Long-term environmental monitoring for preventive conservation of external historical plasterworks

Journal of Building Engineering LONG-TERM ENVIRONMENTAL MONITORING FOR PREVENTIVE CONSERVATION OF EXTERNAL HISTORICAL PLASTERWORKS --Manuscript Draft-- Manuscript Number: JBE-D-21-03287R2 Article Type: Research Paper Section/Category: Building materials Keywords: plasterwork; temperature; relative humidity; architectural heritage; preservation Corresponding Author: VICENTE FLORES-ALÉS, Ph.D. Universidad de Sevilla SEVILLE, SPAIN First Author: Marta Torres-González, PhD aspirant Order of Authors: Marta Torres-González, PhD aspirant Carlos Rubio-Bellido, Ph. D. David Bienvenido-Huertas, Ph. D. Juan Manuel Alducin-Ochoa VICENTE FLORES-ALÉS, Ph.D. Abstract: In this study, an analysis of the environmental conditions in a Csa climatic zone for the conservation of plasterwork of the Real Alcázar of Seville is carried out. The measurements obtained from in-situ monitoring are compared with the measurements provided by AEMET (State Meteorological Agency in Spain) during the reference year and the study is completed by estimating future environmental conditions using two alternative approaches: a morphing process from the EPW of the climatic zone and the application of M5P data mining algorithms. An optimal temperature range is established for the conservation of the plasterwork that prevents their dehydration or the freezing of water particles contained. The transformation of gypsum into bassanite, the risks associated with exposures to high relative humidity and the consequences of the slight hygroscopicity of the material and the environmental conditions that must be developed to favor the growth of mold on the surface or the cracking of polychromies that embellish these plaster decorations on numerous occasions are analyzed. The results obtained allow us to establish preventive conservation measures not only on the plasterwork but also on the Real Alcázar of Seville and that architectural heritage located in the subtropical dry-summer climate. Suggested Reviewers: David Sanz, Ph. D. Lecturer, Polytechnic University of Madrid: Universidad Politecnica de Madrid [email protected] He is an expert research about heritage and traditional materials. He is specialized in the study of gypsum. Ana Silva, Ph. D. Research, University of Lisbon Higher Technical Institute: Universidade de Lisboa Instituto Superior Tecnico [email protected] Dra. Silva is a researcher with an extensive experience in building service life prediction and durability. Opposed Reviewers: Response to Reviewers: Dear editor: We have received the second revision of our manuscript JBE-D-21-03287R1 "LONGTERM ENVIRONMENTAL MONITORING FOR PREVENTIVE CONSERVATION OF EXTERNAL HISTORICAL PLASTERWORKS". Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation #Reviewer 1 now consider it in conditions for acceptance, but the review of #Reviewer2 is the same that we received previouslyit means the same comments and the same pdf attached for minor comments that we considered in the second version of our manuscript. As we indicated in the previous revision, all the suggestions have been taken into consideration. The text was rewriting to improve the comprehension and highlighted in yellow colour Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation Journal of Building Engineering Dear Editor: Let me introduce myself, my name is Vicente Flores-Alés and I am a Material Science Lecturer at the University of Sevilla (Spain). I submit the manuscript entitled “LONGTERM MONITORING FOR PREVENTIVE CONSERVATION OF HISTORICAL PLASTERWORKS” for your consideration in order to be published in Journal of Building Engineering. . The aim of this paper is evaluating the influence of environmental conditions on the conservation of historical plasterwork. The chosen case study is the Real Alcázar of Seville and was based on (i) the annual monitoring of the environmental parameters Temperature and Relative Humidity, (ii) the estimation of hourly values in future scenarios using artificial intelligence and (iii) the use of the threshold values for the conservation of plasterwork defined in the literature. Look forward to your favourable consideration. Most sincerely, Vicente Flores Alés Authors: Marta Torres-González, Carlos Rubio-Bellido, David Bienvenido-Huertas, Juan Manuel Alducin-Ochoa, Vicente Flores-Alés(*) [*Corresponding author] Mailing address: Vicente Flores Alés, E.T.S.I.E., Avda/ Reina Mercedes, nº4A, 41012-Sevilla ESPAÑA. E-mail address: vflor[email protected] Phone number: 34 954556656 Fax: 34 954556691 Cover Letter REVIEWER #1: Thank you for your comments. REVIEWER #2: As we indicated in the previous revision, all the suggestions have been taken into consideration. The text was rewriting to improve the comprehension and highlighted in yellow colour Response to Reviewers (R2) · The influence of temperature and relative humidity in the preservation of plasterwork. · Comparative analysis between different approaches. · Climate change will affect the maintenance of materials. · Future estimations allow to establish mitigation actions on architectural heritage. Highlights LONG-TERM ENVIRONMENTAL MONITORING FOR PREVENTIVE CONSERVATION OF EXTERNAL HISTORICAL PLASTERWORKS Marta Torres-González a,b *, Carlos Rubio-Bellido a, David Bienvenido-Huertas a, JM AlducinOchoa a, V. Flores-Alés a, a Department of Architectural Construction II, University of Seville b Department of Civil Engineering, Architecture and Georesources, University of Lisbon * Corresponding author: [email protected] , Av Reina Mercedes 4, Seville, 41012 KEYWORDS: plasterwork; temperature; relative humidity; architectural heritage; preservation; ABSTRACT: In this study, an analysis of the environmental conditions in a Csa climatic zone for the conservation of plasterwork of the Real Alcázar of Seville is carried out. The measurements obtained from in-situ monitoring are compared with the measurements provided by AEMET (State Meteorological Agency in Spain) during the reference year and the study is completed by estimating future environmental conditions using two alternative approaches: a morphing process from the EPW of the climatic zone and the application of M5P data mining algorithms. An optimal temperature range is established for the conservation of the plasterwork that prevents their dehydration or the freezing of water particles contained. The transformation of gypsum into bassanite, the risks associated with exposures to high relative humidity and the consequences of the slight hygroscopicity of the material and the environmental conditions that must be developed to favor the growth of mold on the surface or the cracking of polychromies that embellish these plaster decorations on numerous occasions are analyzed. The results obtained allow us to establish preventive conservation measures not only on the plasterwork but also on the Real Alcázar of Seville and that architectural heritage located in the subtropical dry-summer climate. Manuscript File [For Revision, Please upload clean version of Revised manuscript] Click here to view linked References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 1. INTRODUCTION 1.1 Thermo-hygrometric conditions and the conservation of the architectural heritage. The preventive conservation of heritage is crucial so that future generations can know the identity traits of their culture [1]. Therefore, the monitoring and analysis of specific environmental conditions [2] and their impact on architectural heritage, as well as future vulnerabilities due to global warming [3], are currently a challenge. There are numerous studies that analyze both the impact of environmental conditions on architectural heritage [4,5] and the conditioning of interior spaces to favor the preservation of objects (e.g., objects that belong to exhibitions) [6,7]. In both cases, there are national and international guidelines and standards that establish the interior air conditioning conditions of these buildings for such purposes, such as UNI 10829 Standard [8], UNE EN 15757 [9], ASHRAE STANDARD 55-2020[10] or the British PAS 198 [11]. Likewise, EN 15759-1 [12] also considers the possibility of using heating systems to achieve thermal comfort and for the optimal conservation of heritage elements. In addition to international standards, numerous studies have carried out thermo-hygrometric monitoring of architectural heritage in recent decades, resulting in a very useful tool to assess the state of conservation of the building and to know the causes of the degradation in order to determine a future restoration intervention or to establish preventive conservation criteria, ruling out the use of destructive techniques. Thus, long-term studies of up to 20 years duration [13,14] and studies in which monitoring is limited to one year [15] have been published. In this regard, it is common to evaluate at least annual monitoring periods that are complemented by thermal models on which different estimates of environmental conditions are established. Environmental monitoring and simulation tools make it possible to establish estimates of different factors, such as the modification of operational patterns, thermal comfort [16-18], or 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 energy conservation measures [19]. However, the impact of climate change, the basis of numerous investigations related to cities [20,21] has been sparsely discussed in reference to preventive conservation of heritage and long-term adverse aspects, which could become very significant [22,23]. 