Microclimate monitoring of Ariadne’s house (Pompeii, Italy) for preventive conservation of fresco paintings
© Merello et al.; licensee Chemistry Central Ltd. 2012
Received: 7 August 2012
Accepted: 23 November 2012
Published: 28 November 2012
Ariadne’s house, located at the city center of ancient Pompeii, is of great archaeological value due to the fresco paintings decorating several rooms. In order to assess the risks for long-term conservation affecting the valuable mural paintings, 26 temperature data-loggers and 26 relative humidity data-loggers were located in four rooms of the house for the monitoring of ambient conditions.
Data recorded during 372 days were analyzed by means of graphical descriptive methods and analysis of variance (ANOVA). Results revealed an effect of the roof type and number of walls of the room. Excessive temperatures were observed during the summer in rooms covered with transparent roofs, and corrective actions were taken. Moreover, higher humidity values were recorded by sensors on the floor level.
The present work provides guidelines about the type, number, calibration and position of thermohygrometric sensors recommended for the microclimate monitoring of mural paintings in outdoor or semi-confined environments.
Pompeii was a village of ancient Rome located about 26 km southeast of the modern city of Naples (Italy). Around the year 62 AD, an earthquake severely damaged Pompeii and other nearby locations. Since then, the city was rebuilt until the eruption of Mount Vesuvius in 79 AD. Pompeii at that time had a population of about 15,000 inhabitants. The violent eruption of this volcano buried the city and preserved the ruins for centuries. Thick layers of ash covered Pompeii and Herculaneum, both towns placed at the base of Mount Vesuvius. Their names and locations were forgotten until the 18th century when a new interest for antiquity led to excavations . Herculaneum was rediscovered in 1738 and Pompeii in 1748. Since then, a campaign was launched to unearth both cities, revealing many intact buildings and valuable wall paintings.
Pompeii offers a picture of Roman life in the first century. The forum, baths, many houses, and some villas remained in a surprisingly good state of conservation. The site was declared World Heritage by UNESCO in 1997 and has become a popular tourist destination, with 2.5 million visitors in 2007. Pompeii is an open-air museum of 1,500 buildings comprising around 20,000 m2 of mural fresco paintings . Pompeian interior frescoes have become a main source of knowledge about Roman painting up to 79 AD, not because the city was so important at that time but because its tragic end has preserved in good conditions all its wall frescoes for posterity to study .
An imposed moratorium stopped excavations in the site to focus the efforts in maintaining the unburied ruins and leaving the remaining excavations for future generations. At present, the access to the ruins is more restricted for tourists, and less than one third of the houses open in the 1960s are currently available for public visits. This is due to the endless maintenance works to prevent the deterioration of buildings already unearthed . Apart from ruin fallings, at least 150 m2 of fresco paintings and plaster works are lost every year due to lack of maintenance.
Description of Ariadne’s house
Ariadne’s house is a Pompeian domus (i.e., a single-family house owned by the upper classes) situated in a privileged location at the city center (Regio VII, insula 4). It is a domus of Hellenistic inspiration with an extension of 1,700 m2. The elevation above sea level of Pompeii is about 14 m. The prevailing wind directions are northeast (October to February) and south (March to September), with an average wind speed of approximately 2.57 m/s (http://es.windfinder.com). The house was built at the end of the second century BC and was heavily damaged during the earthquake of 62 AD, not having been finished its reconstruction on the fateful date of Mount Vesuvius eruption . Most inner walls of the house were originally decorated with frescoes. The complete pictorial collection of Ariadne’s house has a remarkable quality and was probably created by the same artists’ team.
Room 2 is comprised by four walls, the northwest one with a large window and a doorway in the SW wall. As a result, the indoor microclimate is more isolated from the outside environment compared with the other rooms under study that are delimited by three walls. Rooms 1 and 3 are open to the courtyard by their NE side. Room 4 is open to the atrium by its west side. It is the largest one and has only one wall decorated with frescoes that presents a semicircular shape (Figure 1).
