock and the uneven surfaces of the rock statues. The presence of irregularities on the statues’ surfaces poses a challenge for the small metal ball in the surface hardness tester to achieve the desired impact angle of 90 degrees horizontally, resulting in deviations in the results. Accordingly, it is evident that the irregular features of rock-hewn heritages inevitably have an impact on the accuracy of on-site measurements. To account for this, it is necessary to increase the number of samples taken within an acceptable range in conservation practice, in order to better approximate the actual conditions.

5 Conclusion

The moisture content and distribution of Ciyunling statues were ascertained through in situ measurements and laboratory calibration tests. The limestone statues exhibit a higher moisture content than the background rock walls between the statues. Consequently, the statues may display greater vulnerability to the ambient environment and water-induced deterioration. The primary water source for these statues may be the gaseous water in the atmosphere, which accumulates within surface pores via capillary condensation. The sampling strategy of the on-site moisture content plays a crucial role in effectively portraying the overall depiction of the moisture distribution of the statues. Hence, it is advisable to consider this aspect during fieldwork in such historically significant rock-hewn heritage. The methodology applied in this study can be convincingly believed to be available for tackling this issue and contributing to future remedies.

Considering the escalating occurrence of extreme high-temperature events in recent years, future research could focus on examining the dynamic variations in moisture content and the response of rock properties under high heat and high humidity conditions. It will enhance our understanding of the potential effects of climate change on the long-term preservation of limestone rock-hewn heritage.