1.2 Relative humidity and temperature influence in the preservation of historical plasterwork. There are studies about how ambient conditions could affect constructive materials [24–28] but are insufficient with respect to gypsum or plasterwork. In this sense, it is worth highlighting the importance of follow-up campaigns to identify possible risks for the conservation of the plasterwork, prevent irreversible damage and advance conservation strategies. However, to account for all the possible damage mechanisms, a comprehensive methodology should be used to evaluate some risk indexes based on ambient temperature (T) and relative humidity (RH) measurements. Gypsum, due to its porous microstructure formed by crystalline groups of hydrated calcium sulfate, can incorporate water molecules inside, in the form of water vapor or even in the form of liquid water, when environmental conditions favor capillary condensation. Furthermore, gypsum has the ability to eliminate this water when environmental conditions change, reducing the content of water vapor in contact with the material [29]. Variations in T and RH can induce phase transitions. In general, gypsum (CaSO4·2H2O) dehydration occurs in dry environments [30], which is the reason why temperature below 40°C is desirable for the correct preservation of plasterwork. However, there may be situations of high temperatures combined with high relative humidity in which there are no alterations in the mineral phases of the gypsum [31]. In this regard, Winkler & Wilhelm (1970) indicated the relationship that must exist between ambient temperature and relative humidity to transform gypsum (CaSO4·2H2O) into bassanite (CaSO4·0.5H2O) [31], In this case, the material is likely to 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 lose stability if exposure to these unfavorable environmental conditions is prolonged over time [32] (Fig. 1-A, 1-B) . Fig. 1 (A) T-HR ratio for gypsum-bassanite transformation at 1 atmosphere pressure [31] and (B) Relationship between the relative humidity and the moisture content of the gypsum [33]. However, this transformation is part of a reversible process in plasterwork, which means that a rehydration would allow the recovery of the initial properties as long as the dehydration had not been prolonged in time until it remained constant, due to a change in climatic conditions, causing the mechanical weakening of the plasterwork [32]. Additionally, gypsum has low hygroscopicity [33,34] and medium-low solubility [35] that do not represent a direct risk in the preservation of plasterwork [36]. However, it is important to control the environmental conditions because the condensation of moisture on the surface of the plasters favors the alteration of the polychrome binders and their possible detachment [33]. The adsorption hygroscopicity of gypsum with 80% RH is only 4 g/kg (0.4%) [33], reaching a high adsorption hygroscopicity only by capillary saturation (99% RH) [34]. Likewise, humidity favors the nesting of biotic organisms and the alteration derived from their activity [33]. Therefore, an RH close to or greater than 90% weakens the structure of the plasterwork and favors the proliferation of microorganisms and biological agents, as well as the fixation of environmental pollutants.[37–39] (Fig. 01-B). 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 2.2. Measurement procedure To determine ambient temperature and relative humidity conditions, a Lascar thermohygrometer, model EL-USB 2 LCD with two recording channels, within the range -35ºC to 80ºC and 0% RH to 100% RH, with a resolution of ± 0.5ºC and ± 3% RH was used. In this case, the measurements have been carried out over a period of one year, at uninterrupted one-hour intervals from March 2020 to February 2021 with a thermo-hygrometer placed in the southeast gallery of the Courtyard of the Maidens (Fig. 6). The outdoor measurements were obtained through the AEMET (State Meteorological Agency in Spain). The meteorological station used was the one located in Seville. This meteorological station is equipped with a VAISALA HMP45D probe with a measuring range from 40 to 60 ºC and 0.8 to 100% RH, and an accuracy of ± 0.2°C and ± 2% RH. The measurement data obtained are also from one-hour intervals. To determine the interval of variation of RH by months, the average of the minimum and maximum daily humidities has been calculated and the monthly arithmetic means of these values have been obtained. These data have been studied following the indications of the UNEEN 15757: 2011 standard[9]. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Fig. 6 Location of the monitoring equipment on the ground floor of the Pedro I Palace. 2.3. Obtaining future climate data According to the data obtained during the annual monitoring, two approaches have been carried out to obtain future time series: (i) estimation of future outdoor climate data through the EPW morphing process of the climate zone of the case of study, and (ii) the estimation of the time series of the interior point through the application of data mining algorithms. 2.3.1. Morphing from the EPW of the climate zone of the case study The first of them was based on the climatic files provided by the Spanish Technical Building Code (CTE)[53] for the climate zone of Seville, Csa (Mediterranean) according to Köppen – Geiger classification [54]. In the first place, T and RH data resulting from the monitoring have been entered in the Seville climate archive to be able to carry out the evaluation of future scenarios. Secondly, T and RH data measured outdoors have been compared with those existing in the CTE to see if it is feasible to establish considerations, not only in the case of study but also in plasterwork in similar "subtropical dry-summer" climates. The calibration criteria of the Federal Energy Management Program (FEMP)[55], ASHRAE GUIDELINE Standard 14-2002 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 [56], [57] and the International Performance Measurement and Verification Protocol (IPMVP) [58] have been followed. This validation process makes it possible to verify if the results of the estimations for the climate zone of Seville are related to the real values monitored with a margin of error. For this, the following formulas are used: 𝑀𝐵𝐸 =∑(𝑚𝑖−𝑠𝑖) 𝑛 𝑖=1 𝑛 (1) 𝑀𝐵𝐸 =∑(𝑚𝑖−𝑠𝑖) 𝑛 𝑖=1 𝑛 (2) 𝑁𝑀𝐵𝐸 =1 𝑚 ∙∑(𝑚𝑖−𝑠𝑖) 𝑛 𝑖=1 𝑛−𝑝 (3) In order to scale the results of the Mean Bias Error (MBE) (Equation 1), Normalized Mean Bias Error (NMBE) is used by dividing MBE by the mean of measured values (𝑚), giving the global difference between the real values and the predicted ones (Equation 2); In these equations mi is the measured value, si is the simulated one and n the number of measured data points. The Coefficient of Variation of the Root Mean Square Error CV (RMSE) measures the variability of the errors between measured and simulated values (Equation 3). The limit values associated with these parameters vary depending on whether they are hourly or monthly. Considering that the environmental variables are hourly, the limit values are between -10% and 10% in NMBE and less than 30% in CV (RMSE) [56,57]. To estimate future climate scenarios in the climate zone, the CCWorldWeatherGen tool from the UK Met Office Hadley Center Coupled Model 3 HadCM3 has been used. Through a morphing process, this tool generates weather files adapted to climate change from any location in the world and creates files compatible with most building performance simulation programs [59]. The morphing of the climate files has been carried out for scenario A2 of greenhouse gas emissions according to the IPCC [60], resulting in climate scenario files for years 2050 and 2080. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 2.3.2. Estimation of future time series of the case study The data mining approach of this study was performed using the M5P algorithm. This algorithm allows estimating the time series for the years 2050 and 2080 in the measured interior point of the case study. The M5P algorithm (also known as M5') is an evolution of the Classification and regression tree (CART) [61,62]. Unlike CART, M5P combines decision trees with multivariate regression; a decision tree is built following the same inverted tree structure offered by CART, but a multiple linear regression (MLR) model is fitted on each sheet (Equation 4). Therefore, the algorithm works by developing an MLR model in each subregion (Fig. 7). In the process of developing the M5P algorithm tree, instead of maximizing the information gain, the internal variation of the subsets for the class values of each branch is minimized. Once the model is built, pruning reduces overfitting. The advantages of the models generated by this algorithm are that they efficiently handle large amounts of numerical variables and are robust in the absence of values in the instances of the analyzed dataset [63,64]. Thus, its use for the characterization of different aspects of buildings has increased in recent years [65,66]. 𝑌 𝑀𝐿𝑅 = 𝛽0+∑ (𝛽𝑖𝑥𝑖) 𝑣 𝑖=1 + 𝜀 (4) Where 𝛽0 is the independent term, 𝛽𝑖 are the regression coefficients, 𝑥𝑖 are the predictor variables, and 𝜀 is the error. Fig. 7. Scheme of an M5P model. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 The approach designed had a configuration of 6 input variables corresponding to the hourly values of temperature, relative humidity, dew point temperature, horizontal infrared radiation intensity, global horizontal radiation, and direction normal solar irradiance, diffuse horizontal solar irradiance of the time series of the outer point of Seville. The output variables correspond to the relative humidity and the temperature at the interior point of the case study. For each output variable a different M5P was designed. The training of each M5P model was carried out using the data from the current scenario. Once the models were trained, time series of the years 2050 and 2080 were estimated. It is worth noting that the M5P models were validated using the coefficient of determination and the MBE index as statistical parameters. 