An extensive scientific literature is available about the analysis of materials and characterization of mural paintings in Pompeii [5–8] and other Roman cities [9, 10]. In the benchmark of a conservation project started in 2008, a series of works were carried out in Ariadne’s house: microclimate monitoring, electromagnetic radiation measurements, study of materials, and photographic report. The purpose was to assess the conservation state of frescoes and the most convenient future restoration works of Ariadne’s lodgings still containing mural paintings. These were all covered rooms (1 to 4 in Figure 1), whose roofs have preserved wall frescoes during the last decades.
The multidisciplinary team has studied the problems of salt efflorescence, atmospheric pollution and other pathologies associated with the materials in order to determine the causes of fresco degradation and propose conservation actions . It was found that all pigments are from inorganic origin. The deterioration of frescoes due to inappropriate temperatures in the rooms assessed by thermography was also discussed. The multidisciplinary approach has been fundamental to provide the basic guidelines for future in-depth restoration works. Regarding the microclimate study, a set of 26 thermohygrometric probes comprised by temperature and relative humidity (RH) data-loggers were installed for monitoring ambient conditions inside the roofed rooms. The statistical analysis of data recorded by these sensors will be the subject of the present research.
Different reported studies have measured air temperature, RH and other parameters inside museums for the preventive conservation of their collections [12–15]. Similar works have monitored the indoor environment in churches, as they contain valuable artifacts [16–21]. However, very few studies have characterized ambient conditions of outdoor archaeological sites [22–24] or semi-confined environments . The present work reports for the first time a microclimate research conducted in the ruins of Pompeii, which is of relevant interest because inappropriate conditions of temperature and RH are causing severe damages on the valuable fresco paintings. Moreover, results reported here will provide guidelines to establish thermohygrometric monitoring systems in similar open-air archaeological sites for preventive conservation.
Description of data-loggers
The use of autonomous devices was the best option in this case due to the lack of electric supply in Ariadne’s house. Each RH data-logger (Datalog Hygrochron DS1923) contains a humidity sensor with an accuracy of ± 5% . Although this model can also record temperatures, it was decided to use independent devices (Datalog Thermochron DS1922L) for the temperature monitoring , which has the same accuracy (±0.5°C) as DS1923. The reason was to expand the data storage capacity of the monitoring system.
A set of 52 data-loggers, 26 of each model, were purchased directly from the manufacturer (Maxim Integrated Products, Inc., Sunnyvale, CA) and they were calibrated prior to their installation as described ahead. These devices resemble button-like batteries, with 17.4 mm of diameter and 5.9 mm of height. Given their small size, each pair of DS1923 and DS1922L data-loggers were assembled together by means of a metallic structure and will be referred to hereafter as thermohygrometric probe. This probe is isolated by an external PVC structure of dark-gray color, which allows a convenient fixing to the wall, and has a cylindrical shape with a diameter of 6 cm. It is filled with polyurethane foam to isolate the data-loggers from the PVC structure. The assembly allows both data-loggers to be directly in contact with the ambient air, and therefore it was assumed that the supporting structure had a negligible effect on the recorded data.
Results and discussion
Calibration of sensors
Height and bias a of temperature and RH data-loggers
Probe #2 was the one with highest RH values recorded in the second period (Figure 2c, bias = 1.9%), but it was the third one with lowest RH in the first stage (Figure 2a, bias = −1.1%). Thus, from the first to the second period, the shift is 1.1+1.9 = 3%. This error seems too high as the purpose of the present study is to discuss the slight differences recorded among sensors, and no bias correction was applied to #2. This abnormal performance was caused by an accidental water drop over this data-logger during its uninstallation. If RH data from #2 are disregarded for the calibration study, it turns out that the correlation between the average RH recorded by each data-logger in the first vs. the second calibration period is statistically significant (r = 0.701, p = 0.0001). Again, this result reveals certain bias, and the same procedure described for temperature was applied to estimate the errors. The resulting bias estimations range from −1.5% (#19) to +0.8% (#5). This range is much lower than the measurement error of ±5% indicated in the manufacturer data sheet . It was decided to correct all temperature and RH values collected in Ariadne’s house according to the estimated biases (Table 1) in order to improve the reproducibility of the monitoring system.
In order to determine more accurately the real bias, a further experiment was conducted with aqueous solutions of two salts. Results confirmed that the bias was small and irrelevant for the purpose of the present study.
Descriptive data analysis of mean trajectories
Knowledge about ideal or limit values of microclimate parameters for conservation of cultural heritage is still poor . The Italian UNI 10829  and DM 10/2001  are currently the approved standards on this issue. They do not provide guidelines for the conservation of outdoor paintings, but nevertheless the admissible values of RH and temperature indicated by both standards can be taken as a reference for the conservation of Pompeian frescoes. According to , the recommended range of RH and temperature for mural paintings is 55 – 65% and 10 – 24°C, respectively. These ranges are narrower than the ones suggested by : 45 – 60% and 6 – 25°C. Taking into account that both standards were developed for indoor conditions, we decided to consider here the intervals given by , which are wider and could be better extrapolated for outdoor conditions.
Daily oscillations of temperature and RH in summer can be up to 25°C and 40%, respectively (Figure 4). High variability is inappropriate for the conservation of fresco paintings, which encourages some kind of corrective action to reduce the temperature at least in summer . A monitoring system will be necessary to study the effectiveness of any action of this kind, and to assess the risks for the long-term conservation of frescoes.
Figure 4a shows that all RH trajectories are quite parallel. Thus, the average RH of all values recorded by each sensor during the monitoring study characterizes the basic dissimilarities among RH time series. In the case of temperature, not all trajectories follow the same pattern. Consequently, the average value is not enough to describe conveniently the differences among temperature trajectories, and additional parameters were studied using bivariate plots and analysis of variance as described below.
Detection of outliers
In order to check if the atypical peaks of temperature recorded by #3 occurred all the year round, we calculated the differences between #3 and #4. The latter was taken as a reference data-logger in the same room with a normal performance. Next, the maximum value of this difference was calculated for the 48 time observations corresponding to each day: MAX .
The sudden variations observed in certain trajectories at particular day frames can be seen as a mistake in the experimental setup. This is a problem for the statistical analysis, and the abnormal data of probes #2, #3 and #9 were discarded after carefully checking the values. Of course, direct sunshine on sensors produces abnormal temperature variations that should be avoided, but in long-term monitoring it is not evident how to determine the different trajectories of solar radiation incident on all walls during the year. This is an important recommendation for future studies of outdoor ambient monitoring.
Daily mean trajectories
Daily variations of temperature are more pronounced in rooms 1 and 3 (Figure 9a). Ceramic tiles protecting room 4 reduce the amplitude of the daily cycles (Figure 9b), which clearly indicates that this type of roof seems more convenient for the conservation of frescoes than the transparent polycarbonate sheets covering the other rooms. Actually, marked oscillations of thermohydrometric parameters represent a serious risk for the long-term preservation of frescoes.
All temperature trajectories seem to converge at night to a common pattern, but their variability is much higher at midday, with a range of about 15°C comparing the values of #9 vs. #11. Moreover, the shape of these trajectories differs at midday as the maximum value is reached around 9:00 AM in some cases (e.g. #8, #9 or #21) while other probes registered the maximum temperature about 12:00 AM (e.g. #13, #14 or #20). In the case of RH, the highest variability is observed at about 10:00 AM. At this time, the range among daily RH trajectories is about 45% (#21 vs. #25).
Analysis of variance (ANOVA)
Different statistical parameters were calculated for each day and each temperature data-logger: minimum (Tmin), maximum (Tmax) and average. Similarly, daily averages of RH, minimum (RHmin) and maximum (RHmax) values were also obtained. In order to study the effect of height and wall orientation where the probe was located, different ANOVA models were carried out with these parameters. In all cases, two factors were considered: day and data-logger. The effect of day is not of interest here, and results will only be discussed according to the second factor. The best results were obtained with RH recorded in summer (from 7/24/2008 to 9/22/2008). Alternative ANOVAs using only winter data or both periods were also checked, but the interpretation was less clear.