3. RESULTS AND DISCUSSION 3.1 Thermo-hygrometric conditions of the Palace of Pedro I The results obtained from the monitoring carried out between March 2020 and February 2021 in the southeast gallery of the Courtyard of the Maidens are shown in Fig. 8 and table 1: Fig. 8 Environmental conditions in the southeast corner of the Courtyard of the Maidens from March 2020 to February 2021. The measurement of the T and RH parameters is relevant mainly for the study of the fluctuations and thermal jumps that may occur during the day-night hours or during the different seasons of the year. In this regard, it is observed that the most pronounced thermal jumps between day and night hours do not exceed 5ºC (July and August), a fact that -by itself1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 does not represent any risk in terms of material expansion. Finally, a high RH in winter is recorded and must be controlled with adequate ventilation, although it is within the acceptable range for the correct preservation of plasterwork. Table 1. Summary of environmental parameters from 01/03 /2020 to 28 / 02/2021. Months RELATIVE HUMIDITY (%) TEMPERATURE (° C) Min. Values registered Max. values registered Δ (max-min) Min. Values registered Max. values registered Δ (max-min) Average SD Average SD Average SD Average SD Feb 20 55.3 11,001 73.8 9,454 18.5 14 1,319 18.1 1,577 4.1 Ap. 20 60.6 6,672 77.7 4.79 17.1 15.7 1,534 19.2 1,558 3.5 May 20 43.3 10,818 64.1 9,947 20.8 21.3 2,337 26.1 2,832 4.8 Jun 20 41.4 5,608 62 7.87 20.6 22.3 2,434 27.3 2,314 5 Jul. 20 38 5,877 56 7.43 18 27.6 1,448 31.8 1,328 4.2 Aug. 20 38 8,637 56.5 7,966 18.5 26.3 1,675 30.5 1,581 4.2 Sept. 20 44.1 9,493 58.3 8,287 14.2 24 1.9 27.4 2,105 3.4 Oct. 20 49.3 12,985 63.1 11,674 13.8 18.1 1,996 21.3 2,136 3.2 Nov. 20 70 8,459 78.6 7,134 8.6 15.6 1,964 17.9 1,965 2.3 Dec. 20 74.2 10,822 84.4 8,133 10.2 11.4 2,366 13.7 1,935 2.3 Jan. 21 75.2 9,545 86.1 8,141 10.9 8.6 3,493 11.2 3,093 2.6 Feb. 21 77.4 5,562 83.1 5,129 5.7 12.9 1,127 15.1 0.854 2.2 Fig. 9 Graphical representation of ambient condition (T and RH) from 01/03 /2020 to 28 / 02/2021. From a detailed analysis of the measurements obtained (Table 1 and Fig 9), the risk due to high RH is concentrated from November to February, both inclusive; being January the most unfavorable month since 27.02% of the measurements recorded during that month exceed the limit of 90% RH. Likewise, the risk of dehydration of the gypsum and transformation into bassanite (CaSO4·1/2H2O) occurs from May to October, both inclusive; being July and August 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 the most critical months for such purposes. It is worth to highlight that a 56.71% of the measurements recorded in the year remain below the limit established in the matrix (Fig. 10). The data obtained correspond to a total of 8,760 measurements, of which only 11.76% are outside the limits established in the matrix that relates thermo-hygrometric conditions with the mineralogical phases of calcium sulfate, depending on their degree of hydration. Fig. 10 Dispersion graph of the measurements from 01/03 /2020 to 28 / 02/2021 at the southeast corner of the Courtyard of the Maidens. Regarding the type of risk detected, it can be seen in Fig. 10 how it is mainly distinguished between risk due to internal structural transformations (associated with the summer months) and risk due to high RH (associated with the winter months). The 7.95% of the hourly records correspond to measurements that are below the curve established by Winkler& Wilhelm [31] , while the risk due to a high RH corresponds to 3.65% and the risk due to a low T corresponds to only 10 observations throughout the year, so the current climatic conditions in the Courtyard of the Maidens are favorable in this sense. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Fig. 11 Monthly representation of the hourly measurements obtained in the Courtyard of the Maidens from March 2020 to February 2021. However, the results obtained from monitoring the climate of the city of Seville show different trends in conservation risks (Fig. 11). As it can be seen, although the most prominent risks are once again that of internal transformations in summer and high RH in winter, there is a greater number of observations that remain within these risks. Thus, the observations at risk of internal transformations correspond to values between 348 and 360 in July and August, while in the Courtyard of the Maidens the values were 229 and 168, respectively. Likewise, it was also detected how the risk due to high RH extends to a greater number of months than in the case of the Courtyard of the Maidens and with a significantly higher number of observations. Fig. 12 Monthly representation of the hourly measurements obtained in Seville from March 2020 to February 2021. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Therefore, significant variations are detected in the results obtained with the two approaches. While the approach to the analysis of the Courtyard of the Maidens point (with a clear microclimate) presents less severe conditions for the conservation of plasterwork, outside the conditions would be different, presenting a greater risk due to the climate in these elements. Likewise, the variability detected between the two approaches shows the great variation that the analysis approaches of the influence of climate can present on the conservation of plasterwork. Thus, the use of a joint approach to the outdoor climate and microclimates in this type of case study allows us to show a broader perspective of the risk of conservation of plasterwork. 3.2 Calibration of outdoor measurements. The climate file of Seville. Despite the possible risks detected in the current scenario, climate change can generate a change in the conservation conditions of plasterwork. For that, the future time series for the two previously validated approaches used were estimated in this study. In the case of the outdoor climate approach, T and RH parameters measured during the outdoor monitoring are compared with T and RH data extracted from the CTE for the climate zone of Seville Csa (Mediterranean). As the environmental variables analyzed are hourly, the limit values are between -10% and 10% in NMBE and less than 30% in CV (RMSE). According to Table 2, it is detected that the CV (RMSE) values are within the established error limits. However, it is detected that T is slightly higher (0.88%) than the limits in the NMBE and in matters related to RH the value exceeds the limit in the NMBE by 6.40% (Table 2). From these results we can establish the following considerations: In the first place, considering that a measurement is being compared under specific environmental conditions with a typical meteorological year according to the criteria established in the CTE and based on the available meteorological stations, the level of approximation is good, especially if we consider the CV (RMSE). 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Table 2. Calibration of measured data in RAS compared to CTE. Limit values according to ASHRAE Guideline14. Variable NMBE [%] CV (RMSE) [%] Obtained values Limit values Obtained values Limit values Temperature (T) 11.03 ± 10 25.34 30 Relative humidity (RH) 15.52 ± 10 27.08 30 Therefore, to make predictions for climate scenarios on the case study, it is necessary to use the actual measured values of T and RH. However, in order to establish considerations on similar Csa dry-summer subtropical or Mediterranean climates, we can consider the approximation to the real measurements as acceptable, recommending the verification of the specific microclimatic conditions using procedures similar to the one used. In the case of the interior point approach, future estimates were obtained by M5P algorithm. Table 3 shows the values associated with each M5P model, while Fig. 13 shows the cloud of points between the actual and estimated values of the training and validation phase. As it can be seen, the coefficient of determination ranged between 74% and 87%, while the mean error ranged between 2.6ºC and 8% RH in each variable. The estimates were more limited in the RH variable in a similar way to that detected in other studies for estimating future time series of RH [14]. In any case, the results can be considered valid to have a knowledge of the future time series expected in the addressed case study. Table 3. Performance obtained by the M5P models in the training and validation phase. Approach Output variable Determination coefficient (R2) MBE M5P 1 Temperature (T) 86.53% 2.6 M5P 2 Relative Humidity (RH) 74.93% 8.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 [9] A.E. de N. UNE, UNE-EN 15757: Conservation of Cultural Property - Specifications for temperature and relative humidity to limit climate-induced mechanical damage in organic hygroscopic materials, (2011). [10] Ansi/Ashrae, ANSI/ASHRAE 55:2004 Thermal Environmental Conditions for Human Occupancy, Ashrae. (2004) 30. [11] PAS 198:2012 - Specification for managing environmental conditions for cultural collections – BSI British Standards, (n.d.). http://shop.bsigroup.com/ProductDetail/?pid=000000000030219669. [12] UNE-EN-15759-1, Conservation of cultural property. Indoor climate. Part 1: Guidelines for heating churches, chapels and other places of worship, (2012). [13] C. Bonacina, P. Baggio, F. Cappelletti, P. Romagnoni, A.G. Stevan, The Scrovegni Chapel: The results of over 20 years of indoor climate monitoring, Energy Build. 95 (2015) 144–152. https://doi.org/10.1016/j.enbuild.2014.12.018. [14] D. Bienvenido-Huertas, M. León-Muñoz, J.J. Martín-del-Río, C. Rubio-Bellido, Analysis of climate change impact on the preservation of heritage elements in historic buildings with a deficient indoor microclimate in warm regions, Build. Environ. 