An additional effect of sensor height is also apparent in room 3 (Figure 11b). Actually, comparing probes with the same orientation, those on the floor recorded significantly higher RHmin values. Results reported here indicate that data-loggers located on the ground have a different pattern, and consequently all of them will be treated as an independent cluster in the bivariate plots presented in the next section.
Different ANOVAs were also performed with daily average temperatures, Tmax and Tmin, but results did not provide a clear interpretation regarding the effect of height and wall orientation on these parameters. Further studies with multivariate statistical methods will be necessary for this purpose. The fact that RH provides more information than temperatures was also found in a recent study .
Summarizing, sensor height (floor vs. upper position) was the main cause of RH variability in the room with four walls. An additional effect of wall orientation on RHmin was observed in rooms 1 and 3 that are quite similar. Consequently, it appears that the location of probes was appropriate, and similar criteria should be applied in further studies of outdoor monitoring. In closed rooms, it seems advisable to install sensors on the floor, and two different orientations might be enough for wall sensors. By contrast, in outdoor sites or lodgings with a high rate of air exchange, special care should be taken in wall orientation. In all cases, direct sunlight on data-loggers should always be avoided to prevent abnormal measurements.
In order to characterize the dissimilarities among probes, different bivariate plots were obtained and visually inspected. Figure 14a displays average RH vs. average temperatures. Taking into account the relationship between RH and AH (Equation (1)), RH was plotted as a function of temperature inside this figure by considering AH=0.0109 kg/m3 (Equation (2)). This value is the one that achieves the best goodness-of-fit in Figure 14a and it was estimated using non-linear regression after disregarding floor probes and #1.
According to Figures 14, 15 and 16 probes on the floor recorded higher RH and lower temperatures. In particular, most of them yielded the highest mean RH along the year (Figure 14a), and the highest RHmin values in winter (Figure 16b). Moreover, they yield the lowest mean temperatures along the year (Figure 14a) and the lowest Tmax in winter (Figure 16a). As an exception, probe #8 registered much higher temperatures (Figure 14a). The reason is unclear, but taking into account that #8 was oriented to the south, perhaps it received more diffuse solar radiation. The same reason might explain why probes #9 and #21 recorded the highest Tmax (Figure 16a) and the lowest RHmin values (Figures 11 and 16b), given that they were also south-oriented. This hypothesis is mere speculation and further studies using more data-loggers will be necessary to investigate this issue.
Assuming that AH is constant inside all rooms, the curve depicted in Figure 14a highlights that floor probes #14 and #16 (room 2) deviate from the theoretical relationship between RH and temperature. The RH mean daily time series of both probes is also atypical (Figure 9d) as the daily decrease of RH registered at sunrise is delayed in about 30 min with respect to the other trajectories. This abnormal performance might be caused by diffusion of water vapor by capillarity through the ground to the boundary layer in contact with the data-logger. The fact that higher RHmin values were recorded by floor sensors is also consistent with this hypothesis. Room 2 is the only one comprised by four walls, but it is uncertain if this issue could also explain their atypical performance.
Probe #1 also appears as an outlier in Figure 14, which suggests that AH outside Ariadne’s house might be higher than indoors. Moreover, this probe recorded the lowest Tmin (Figure 15) and higher RHmin values in winter with respect to wall sensors (Figure 16b). Results reveal that the microclimate reflected by #1 was slightly different, which suggests using one or more outside sensors in similar monitoring studies.
According to DM 10/2001 , the recommended interval of RH is 45-60%. Regarding RHmin, only probes in room 4 satisfy this condition in summer and winter (Figure 16b). By contrast, RHmin tends to be lower than 45% in rooms 1 and 3 (Figure 15), which indicates ambient conditions too dry for an appropriate conservation of frescoes particularly in summer.
Figures 14, 15, 16 reveal that ambient conditions in rooms 2 and 4 were cooler in summer, more humid and more stable (i.e., with less variability among probes) than in rooms 1 and 3. In winter, temperatures in rooms 2 and 4 tended to be less cold (Figure 16a), which appears more appropriate for preserving the frescoes. The observed differences among lodgings can be explained by the roof type and number of walls of the room. Actually, room 2 is characterized by an environment more isolated from outdoor conditions and room 4 is partly covered with ceramic tiles that prevent incoming sunlight. The others are roofed with transparent sheets that produce a greenhouse effect, which tends to increase the temperature. Thus, opaque shelters seem more convenient. Taking into account that probe #5 was located under a glass, the greenhouse effect could also explain the higher temperatures recorded by this data-logger with respect to the others on the floor of room 3 (Figure 14a).
Experts in restoration of frescoes are currently assessing the conservation state of paintings in Ariadne’s house. Preliminary results suggest that frescoes in room 2 are better preserved that the rest, which is of interest because room 2 presents ambient conditions closer to the range recommended by DM 10/2001 . This issue supports the use of this standard for discussing the statistical data analysis.
Given the unfavorable environment revealed in the present study, the transparent roof in rooms 1–3 was replaced in December 2009 by undulating roof sheets made of fiber cement with a thickness of about 5 mm. The conclusion that transparent roofs are inappropriate for the conservation of outdoor wall paintings might seem obvious for experts in the field, but the competing authorities requested empirical evidence prior to approving the roof change. The effectiveness of this corrective measure will be assessed in a future study.
The statistical analysis of RH and temperatures recorded in Ariadne’s house has revealed that autonomous data-loggers with a sampling rate of two readings per hour seem appropriate. Considering that few works have monitored microclimate environments of outdoor or semi-confined archaeological sites, results reported here are of interest for similar studies regarding the number, type and position of data-loggers, as well as for the subsequent statistical data analysis. The first step of the analysis should be the bias correction of all measurements based on the calibration experiment.
Descriptive tools have been effective for highlighting the dissimilarities among probes. The mean daily trajectories displayed in Figure 9 provide useful information about the difference in average values and shapes among data-loggers. This type of plot is used in several fields of science to visualize and synthesize differences in recorded values, but it is rarely applied in microclimate monitoring where trajectories of a few days randomly chosen are usually presented.
By means of ANOVA and bivariate plots, it was found an effect of roof type, wall orientation and sensor height on microclimate conditions. Higher RH values were recorded by data-loggers on the floor, and the south orientation was characterized by lower RH values because it tended to receive more sunlight radiation. The differences among probes suggest that at least four data-loggers per room would be required in similar studies to assess the effect of different factors and to detect abnormal trajectories. One or more control sensors are also recommended. Regarding the number of walls of the room, rooms with four walls will require less number of sensors at wall levels, where two different orientations might be enough. At floor level, north orientation and a complementary one (south or east) should also be monitored. If several lodgings present similar characteristics, orientation, size and number of walls, it seems sufficient to monitor just one of them.
A tight correlation was found between average RH and temperatures, which suggests that it is not necessary to use the same number of RH and temperature sensors in open-air archeological sites. Redundant information will also result if lodgings with very similar characteristic are monitored, which reveals the importance of choosing properly the right location for probes.
Results reveal that summer conditions in Pompeii are more unfavorable for the conservation of frescoes according to the standard DM 10/2001 because the microclimate was too dry and hot. Consequently, ambient conditions in similar studies should be basically monitored in summer. Transparent roofs produce a greenhouse effect that increases the temperature, which is undesirable. Based on the results, corrective measures were taken by replacing the type of roof. The effectiveness of such measures will be assessed in future studies.
Installation of data-loggers
One target was to study the vertical gradient of RH and temperature. For this purpose, data-loggers were placed in the four rooms at three levels: floor, intermediate (15 – 200 cm) and upper position (200 – 338 cm). Thermographic studies carried out previously  were also taken into consideration to decide the best location of probes. The intermediate position was omitted in room 2 because it is more isolated from the outside environment and, hence, lower thermohygrometric gradients were expected. Data-loggers in room 4 were not located on the ground because the floor was full of rubble, which might result in microclimate conditions not directly comparable with respect to the floor of the other rooms.