200 (2021) 107959. https://doi.org/10.1016/j.buildenv.2021.107959. [15] G. Zonno, R. Aguilar, R. Boroschek, P.B. Lourenço, Analysis of the long and short-term effects of temperature and humidity on the structural properties of adobe buildings using continuous monitoring, Eng. Struct. 196 (2019) 109299. https://doi.org/10.1016/j.engstruct.2019.109299. [16] R.P. Kramer, M.P.E. Maas, M.H.J. Martens, A.W.M. van Schijndel, H.L. Schellen, Energy conservation in museums using different setpoint strategies: A case study for a state-ofthe-art museum using building simulations, Appl. Energy. 158 (2015) 446–458. https://doi.org/10.1016/j.apenergy.2015.08.044. [17] M. Napp, T. Kalamees, Energy use and indoor climate of conservation heating, dehumidification and adaptive ventilation for the climate control of a mediaeval church in a cold climate, Energy Build. 108 (2015) 61–71. https://doi.org/10.1016/j.enbuild.2015.08.013. [18] C. Cornaro, V.A. Puggioni, R.M. Strollo, Dynamic simulation and on-site measurements for energy retrofit of complex historic buildings: Villa Mondragone case study, J. Build. Eng. 6 (2016) 17–28. https://doi.org/10.1016/j.jobe.2016.02.001. [19] T. Cardinale, G. Rospi, N. Cardinale, The influence of indoor microclimate on thermal comfort and conservation of artworks: The case study of the Cathedral of Matera (South Italy), in: Energy Procedia, Elsevier Ltd, 2014: pp. 425–432. https://doi.org/10.1016/j.egypro.2014.10.398. [20] A. Revi, D. Satterthwaite, F. Aragon-Durand, J. Corfee-Morlot, R.. R. Kiunsi, M. Pelling, D. Roberts, W. Solecki, S.P. Gajjar, A. Sverdlik, Towards transformative adaptation in cities: the IPCC’s Fifth Assessment, Environ. Urban. 26 (2014) 11–28. https://doi.org/10.1177/0956247814523539. [21] A.J. Prieto, K. Verichev, A. Silva, J. de Brito, On the impacts of climate change on the functional deterioration of heritage buildings in South Chile, Build. Environ. 183 (2020) 107138. https://doi.org/10.1016/j.buildenv.2020.107138. [22] V. Rajčić, A. Skender, D. Damjanović, An innovative methodology of assessing the 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 climate change impact on cultural heritage, Int. J. Archit. Herit. 12 (2018) 21–35. https://doi.org/10.1080/15583058.2017.1354094. [23] A. Haugen, C. Bertolin, G. Leijonhufvud, T. Olstad, T. Broström, A methodology for long-term monitoring of climate change impacts on historic buildings, Geosci. 8 (2018). https://doi.org/10.3390/geosciences8100370. [24] S. Ramírez, M. Zarzo, A. Perles, F.J. García-Diego, A methodology for discriminant time series analysis applied to microclimate monitoring of fresco paintings, Sensors (Switzerland). 21 (2021) 1–29. https://doi.org/10.3390/s21020436. [25] M. Gómez Heras, La Temperatura en los Materiales del Patrimonio, (2012) 87–95. http://digital.csic.es/handle/10261/46794. [26] Z. Pavlík, J. Fořt, R. Černý, An in situ monitoring system for the study of environmental influences on durability and the destructive process of building materials and structures, WIT Trans. Modelling Simul. 55 (2013) 287–296. https://doi.org/10.2495/CMEM130231. [27] S. Michalski, Temperatura Incorrecta, Can. Conserv. Institute. ICCROM. (2009). [28] S. Michalski, Humedad relativa Incorrecta, Can. Conserv. Institute. ICCROM. (2009). [29] L. Villanueva Domínguez, A. García Santos, Manual del yeso, CIE Inversiones Editoriales, 2001. [30] D. Freyer, W. Voigt, Crystallization and Phase Stability of CaSO4 and CaSO 4 - Based Salts, Monatshefte Fur Chemie. 134 (2003) 693–719. https://doi.org/10.1007/s00706003-0590-3. [31] E.M. Winkler, E.J. Wilhelm, Salt burst by hydration pressures in architectural stone in urban atmosphere, Geol. Soc. Am. Bull. 81 (1970) 567–572. https://doi.org/https://doi.org/10.1130/0016-7606(1970)81[567:SBBHPI]2.0.CO;2. [32] L. Ritterbach, P. Becker, Temperature and humidity dependent formation of CaSO4·xH2O (x = 0...2) phases, Glob. Planet. Change. 187 (2020). https://doi.org/10.1016/j.gloplacha.2020.103132. [33] E. Goossens, Moisture transfer properties of coated gypsum, Technische Universiteit Eindhoven, Faculteit Bouwkunde, 2003. https://doi.org/10.6100/IR571306. [34] C.A.C. Mesquita, Revestimientos Continuos Interiores de Varias Capas con Características de Barrera de Vapor e Higroscopicidad, Doctoral dissertation, Polytechnic University of Madrid, 2012. [35] R. Rubio Domene, Yeserías de la Alhambra. Historia, técnica y conservación, Patronato de la Alhambra y Generalife. University of Granada, Granada, 2010. [36] M. Torres-González, F.J. Alejandre, V. Flores-alés, A.I. Calero-castillo, F.J. Blasco-lópez, Analysis of the state of conservation of historical plasterwork through visual inspection and non-destructive tests . The case of the upper frieze of the Toledanos Room ( The Royal Alcázar of Seville , Spain ), J. Build. Eng. 40 (2021) 1–14. https://doi.org/https://doi.org/10.1016/j.jobe.2021.102314. [37] R. Campos de Alvear, The maintenance and the preventive preservation measures of the cultural goods in the Royal Alcázar of Seville, Apunt. Del Alcázar Sevilla. 18 (2018) 71– 87. [38] E. Correa Gómez, R. Rubio Domene, El yeso. Las decoraciones de yeso en época nazarí, in: P. de la A. y Generalife (Ed.), Man. Buenas Prácticas. Restauración Madera, Yeso y 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Cerámica, Patronato, Consejería de Educación, Cultura y Deporte de la Junta de Andalucía, Granada, 2014: pp. 43–52. [39] J.M. Cabrera Garrido, La influencia de los contaminantes en el Patrimonio artístico Nacional, Econ. Ind. 107 (1972) 51–60. [40] H. Viitanen, T. Ojanen, Improved Model to Predict Mold Growth in Building Materials, in: Therm. Perform. Exter. Envel. Whole Build. X–Proceedings CD, 2007: pp. 2–7. [41] G.A. Tsongas, F. Rioroan, Minimum conditions for visible mold growth, ASHRAE J. 58 (2016) 32–43. [42] D. Sanz Arauz, Análisis del yeso empleado en revestimientos exteriores mediante técnicas geológicas, Doctoral dissertation, Universidad Politécnica de Madrid, 2009. http://oa.upm.es/1711/1/DAVID_SANZ_ARAUZ.pdf. [43] A.I. Calero Castillo, A. García Bueno, O. López Cruz, V.J. Medina Flórez, La policromía original de las yeserías del Patio de las Doncellas del Real Alcázar de Sevilla. Materiales constitutivos y técnicas de ejecución, Arqueol. y Territ. Mediev. 24 (2017) 255–290. https://doi.org/10.17561/aytm.v24i0.9. [44] C. Sabbioni, P. Brimblecombe, M. Cassar, The Atlas of Climate Change Impact on European Cultural Heritage: Scientific Analysis and Management Strategies, Anthem Press, London, UK, 2010. [45] A. Almagro Gorbea, El Alcázar de Sevilla Un palacio musulmán para un rey cristiano, in: Cris. y Musulmanes En La Península Ibérica La Guerr. La Front. y La Convivencia. XI Congr. Estud. Mediev., 2007: pp. 331–365. [46] F.J. Blasco López, Yeserías medievales de tradición islámica del Real Alcázar de Sevilla: Revisión Historiográfica, Metodología para la caracterización, evaluación de su durabilidad y elaboración de un inventario, Doctoral dissertation, University of Seville, 2011. [47] A.I. Calero Castillo, Materiales, técnicas y procedimientos en la decoración arquitectónica. Aplicaciones a la conservación y restauración de las yeserías del Patio de las Doncellas. Real Alcázar de Sevilla., Doctoral dissertation, Universidad de Granada, 2016. http://hdl.handle.net/10481/43864. [48] A. Pleguezuelo, Tile-work in the mudéjar palace in the Royal Alcázar of Seville. A Preliminary Visual Analysis., Apunt. Del Alcázar Sevilla. 16 (2015) 219–230. [49] C. Enríquez Díaz, J.R. Baeza Álvarez, A project for the restoration of the tilings of the ground floor of the mudéjar palace, Apunt. Del Alcázar Sevilla. 19 (2019) 65–77. [50] C. Cañas Palop, Las armaduras de cubiertas mudéjares del palacio de Pedro I, del Alcázar de Sevilla: análisis integral y propuestas para la restauración, Doctoral dissertation, University of Seville, 2006. https://dialnet.unirioja.es/servlet/tesis?codigo=23333. [51] S. Fernández Aguilera, Portaventaneros mudéjares en el Real Alcázar de Sevilla, Archivo Hi, Diputación Provincial de Sevilla, Sevilla, 2012. [52] F.M. Tubino, Estudios sobre el arte en España. La arquitectura hispano-visigoda y árabe española. El Alcázar de Sevilla. Una iglesia mozárabe, 1886. [53] The Government of Spain, Royal Decree 314/2006. Approving the Spanish Technical Building Code, Madrid, Spain, 2013. [54] M. Kottek, J. Grieser, C. Beck, B. Rudolf, F. Rubel, World map of the Köppen-Geiger 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 climate classification updated, Meteorol. Zeitschrift. 15 (2006) 259–263. https://doi.org/10.1127/0941-2948/2006/0130. [55] L. Webster, J. Bradford, Measurement and Verification for Federal Energy (M&V Guidelines), 2008. [56] ANSI/ASHRAE, ASHRAE Guideline 14-2002 Measurement of Energy and Demand Savings, Ashrae. 8400 (2002) 170. [57] G.R. Ruiz, C.F. Bandera, Validation of calibrated energy models: Common errors, Energies. 10 (2017). https://doi.org/10.3390/en10101587. [58] Efficiency Valuation Organization, International Performance Measurement & Verification Protocol, Handb. Financ. Energy Proj. I (2016) 122. [59] M.F. Jentsch, A.S. Bahaj, P.A.B. James, CCWorldWeatherGen, Climate change world weather file generator, Version 1.8, Sustain. Energy Res. Gr. (2013). [60] I.P. on C. Change, Summary for Policymakers, in: Intergovernmental Panel on Climate Change (Ed.), Clim. Chang. 2013 - Phys. Sci. Basis, Cambridge University Press, Cambridge, 2014: pp. 1–30. https://doi.org/10.1017/CBO9781107415324.004. [61] J.R. Quinlan, others, Learning with continuous classes, in: 5th Aust. Jt. Conf. Artif. Intell., 1992: pp. 343–348. [62] Y. Wang, I.H. Witten, Induction of model trees for predicting continuous classes, in: Eur. Conf. Mach. Learn., Prague: University of Economics, Faculty of Informatics and Statistics, 1997. [63] A. Behnood, V. Behnood, M. Modiri Gharehveran, K.E. Alyamac, Prediction of the compressive strength of normal and high-performance concretes using M5P model tree algorithm, Constr. Build. Mater. 