Probe #5 was placed inside a glass box that protects a mosaic on the floor of room 3 in order to assess the microclimate inside this box. All probes on the ground except #5 were placed near the walls (Figure 17), as otherwise they would hamper the free movement of visitors inside the room. The rest were installed hanging on walls except the reference probe #1.
Calibration of sensors
A calibration experiment is another compulsory step in a methodology of microclimate monitoring. Two different calibration experiments were conducted. The first was carried out in two periods (at the beginning and at the end of the experiment) in order to study if measurements from one or more sensors were biased with respect to the average recorded by all sensors. During both periods, all data-loggers were kept in a small compartment in contact with the indoor air of the laboratory whose conditions varied during the experiment.
For this calibration experiment, all data-loggers were confined inside an 8-liter compartment (20×10×40 cm) that was not sealed and allowed the inside air to reach an equilibrium of RH and temperature with respect to the outside. This experiment was carried out during a 3-day period (07/18/2008 to 07/21/2008) prior to the installation of probes in Ariadne’s house and during 28 days after the monitoring study (07/31/2009 to 08/28/2009), once all data-loggers were back in the laboratory. In the first calibration period, temperature ranged from 27.5°C to 29.5°C and RH ranged from 49% to 60%. In the second period, the range of temperature was 21.5 – 26°C, and 58 – 82% in the case of RH.
A further experiment was carried out with aqueous solutions of two salts (lithium chloride and sodium chloride) according to the standard ASTM E 104–02  in order to study the measurement errors of RH data-loggers.
Frequency of data recording
The monitoring study started on July 23rd 2008 and ended on July 30th 2009, resulting a period of 372 days. All data-loggers were programmed to register one measurement every 30 minutes, which implies 1,440 recorded values per month (i.e., 30 days × 24 hours/day × 2 data/hour). Taking into account that these devices are able to store 8,192 registries, it turned out that they could operate for about 5.7 months, which allowed the scientific team to travel conveniently from Spain to Pompeii three times a year to collect the data.
Analysis of variance (ANOVA)
Taking into account that similarities and dissimilarities among trajectories might be different in summer versus winter, a summer period of 61 days was considered from 7/24/2008 to 9/22/2008 (i.e., from the first day of monitoring until the autumnal equinox). Similarly, a winter period of 90 days was considered from 12/21/2008 (winter solstice) to 31/3/2009. For each day and each data-logger, different statistical parameters of RH and temperature were calculated: daily averages, maximum and minimum values. These data were arranged in a matrix comprised by 3,926 rows (i.e., 61+90 days × 26 data-loggers). Four factors were considered and included in the matrix: day, season (summer or winter), data-logger code and room code. In order to study if the differences among probes in the same room were statistically significant, multifactor ANOVAs were carried out considering two factors (day and data-logger) and selecting data recorded in summer in one room. Next, the analysis was repeated for each room and considering only the summer or winter periods. All ANOVAs were performed using the software Statgraphics 5.1 .
This work was partially supported by the Spanish Government (Ministerio de Ciencia e Innovación) under projects HAR2010-21944-C02-01 and HAR2010-21944-C02-02.
- Ribera A, Olcina M, Ballester C: Pompeya Bajo Pompeya, las Excavaciones en la Casa de Ariadna. 2007, Valencia: Fundación MARQGoogle Scholar
- World Monuments Fund: World Monuments Watch: 100 Most Endangered Sites. 1996, New York: World Monuments FundGoogle Scholar
- Anter KF: Colours in Pompeiian cityscape: Adding pieces to the puzzle. Color Res Appl. 2006, 31 (4): 331-340. 10.1002/col.20228.View ArticleGoogle Scholar
- Harris J: Protecting Pompeii and the Italian heritage in 2012. http://www.i-italy.org/bloggers/18935/protecting-pompeii-and-italian-heritage-2012,
- Augusti S: La Tecnica Dell’antica Pittura Parietale Pompeiana. 1950, Napoli: Gaetano Macchiaroli EditoreGoogle Scholar
- Miriello D, Barca D, Bloise A, Ciarallo A, Crisci GM, De Rose T, Gattuso C, Gazineo F, La Russa MF: Characterisation of archaeological mortars from Pompeii (Campania, Italy) and identification of construction phases by compositional data analysis. J Arch Sci. 2010, 37: 2207-2223. 10.1016/j.jas.2010.03.019.View ArticleGoogle Scholar
- Castriota M, Cosco V, Barone T, De Santo G, Carafa P, Cazzanelli E: Micro-Raman characterizations of Pompei’s mortars. J Raman Spectrosc. 2008, 39 (2): 295-301. 10.1002/jrs.1877.View ArticleGoogle Scholar
- Maguregui M, Knuutinen U, Castro K, Madariaga JM: Raman spectroscopy as a tool to diagnose the impact and conservation state of Pompeian second and fourth style wall paintings exposed to diverse environments (House of Marcus Lucretius). J Raman Spectrosc. 2010, 41 (11): 1400-1409. 10.1002/jrs.2671.View ArticleGoogle Scholar
- Genestar C, Pons C, Más A: Analytical characterisation of ancient mortars from the archaeological Roman city of Pollentia (Balearic Islands, Spain). Anal Chim Acta. 2006, 557: 373-379. 10.1016/j.aca.2005.10.058.View ArticleGoogle Scholar
- Duran A, Perez-Maqueda LA, Poyato J, Perez-Rodriguez JL: A thermal study approach to roman age wall painting mortars. J Therm Anal Calorim. 2010, 99 (3): 803-809. 10.1007/s10973-009-0667-2.View ArticleGoogle Scholar
- Pérez MC, García Diego F-J, Merello P, D’Antoni P, Fernández Navajas A, Ribera Lacomba A, Ferrazza L, Pérez Miralles J, Baró JL, Merce P, D’Antoni H, Curiel Esparza J: Ariadne’s house (Pompeii, Italy) wall paintings: a multidisciplinary study of its present state focused on a future restoration and preventive conservation. Mater Constr. 10.3989/mc.2012.00812. in pressGoogle Scholar
- Bernardi A: Microclimate in the British Museum. London. Museum Manag Curat. 1990, 9: 169-182.View ArticleGoogle Scholar
- Bernardi A, Camuffo D: Microclimate in the Chiericati Palace Municipal Museum. Vicenza. Museum Manag Curat. 1995, 14: 5-18.View ArticleGoogle Scholar
- Camuffo D, Bernardi A, Sturaro G, Valentino A: The microclimate inside the Pollaiolo and Botticelli rooms in the Uffizi Gallery. Florence. J Cult Herit. 2002, 3: 155-161. 10.1016/S1296-2074(02)01171-8.View ArticleGoogle Scholar
- La Gennusa M, Rizzo G, Scaccianoce G, Nicoletti F: Control of indoor environments in heritage buildings: Experimental measurements in an old Italian museum and proposal of a methodology. J Cult Herit. 2005, 6 (2): 147-155. 10.1016/j.culher.2005.03.001.View ArticleGoogle Scholar
- Camuffo D, Sturaro G, Valentino A: Thermodynamic exchanges between the external boundary layer and the indoor microclimate at the Basilica of Santa Maria Maggiore, Rome, Italy: the problem of conservation of ancient works of art. Bound Lay Meteorol. 1999, 92: 243-262. 10.1023/A:1002026711404.View ArticleGoogle Scholar