142 (2017) 199–207. https://doi.org/10.1016/j.conbuildmat.2017.03.061. [64] L. Lin, Q. Wang, A.W. Sadek, A combined M5P tree and hazard-based duration model for predicting urban freeway traffic accident durations, Accid. Anal. Prev. 91 (2016) 114– 126. https://doi.org/10.1016/j.aap.2016.03.001. [65] F. Afsarian, A. Saber, A. Pourzangbar, A.G. Olabi, M.A. Khanmohammadi, Analysis of recycled aggregates effect on energy conservation using M5″ model tree algorithm, Energy. 156 (2018) 264–277. https://doi.org/10.1016/j.energy.2018.05.099. [66] C.F. Jeffrey Kuo, C.H. Lin, M.H. Lee, Analyze the energy consumption characteristics and affecting factors of Taiwan’s convenience stores-using the big data mining approach, Energy Build. 168 (2018) 120–136. https://doi.org/10.1016/j.enbuild.2018.03.021. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Figure 1 Click here to access/download;Figure (Do not upload picture, .svg format. Upload only PDF, eps or tiff format);Figure 01.jpg Figure 2 Click here to access/download;Figure (Do not upload picture, .svg format. Upload only PDF, eps or tiff format);Figure 02.jpg Figure 3 Click here to access/download;Figure (Do not upload picture, .svg format. Upload only PDF, eps or tiff format);Figure 03.jpg Figure 4 Click here to access/download;Figure (Do not upload picture, .svg format. Upload only PDF, eps or tiff format);Figure 04.jpg Figure 5 Click here to access/download;Figure (Do not upload picture, .svg format. Upload only PDF, eps or tiff format);Figure 05.jpg Figure 6 Click here to access/download;Figure (Do not upload picture, .svg format. Upload only PDF, eps or tiff format);Figure 06.jpg Figure 13 Click here to access/download;Figure (Do not upload picture, .svg format. Upload only PDF, eps or tiff format);Figure 13.jpg Figure 14 Click here to access/download;Figure (Do not upload picture, .svg format. Upload only PDF, eps or tiff format);Figure 14.jpg Table 1. Summary of environmental parameters from 01/03 /2020 to 28 / 02/2021. Months RELATIVE HUMIDITY (%) TEMPERATURE (° C) Min. Values registered Max. values registered Δ (maxmin) Min. Values registered Max. values registered Δ (maxmin) Averag e SD Averag e SD Averag e SD Averag e SD Feb 20 55.3 11,001 73.8 9,454 18.5 14 1,319 18.1 1,577 4.1 Ap. 20 60.6 6,672 77.7 4.79 17.1 15.7 1,534 19.2 1,558 3.5 May 20 43.3 10,818 64.1 9,947 20.8 21.3 2,337 26.1 2,832 4.8 Jun 20 41.4 5,608 62 7.87 20.6 22.3 2,434 27.3 2,314 5 Jul. 20 38 5,877 56 7.43 18 27.6 1,448 31.8 1,328 4.2 Aug. 20 38 8,637 56.5 7,966 18.5 26.3 1,675 30.5 1,581 4.2 Sept. 20 44.1 9,493 58.3 8,287 14.2 24 1.9 27.4 2,105 3.4 Oct. 20 49.3 12,985 63.1 11,674 13.8 18.1 1,996 21.3 2,136 3.2 Nov. 20 70 8,459 78.6 7,134 8.6 15.6 1,964 17.9 1,965 2.3 Dec. 20 74.2 10,822 84.4 8,133 10.2 11.4 2,366 13.7 1,935 2.3 Jan. 21 75.2 9,545 86.1 8,141 10.9 8.6 3,493 11.2 3,093 2.6 Feb. 21 77.4 5,562 83.1 5,129 5.7 12.9 1,127 15.1 0.854 2.2 Table 1 Click here to access/download;Table;Table 1.docx Table 2. Calibration of measured data in RAS compared to CTE. Limit values according to ASHRAE Guideline14. Variable NMBE [%] CV (RMSE) [%] Obtained values Limit values Obtained values Limit values Temperature (T) 11.03 ± 10 25.34 30 Relative humidity (RH) 15.52 ± 10 27.08 30 Table 2 Click here to access/download;Table;Table 2.docx Table 3. Performance obtained by the M5P models in the training and validation phase. Approach Output variable Determination coefficient (R2) MBE M5P 1 Temperature (T) 86.53% 2.6 M5P 2 Relative Humidity (RH) 74.93% 8.0 Table 3 Click here to access/download;Table;Table 3.docx Declaration of interests x The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Journal of Building Engineering Dear Editor: Manuscript title: “LONG-TERM MONITORING FOR PREVENTIVE CONSERVATION OF HISTORICAL PLASTERWORKS” As corresponding author, I certify that the co-authors of this manuscript have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript. Vicente Flores-Alés Universidad de Sevilla As corresponding author, I certify that the co-authors of this manuscript have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript. Conflict of Interest AUTHOR STATEMENT M. Torres-González: Investigation, data discussion, Writing, Review & Editing C.Rubio-Bellido: Conceptualization, Investigation, data discussion, Validation D. Bienvenido-Huertas: Conceptualization, Investigation, Software, data discussion J.M. Alducin-Ochoa:, Investigation, Statistical analysis, Formal analysis V. Flores-Alés: Conceptualization, Methodology, Investigation, Validation Author Statement LONG-TERM ENVIRONMENTAL MONITORING FOR PREVENTIVE CONSERVATION OF EXTERNAL HISTORICAL PLASTERWORKS Marta Torres-González a,b *, Carlos Rubio-Bellido a, David Bienvenido-Huertas a, JM AlducinOchoa a, V. Flores-Alés a, a Department of Architectural Construction II, University of Seville b Department of Civil Engineering, Architecture and Georesources, University of Lisbon * Corresponding author: [email protected] , Av Reina Mercedes 4, Seville, 41012 KEYWORDS: plasterwork; temperature; relative humidity; architectural heritage; preservation; ABSTRACT: In this study, an analysis of the environmental conditions in a Csa climatic zone for the conservation of plasterwork of the Real Alcázar of Seville is carried out. The measurements obtained from in-situ monitoring are compared with the measurements provided by AEMET (State Meteorological Agency in Spain) during the reference year and the study is completed by estimating future environmental conditions using two alternative approaches: a morphing process from the EPW of the climatic zone and the application of M5P data mining algorithms. An optimal temperature range is established for the conservation of the plasterwork that prevents their dehydration or the freezing of water particles contained. The transformation of gypsum into bassanite, the risks associated with exposures to high relative humidity and the consequences of the slight hygroscopicity of the material and the environmental conditions that must be developed to favor the growth of mold on the surface or the cracking of polychromies that embellish these plaster decorations on numerous occasions are analyzed. The results obtained allow us to establish preventive conservation measures not only on the plasterwork but also on the Real Alcázar of Seville and that architectural heritage located in the subtropical dry-summer climate. Revised Manuscript with Changes Marked Click here to view linked References 1. INTRODUCTION 1.1 Thermo-hygrometric conditions and the conservation of the architectural heritage. The preventive conservation of heritage is crucial so that future generations can know the identity traits of their culture [1]. Therefore, the monitoring and analysis of specific environmental conditions [2] and their impact on architectural heritage, as well as future vulnerabilities due to global warming [3], are currently a challenge. There are numerous studies that analyze both the impact of environmental conditions on architectural heritage [4,5] and the conditioning of interior spaces to favor the preservation of objects (e.g., objects that belong to exhibitions) [6,7]. In both cases, there are national and international guidelines and standards that establish the interior air conditioning conditions of these buildings for such purposes, such as UNI 10829 Standard [8], UNE EN 15757 [9], ASHRAE STANDARD 55-2020[10] or the British PAS 198 [11]. Likewise, EN 15759-1 [12] also considers the possibility of using heating systems to achieve thermal comfort and for the optimal conservation of heritage elements. In addition to international standards, numerous studies have carried out thermo-hygrometric monitoring of architectural heritage in recent decades, resulting in a very useful tool to assess the state of conservation of the building and to know the causes of the degradation in order to determine a future restoration intervention or to establish preventive conservation criteria, ruling out the use of destructive techniques. Thus, long-term studies of up to 20 years duration [13,14] and studies in which monitoring is limited to one year [15] have been published. In this regard, it is common to evaluate at least annual monitoring periods that are complemented by thermal models on which different estimates of environmental conditions are established. Environmental monitoring and simulation tools make it possible to establish estimates of different factors, such as the modification of operational patterns, thermal comfort [16-18], or energy conservation measures [19]. However, the impact of climate change, the basis of numerous investigations related to cities [20,21] has been sparsely discussed in reference to preventive conservation of heritage and long-term adverse aspects, which could become very significant [22,23]. 