- Tabunschikov Y, Brodatch M: Indoor air climate requirements for Russian churches and cathedrals. Indoor Air. 2004, 14 (Suppl 7): 168-174.View ArticleGoogle Scholar
- Loupa G, Charpantidou E, Kioutsioukis I, Rapsomanikis S: Indoor microclimate, ozone and nitrogen oxides in two medieval churches in Cyprus. Atmos Environ. 2006, 40: 7457-7466. 10.1016/j.atmosenv.2006.07.015.View ArticleGoogle Scholar
- Vuerich E, Malaspina F, Barazutti M, Georgiadis T, Nardino M: Indoor measurements of microclimate variables and ozone in the church of San Vincenzo (Monastery of Bassano Romano – Italy): a pilot study. Microchem J. 2008, 88: 218-223. 10.1016/j.microc.2007.11.014.View ArticleGoogle Scholar
- García-Diego F-J, Zarzo M: Microclimate monitoring by multivariate statistical control: the renaissance frescoes of the cathedral of valencia (Spain). J Cult Herit. 2010, 11 (3): 339-344. 10.1016/j.culher.2009.06.002.View ArticleGoogle Scholar
- Zarzo M, Fernández-Navajas A, García-Diego F-J: Long-term monitoring of fresco paintings in the Cathedral of Valencia (Spain) through humidity and temperature sensors in various locations for preventive conservation. Sensors. 2011, 11 (9): 8685-8710.View ArticleGoogle Scholar
- Maekawa S, Lambert F, Meyer J: Environmental monitoring at Tiwanaku. Mater Res Soc Symp Proc. 1995, 352: 885-892.View ArticleGoogle Scholar
- Lillie M, Smith R, Reed J, Inglis R: Southwest Scottish Crannogs: using in situ studies to assess preservation in wetland archaeological contexts. J Archaeol Sci. 2008, 35 (7): 1886-1900. 10.1016/j.jas.2007.11.029.View ArticleGoogle Scholar
- Verdecchia F, Zoccatelli C, Norelli E, Miandro R: Integrated monitoring network for surface deformation in Capo Colonna archaeological area, Crotone. Italy. IAHS-AISH P. 2010, 339: 345-351.Google Scholar
- Nava S, Becherini F, Bernardi A, Bonazza A, Chiari M, García-Orellana I, Lucarelli F, Ludwig N, Migliori A, Sabbioni C, Udisti R, Valli G, Vecchi R: An integrated approach to assess air pollution threats to cultural heritage in a semi-confined environment: the case study of Michelozzo’s Courtyard in Florence (Italy). Sci Total Environ. 2010, 408 (6): 1403-1413. 10.1016/j.scitotenv.2009.07.030.View ArticleGoogle Scholar
- Hygrochron Temperature/Humidity Logger iButton with 8KB Data-Log Memory. Maxim Integrated Products. http://datasheets.maxim-ic.com/en/ds/DS1923.pdf,
- Temperature Logger iButton with 8KB Data-Log Memory. Maxim Integrated Products. http://datasheets.maxim-ic.com/en/ds/DS1922L-DS1922T.pdf,
- Visco G, Plattner SH, Fortini P, Di Giovanni S, Sammartino MP: Microclimate monitoring in the Carcer Tullianum: temporal and spatial correlation and gradients evidenced by multivariate analysis; first campaign. Chem Cent J. 2012, 6 (Suppl 2): S11-View ArticleGoogle Scholar
- UNI 10829: Works of Art of Historical Importance. Ambient Conditions for the Conservation. Measurement and Analysis. 1999, Milano: UNI Ente Nazionale Italiano di UnificazioneGoogle Scholar
- DM 10/2001: Atto di Indirizzo sui Criteri Tecnico-scientifici e Sugli Standard di Funzionamento e Sviluppo dei Musei. Ministero per i Beni e le Attività Culturali. 2001, Gazzeta Ufficiale della Repubblica Italiana (Official Bulletin of Italian Republica), Rome (Italy), DL 112/1998 art. 150 comma 6Google Scholar
- Camuffo D: Microclimate for Cultural Heritage. 1998, Amsterdam: Elsevier ScienceGoogle Scholar
- ASTM E 104–02: Standard Practice for Maintaining Constant Relative Humidity by Means of Aqueous Solutions. 2012, West Conshohocken, PA: ASTM IntlGoogle Scholar
- Statgraphics Software. http://www.statgraphics.net,
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.