1.2 Relative humidity and temperature influence in the preservation of historical plasterwork. There are studies about how ambient conditions could affect constructive materials [24–28] but are insufficient with respect to gypsum or plasterwork. In this sense, it is worth highlighting the importance of follow-up campaigns to identify possible risks for the conservation of the plasterwork, prevent irreversible damage and advance conservation strategies. However, to account for all the possible damage mechanisms, a comprehensive methodology should be used to evaluate some risk indexes based on ambient temperature (T) and relative humidity (RH) measurements. Gypsum, due to its porous microstructure formed by crystalline groups of hydrated calcium sulfate, can incorporate water molecules inside, in the form of water vapor or even in the form of liquid water, when environmental conditions favor capillary condensation. Furthermore, gypsum has the ability to eliminate this water when environmental conditions change, reducing the content of water vapor in contact with the material [29]. Variations in T and RH can induce phase transitions. In general, gypsum (CaSO4·2H2O) dehydration occurs in dry environments [30], which is the reason why temperature below 40°C is desirable for the correct preservation of plasterwork. However, there may be situations of high temperatures combined with high relative humidity in which there are no alterations in the mineral phases of the gypsum [31]. In this regard, Winkler & Wilhelm (1970) indicated the relationship that must exist between ambient temperature and relative humidity to transform gypsum (CaSO4·2H2O) into bassanite (CaSO4·0.5H2O) [31], In this case, the material is likely to The control and regulation of environmental parameters within a reference interval around average values or typical seasonal cycles of the historical RAS climate will reduce the risk of physical-mechanical damage and will also allow to establish mitigation actions. On the ground floor of the Palace, the interior areas are not physically closed by carpentry or locksmiths and are permanently open; there are large access arches that allow transit between different rooms, as well as between these and the outdoor spaces. In this way, two areas according to exposure to environmental conditions could be defined. An area that includes those plasterworks that receive direct incidence of solar radiation has been considered (yellow zone) (Fig. 5-A), and another zone with those plasterworks not exposed to direct sunlight and with a higher relative humidity (red zone), being the most representative and unfavorable case in terms of the environmental parameters (Fig. 5-B). Fig. 5 A) Solar path on 03/20/2020 in the geographical coordinates corresponding to the Palace of Pedro I (Retrieved from https://www.sunearthtools.com/) B) Two different zones on the ground floor of the Pedro I Palace according to sunlight. Additionally, it has been possible to verify that the humidity and temperature values measured in the interior rooms of the Palace of Pedro I [36] are very similar to those obtained in the southeast gallery of the Courtyard of the Maidens (exterior) because this area does not receive the direct incidence of solar radiation (Fig. 5-A). 2.2. Measurement procedure To determine ambient temperature and relative humidity conditions, a Lascar thermohygrometer, model EL-USB 2 LCD with two recording channels, within the range -35ºC to 80ºC and 0% RH to 100% RH, with a resolution of ± 0.5ºC and ± 3% RH was used. In this case, the measurements have been carried out over a period of one year, at uninterrupted one-hour intervals from March 2020 to February 2021 with a thermo-hygrometer placed in the southeast gallery of the Courtyard of the Maidens (Fig. 6). The outdoor measurements were obtained through the AEMET (State Meteorological Agency in Spain). The meteorological station used was the one located in Seville. This meteorological station is equipped with a VAISALA HMP45D probe with a measuring range from 40 to 60 ºC and 0.8 to 100% RH, and an accuracy of ± 0.2°C and ± 2% RH. The measurement data obtained are also from one-hour intervals. To determine the interval of variation of RH by months, the average of the minimum and maximum daily humidities has been calculated and the monthly arithmetic means of these values have been obtained. These data have been studied following the indications of the UNEEN 15757: 2011 standard[9]. Fig. 6 Location of the monitoring equipment on the ground floor of the Pedro I Palace. 2.3. Obtaining future climate data According to the data obtained during the annual monitoring, two approaches have been carried out to obtain future time series: (i) estimation of future outdoor climate data through the EPW morphing process of the climate zone of the case of study, and (ii) the estimation of the time series of the interior point through the application of data mining algorithms. 2.3.1. Morphing from the EPW of the climate zone of the case study The first of them was based on the climatic files provided by the Spanish Technical Building Code (CTE)[53] for the climate zone of Seville, Csa (Mediterranean) according to Köppen – Geiger classification [54]. In the first place, T and RH data resulting from the monitoring have been entered in the Seville climate archive to be able to carry out the evaluation of future scenarios. Secondly, T and RH data measured outdoors have been compared with those existing in the CTE to see if it is feasible to establish considerations, not only in the case of study but also in plasterwork in similar "subtropical dry-summer" climates. The calibration criteria of the Federal Energy Management Program (FEMP)[55], ASHRAE GUIDELINE Standard 14-2002 [56], [57] and the International Performance Measurement and Verification Protocol (IPMVP) [58] have been followed. This validation process makes it possible to verify if the results of the estimations for the climate zone of Seville are related to the real values monitored with a margin of error. For this, the following formulas are used: 𝑀𝐵𝐸 =∑(𝑚𝑖−𝑠𝑖) 𝑛 𝑖=1 𝑛 (1) 𝑀𝐵𝐸 =∑(𝑚𝑖−𝑠𝑖) 𝑛 𝑖=1 𝑛 (2) 𝑁𝑀𝐵𝐸 =1 𝑚 ∙∑(𝑚𝑖−𝑠𝑖) 𝑛 𝑖=1 𝑛−𝑝 (3) In order to scale the results of the Mean Bias Error (MBE) (Equation 1), Normalized Mean Bias Error (NMBE) is used by dividing MBE by the mean of measured values (𝑚), giving the global difference between the real values and the predicted ones (Equation 2); In these equations mi is the measured value, si is the simulated one and n the number of measured data points. The Coefficient of Variation of the Root Mean Square Error CV (RMSE) measures the variability of the errors between measured and simulated values (Equation 3). The limit values associated with these parameters vary depending on whether they are hourly or monthly. Considering that the environmental variables are hourly, the limit values are between -10% and 10% in NMBE and less than 30% in CV (RMSE) [56,57]. To estimate future climate scenarios in the climate zone, the CCWorldWeatherGen tool from the UK Met Office Hadley Center Coupled Model 3 HadCM3 has been used. Through a morphing process, this tool generates weather files adapted to climate change from any location in the world and creates files compatible with most building performance simulation programs [59]. The morphing of the climate files has been carried out for scenario A2 of greenhouse gas emissions according to the IPCC [60], resulting in climate scenario files for years 2050 and 2080. 2.3.2. Estimation of future time series of the case study The data mining approach of this study was performed using the M5P algorithm. This algorithm allows estimating the time series for the years 2050 and 2080 in the measured interior point of the case study. The M5P algorithm (also known as M5') is an evolution of the Classification and regression tree (CART) [61,62]. Unlike CART, M5P combines decision trees with multivariate regression; a decision tree is built following the same inverted tree structure offered by CART, but a multiple linear regression (MLR) model is fitted on each sheet (Equation 4). Therefore, the algorithm works by developing an MLR model in each subregion (Fig. 7). In the process of developing the M5P algorithm tree, instead of maximizing the information gain, the internal variation of the subsets for the class values of each branch is minimized. Once the model is built, pruning reduces overfitting. The advantages of the models generated by this algorithm are that they efficiently handle large amounts of numerical variables and are robust in the absence of values in the instances of the analyzed dataset [63,64]. Thus, its use for the characterization of different aspects of buildings has increased in recent years [65,66]. 𝑌 𝑀𝐿𝑅 = 𝛽0+∑ (𝛽𝑖𝑥𝑖) 𝑣 𝑖=1 + 𝜀 (4) Where 𝛽0 is the independent term, 𝛽𝑖 are the regression coefficients, 𝑥𝑖 are the predictor variables, and 𝜀 is the error. Fig. 7. Scheme of an M5P model. The approach designed had a configuration of 6 input variables corresponding to the hourly values of temperature, relative humidity, dew point temperature, horizontal infrared radiation intensity, global horizontal radiation, and direction normal solar irradiance, diffuse horizontal solar irradiance of the time series of the outer point of Seville. The output variables correspond to the relative humidity and the temperature at the interior point of the case study. For each output variable a different M5P was designed. The training of each M5P model was carried out using the data from the current scenario. Once the models were trained, time series of the years 2050 and 2080 were estimated. It is worth noting that the M5P models were validated using the coefficient of determination and the MBE index as statistical parameters. 3. RESULTS AND DISCUSSION 3.1 Thermo-hygrometric conditions of the Palace of Pedro I The results obtained from the monitoring carried out between March 2020 and February 2021 in the southeast gallery of the Courtyard of the Maidens are shown in Fig. 8 and table 1: Fig. 8 Environmental conditions in the southeast corner of the Courtyard of the Maidens from March 2020 to February 2021. The measurement of the T and RH parameters is relevant mainly for the study of the fluctuations and thermal jumps that may occur during the day-night hours or during the different seasons of the year. In this regard, it is observed that the most pronounced thermal jumps between day and night hours do not exceed 5ºC (July and August), a fact that -by itself- does not represent any risk in terms of material expansion. Finally, a high RH in winter is recorded and must be controlled with adequate ventilation, although it is within the acceptable range for the correct preservation of plasterwork. Table 1. Summary of environmental parameters from 01/03 /2020 to 28 / 02/2021. Months RELATIVE HUMIDITY (%) TEMPERATURE (° C) Min. Values registered Max. values registered Δ (max-min) Min. Values registered Max. values registered Δ (max-min) Average SD Average SD Average SD Average SD Feb 20 55.3 11,001 73.8 9,454 18.5 14 1,319 18.1 1,577 4.1 Ap. 20 60.6 6,672 77.7 4.79 17.1 15.7 1,534 19.2 1,558 3.5 May 20 43.3 10,818 64.1 9,947 20.8 21.3 2,337 26.1 2,832 4.8 Jun 20 41.4 5,608 62 7.87 20.6 22.3 2,434 27.3 2,314 5 Jul. 20 38 5,877 56 7.43 18 27.6 1,448 31.8 1,328 4.2 Aug. 20 38 8,637 56.5 7,966 18.5 26.3 1,675 30.5 1,581 4.2 Sept. 20 44.1 9,493 58.3 8,287 14.2 24 1.9 27.4 2,105 3.4 Oct. 20 49.3 12,985 63.1 11,674 13.8 18.1 1,996 21.3 2,136 3.2 Nov. 20 70 8,459 78.6 7,134 8.6 15.6 1,964 17.9 1,965 2.3 Dec. 20 74.2 10,822 84.4 8,133 10.2 11.4 2,366 13.7 1,935 2.3 Jan. 21 75.2 9,545 86.1 8,141 10.9 8.6 3,493 11.2 3,093 2.6 Feb. 21 77.4 5,562 83.1 5,129 5.7 12.9 1,127 15.1 0.854 2.2 Fig. 9 Graphical representation of ambient condition (T and RH) from 01/03 /2020 to 28 / 02/2021. From a detailed analysis of the measurements obtained (Table 1 and Fig 9), the risk due to high RH is concentrated from November to February, both inclusive; being January the most unfavorable month since 27.02% of the measurements recorded during that month exceed the limit of 90% RH. Likewise, the risk of dehydration of the gypsum and transformation into bassanite (CaSO4·1/2H2O) occurs from May to October, both inclusive; being July and August the most critical months for such purposes. It is worth to highlight that a 56.71% of the measurements recorded in the year remain below the limit established in the matrix (Fig. 10). The data obtained correspond to a total of 8,760 measurements, of which only 11.76% are outside the limits established in the matrix that relates thermo-hygrometric conditions with the mineralogical phases of calcium sulfate, depending on their degree of hydration. Fig. 10 Dispersion graph of the measurements from 01/03 /2020 to 28 / 02/2021 at the southeast corner of the Courtyard of the Maidens. Regarding the type of risk detected, it can be seen in Fig. 10 how it is mainly distinguished between risk due to internal structural transformations (associated with the summer months) and risk due to high RH (associated with the winter months). The 7.95% of the hourly records correspond to measurements that are below the curve established by Winkler& Wilhelm [31] , while the risk due to a high RH corresponds to 3.65% and the risk due to a low T corresponds to only 10 observations throughout the year, so the current climatic conditions in the Courtyard of the Maidens are favorable in this sense. Fig. 11 Monthly representation of the hourly measurements obtained in the Courtyard of the Maidens from March 2020 to February 2021. However, the results obtained from monitoring the climate of the city of Seville show different trends in conservation risks (Fig. 11). As it can be seen, although the most prominent risks are once again that of internal transformations in summer and high RH in winter, there is a greater number of observations that remain within these risks. Thus, the observations at risk of internal transformations correspond to values between 348 and 360 in July and August, while in the Courtyard of the Maidens the values were 229 and 168, respectively. Likewise, it was also detected how the risk due to high RH extends to a greater number of months than in the case of the Courtyard of the Maidens and with a significantly higher number of observations. Fig. 12 Monthly representation of the hourly measurements obtained in Seville from March 2020 to February 2021. Therefore, significant variations are detected in the results obtained with the two approaches. While the approach to the analysis of the Courtyard of the Maidens point (with a clear microclimate) presents less severe conditions for the conservation of plasterwork, outside the conditions would be different, presenting a greater risk due to the climate in these elements. Likewise, the variability detected between the two approaches shows the great variation that the analysis approaches of the influence of climate can present on the conservation of plasterwork. Thus, the use of a joint approach to the outdoor climate and microclimates in this type of case study allows us to show a broader perspective of the risk of conservation of plasterwork. 3.2 Calibration of outdoor measurements. The climate file of Seville. Despite the possible risks detected in the current scenario, climate change can generate a change in the conservation conditions of plasterwork. For that, the future time series for the two previously validated approaches used were estimated in this study. In the case of the outdoor climate approach, T and RH parameters measured during the outdoor monitoring are compared with T and RH data extracted from the CTE for the climate zone of Seville Csa (Mediterranean). As the environmental variables analyzed are hourly, the limit values are between -10% and 10% in NMBE and less than 30% in CV (RMSE). According to Table 2, it is detected that the CV (RMSE) values are within the established error limits. However, it is detected that T is slightly higher (0.88%) than the limits in the NMBE and in matters related to RH the value exceeds the limit in the NMBE by 6.40% (Table 2). From these results we can establish the following considerations: In the first place, considering that a measurement is being compared under specific environmental conditions with a typical meteorological year according to the criteria established in the CTE and based on the available meteorological stations, the level of approximation is good, especially if we consider the CV (RMSE). 7. ACKNOWLEDGEMENTS This research has been carried out thanks to the financing of the project "DEVELOPMENT AND EVALUATION OF DURABILITY MODELS AND PREVENTIVE CONSERVATION OF DECORATIVE ELEMENTS FROM THE HISTORICAL PLASTERS OF THE ROYAL ALCÁZAR DE SEVILLA" (PGC2018-093470-B-I00) by the Ministry of Science, Innovation and Universities of the Government of Spain and the financing received from the VI PPIT-2021-I.3 from the University of Seville. The authors wish to thank the collaboration of the Board of Trustees of the Real Alcázar of Seville for the facilities provided to achieve visual inspection and measurement of environmental parameters. 8. REFERENCES [1] E. Lucchi, Review of preventive conservation in museum buildings, J. Cult. Herit. 29 (2018) 180–193. https://doi.org/10.1016/j.culher.2017.09.003. [2] D. Camuffo, Microclimate for Cultural Heritage: Conservation, Restoration, and Maintenance of Indoor and Outdoor Monuments, Elsevier, 2019. [3] I. Cook, R. Johnston, K. Selby, Climate Change and Cultural Heritage: A Landscape Vulnerability Framework, J. Isl. Coast. Archaeol. (2019). https://doi.org/10.1080/15564894.2019.1605430. [4] C. Ferreira, J. Barrelas, A. Silva, J. de Brito, I.S. Dias, I. Flores-Colen, Impact of environmental exposure conditions on the maintenance of facades’ claddings, Buildings. 11 (2021). https://doi.org/10.3390/buildings11040138. [5] V. Costanzo, K. Fabbri, E. Schito, M. Pretelli, L. Marletta, Microclimate monitoring and conservation issues of a Baroque church in Italy: a risk assessment analysis, Build. Res. Inf. 0 (2021) 1–19. https://doi.org/10.1080/09613218.2021.1899797. [6] S.P. Corgnati, M. Filippi, Assessment of thermo-hygrometric quality in museums: Method and in-field application to the “ Duccio di Buoninsegna” exhibition at Santa Maria della Scala (Siena, Italy), J. Cult. Herit. 11 (2010) 345–349. https://doi.org/10.1016/j.culher.2009.05.003. [7] M. Zarzo, A. Fernández-Navajas, F.J. García-Diego, Long-term monitoring of fresco paintings in the cathedral of Valencia (Spain) through humidity and temperature sensors in various locations for preventive conservation, Sensors. 11 (2011) 8685–8710. https://doi.org/10.3390/s110908685. [8] N. Italiana, UNI 10829. Works of art of historical importance. Ambient conditions for the conservation. Measurement and analysis. [Beni di interesse storico e artistico. Condizioni ambientali di conservazione. Misurazione ed analisi.], (1999) 1–24. [9] A.E. de N. UNE, UNE-EN 15757: Conservation of Cultural Property - Specifications for temperature and relative humidity to limit climate-induced mechanical damage in organic hygroscopic materials, (2011). [10] Ansi/Ashrae, ANSI/ASHRAE 55:2004 Thermal Environmental Conditions for Human Occupancy, Ashrae. (2004) 30. [11] PAS 198:2012 - Specification for managing environmental conditions for cultural collections – BSI British Standards, (n.d.). http://shop.bsigroup.com/ProductDetail/?pid=000000000030219669. [12] UNE-EN-15759-1, Conservation of cultural property. Indoor climate. Part 1: Guidelines for heating churches, chapels and other places of worship, (2012). [13] C. Bonacina, P. Baggio, F. Cappelletti, P. Romagnoni, A.G. Stevan, The Scrovegni Chapel: The results of over 20 years of indoor climate monitoring, Energy Build. 95 (2015) 144–152. https://doi.org/10.1016/j.enbuild.2014.12.018. [14] D. Bienvenido-Huertas, M. León-Muñoz, J.J. Martín-del-Río, C. Rubio-Bellido, Analysis of climate change impact on the preservation of heritage elements in historic buildings with a deficient indoor microclimate in warm regions, Build. Environ. 200 (2021) 107959. https://doi.org/10.1016/j.buildenv.2021.107959. [15] G. Zonno, R. Aguilar, R. Boroschek, P.B. Lourenço, Analysis of the long and short-term effects of temperature and humidity on the structural properties of adobe buildings using continuous monitoring, Eng. Struct. 196 (2019) 109299. https://doi.org/10.1016/j.engstruct.2019.109299. [16] R.P. Kramer, M.P.E. Maas, M.H.J. Martens, A.W.M. van Schijndel, H.L. Schellen, Energy conservation in museums using different setpoint strategies: A case study for a state-ofthe-art museum using building simulations, Appl. Energy. 158 (2015) 446–458. https://doi.org/10.1016/j.apenergy.2015.08.044. [17] M. Napp, T. Kalamees, Energy use and indoor climate of conservation heating, dehumidification and adaptive ventilation for the climate control of a mediaeval church in a cold climate, Energy Build. 108 (2015) 61–71. https://doi.org/10.1016/j.enbuild.2015.08.013. [18] C. Cornaro, V.A. Puggioni, R.M. Strollo, Dynamic simulation and on-site measurements for energy retrofit of complex historic buildings: Villa Mondragone case study, J. Build. Eng. 6 (2016) 17–28. https://doi.org/10.1016/j.jobe.2016.02.001. [19] T. Cardinale, G. Rospi, N. Cardinale, The influence of indoor microclimate on thermal comfort and conservation of artworks: The case study of the Cathedral of Matera (South Italy), in: Energy Procedia, Elsevier Ltd, 2014: pp. 425–432. https://doi.org/10.1016/j.egypro.2014.10.398. [20] A. Revi, D. Satterthwaite, F. Aragon-Durand, J. Corfee-Morlot, R.. R. Kiunsi, M. Pelling, D. Roberts, W. Solecki, S.P. Gajjar, A. Sverdlik, Towards transformative adaptation in cities: the IPCC’s Fifth Assessment, Environ. Urban. 26 (2014) 11–28. https://doi.org/10.1177/0956247814523539. [21] A.J. Prieto, K. Verichev, A. Silva, J. de Brito, On the impacts of climate change on the functional deterioration of heritage buildings in South Chile, Build. Environ. 183 (2020) 107138. https://doi.org/10.1016/j.buildenv.2020.107138. [22] V. Rajčić, A. Skender, D. Damjanović, An innovative methodology of assessing the climate change impact on cultural heritage, Int. J. Archit. Herit. 12 (2018) 21–35. https://doi.org/10.1080/15583058.2017.1354094. [23] A. Haugen, C. Bertolin, G. Leijonhufvud, T. Olstad, T. Broström, A methodology for long-term monitoring of climate change impacts on historic buildings, Geosci. 8 (2018). https://doi.org/10.3390/geosciences8100370. [24] S. Ramírez, M. Zarzo, A. Perles, F.J. García-Diego, A methodology for discriminant time series analysis applied to microclimate monitoring of fresco paintings, Sensors (Switzerland). 21 (2021) 1–29. https://doi.org/10.3390/s21020436. [25] M. Gómez Heras, La Temperatura en los Materiales del Patrimonio, (2012) 87–95. http://digital.csic.es/handle/10261/46794. [26] Z. Pavlík, J. Fořt, R. Černý, An in situ monitoring system for the study of environmental influences on durability and the destructive process of building materials and structures, WIT Trans. Modelling Simul. 55 (2013) 287–296. https://doi.org/10.2495/CMEM130231. [27] S. Michalski, Temperatura Incorrecta, Can. Conserv. Institute. ICCROM. (2009). [28] S. Michalski, Humedad relativa Incorrecta, Can. Conserv. Institute. ICCROM. (2009). [29] L. Villanueva Domínguez, A. García Santos, Manual del yeso, CIE Inversiones Editoriales, 2001. [30] D. Freyer, W. Voigt, Crystallization and Phase Stability of CaSO4 and CaSO 4 - Based Salts, Monatshefte Fur Chemie. 134 (2003) 693–719. https://doi.org/10.1007/s00706003-0590-3. [31] E.M. Winkler, E.J. Wilhelm, Salt burst by hydration pressures in architectural stone in urban atmosphere, Geol. Soc. Am. Bull. 81 (1970) 567–572. https://doi.org/https://doi.org/10.1130/0016-7606(1970)81[567:SBBHPI]2.0.CO;2. [32] L. Ritterbach, P. Becker, Temperature and humidity dependent formation of CaSO4·xH2O (x = 0...2) phases, Glob. Planet. Change. 187 (2020). https://doi.org/10.1016/j.gloplacha.2020.103132. [33] E. Goossens, Moisture transfer properties of coated gypsum, Technische Universiteit Eindhoven, Faculteit Bouwkunde, 2003. https://doi.org/10.6100/IR571306. [34] C.A.C. Mesquita, Revestimientos Continuos Interiores de Varias Capas con Características de Barrera de Vapor e Higroscopicidad, Doctoral dissertation, Polytechnic University of Madrid, 2012. [35] R. Rubio Domene, Yeserías de la Alhambra. Historia, técnica y conservación, Patronato de la Alhambra y Generalife. University of Granada, Granada, 2010. [36] M. Torres-González, F.J. Alejandre, V. Flores-alés, A.I. Calero-castillo, F.J. Blasco-lópez, Analysis of the state of conservation of historical plasterwork through visual inspection and non-destructive tests . The case of the upper frieze of the Toledanos Room ( The Royal Alcázar of Seville , Spain ), J. Build. Eng. 40 (2021) 1–14. https://doi.org/https://doi.org/10.1016/j.jobe.2021.102314. [37] R. Campos de Alvear, The maintenance and the preventive preservation measures of the cultural goods in the Royal Alcázar of Seville, Apunt. Del Alcázar Sevilla. 18 (2018) 71– 87. [38] E. Correa Gómez, R. Rubio Domene, El yeso. Las decoraciones de yeso en época nazarí, in: P. de la A. y Generalife (Ed.), Man. Buenas Prácticas. Restauración Madera, Yeso y Cerámica, Patronato, Consejería de Educación, Cultura y Deporte de la Junta de Andalucía, Granada, 2014: pp. 43–52. [39] J.M. Cabrera Garrido, La influencia de los contaminantes en el Patrimonio artístico Nacional, Econ. Ind. 107 (1972) 51–60. [40] H. Viitanen, T. Ojanen, Improved Model to Predict Mold Growth in Building Materials, in: Therm. Perform. Exter. Envel. Whole Build. X–Proceedings CD, 2007: pp. 2–7. [41] G.A. Tsongas, F. Rioroan, Minimum conditions for visible mold growth, ASHRAE J. 58 (2016) 32–43. [42] D. Sanz Arauz, Análisis del yeso empleado en revestimientos exteriores mediante técnicas geológicas, Doctoral dissertation, Universidad Politécnica de Madrid, 2009. http://oa.upm.es/1711/1/DAVID_SANZ_ARAUZ.pdf. [43] A.I. Calero Castillo, A. García Bueno, O. López Cruz, V.J. Medina Flórez, La policromía original de las yeserías del Patio de las Doncellas del Real Alcázar de Sevilla. Materiales constitutivos y técnicas de ejecución, Arqueol. y Territ. Mediev. 24 (2017) 255–290. https://doi.org/10.17561/aytm.v24i0.9. [44] C. Sabbioni, P. Brimblecombe, M. Cassar, The Atlas of Climate Change Impact on European Cultural Heritage: Scientific Analysis and Management Strategies, Anthem Press, London, UK, 2010. [45] A. Almagro Gorbea, El Alcázar de Sevilla Un palacio musulmán para un rey cristiano, in: Cris. y Musulmanes En La Península Ibérica La Guerr. La Front. y La Convivencia. XI Congr. Estud. Mediev., 2007: pp. 331–365. [46] F.J. Blasco López, Yeserías medievales de tradición islámica del Real Alcázar de Sevilla: Revisión Historiográfica, Metodología para la caracterización, evaluación de su durabilidad y elaboración de un inventario, Doctoral dissertation, University of Seville, 2011. [47] A.I. Calero Castillo, Materiales, técnicas y procedimientos en la decoración arquitectónica. Aplicaciones a la conservación y restauración de las yeserías del Patio de las Doncellas. Real Alcázar de Sevilla., Doctoral dissertation, Universidad de Granada, 2016. http://hdl.handle.net/10481/43864. [48] A. Pleguezuelo, Tile-work in the mudéjar palace in the Royal Alcázar of Seville. A Preliminary Visual Analysis., Apunt. Del Alcázar Sevilla. 16 (2015) 219–230. [49] C. Enríquez Díaz, J.R. Baeza Álvarez, A project for the restoration of the tilings of the ground floor of the mudéjar palace, Apunt. Del Alcázar Sevilla. 19 (2019) 65–77. [50] C. Cañas Palop, Las armaduras de cubiertas mudéjares del palacio de Pedro I, del Alcázar de Sevilla: análisis integral y propuestas para la restauración, Doctoral dissertation, University of Seville, 2006. https://dialnet.unirioja.es/servlet/tesis?codigo=23333. [51] S. Fernández Aguilera, Portaventaneros mudéjares en el Real Alcázar de Sevilla, Archivo Hi, Diputación Provincial de Sevilla, Sevilla, 2012. [52] F.M. Tubino, Estudios sobre el arte en España. La arquitectura hispano-visigoda y árabe española. El Alcázar de Sevilla. Una iglesia mozárabe, 1886. [53] The Government of Spain, Royal Decree 314/2006. Approving the Spanish Technical Building Code, Madrid, Spain, 2013. [54] M. Kottek, J. Grieser, C. Beck, B. Rudolf, F. Rubel, World map of the Köppen-Geiger climate classification updated, Meteorol. Zeitschrift. 15 (2006) 259–263. https://doi.org/10.1127/0941-2948/2006/0130. [55] L. Webster, J. Bradford, Measurement and Verification for Federal Energy (M&V Guidelines), 2008. [56] ANSI/ASHRAE, ASHRAE Guideline 14-2002 Measurement of Energy and Demand Savings, Ashrae. 8400 (2002) 170. [57] G.R. Ruiz, C.F. Bandera, Validation of calibrated energy models: Common errors, Energies. 10 (2017). https://doi.org/10.3390/en10101587. [58] Efficiency Valuation Organization, International Performance Measurement & Verification Protocol, Handb. Financ. Energy Proj. I (2016) 122. [59] M.F. Jentsch, A.S. Bahaj, P.A.B. James, CCWorldWeatherGen, Climate change world weather file generator, Version 1.8, Sustain. Energy Res. Gr. (2013). [60] I.P. on C. Change, Summary for Policymakers, in: Intergovernmental Panel on Climate Change (Ed.), Clim. Chang. 2013 - Phys. Sci. Basis, Cambridge University Press, Cambridge, 2014: pp. 1–30. https://doi.org/10.1017/CBO9781107415324.004. [61] J.R. Quinlan, others, Learning with continuous classes, in: 5th Aust. Jt. Conf. Artif. Intell., 1992: pp. 343–348. [62] Y. Wang, I.H. Witten, Induction of model trees for predicting continuous classes, in: Eur. Conf. Mach. Learn., Prague: University of Economics, Faculty of Informatics and Statistics, 1997. [63] A. Behnood, V. Behnood, M. Modiri Gharehveran, K.E. Alyamac, Prediction of the compressive strength of normal and high-performance concretes using M5P model tree algorithm, Constr. Build. Mater. 142 (2017) 199–207. https://doi.org/10.1016/j.conbuildmat.2017.03.061. [64] L. Lin, Q. Wang, A.W. Sadek, A combined M5P tree and hazard-based duration model for predicting urban freeway traffic accident durations, Accid. Anal. Prev. 91 (2016) 114– 126. https://doi.org/10.1016/j.aap.2016.03.001. [65] F. Afsarian, A. Saber, A. Pourzangbar, A.G. Olabi, M.A. Khanmohammadi, Analysis of recycled aggregates effect on energy conservation using M5″ model tree algorithm, Energy. 156 (2018) 264–277. https://doi.org/10.1016/j.energy.2018.05.099. [66] C.F. Jeffrey Kuo, C.H. Lin, M.H. Lee, Analyze the energy consumption characteristics and affecting factors of Taiwan’s convenience stores-using the big data mining approach, Energy Build. 168 (2018) 120–136. https://doi.org/10.1016/j.enbuild.2018.03.021.