- Email: [email protected]
GPR technique as a tool for cultural heritage restoration: San Miguel de los Reyes Hieronymite Monastery, 16th century (Valencia, Spain)
Journal of Cultural Heritage 8 (2007) 87e92 http://france.elsevier.com/direct/CULHER/
Case study
GPR technique as a tool for cultural heritage restoration: San Miguel de los Reyes Hieronymite Monastery, 16th century (Valencia, Spain) Francisco Garcı´a Garcı´a a,*, Manuel Ramı´rez Blanco c,1, Isabel Rodrı´guez Abad a,2, Rosa Martı´nez Sala b,3, Isabel Tort Ausina b,4, Javier Benlloch Marco c,1, Jose Luis Montalva´ Conesa b,4 a
Department of Cartographic Engineering, Geodesy and Photogrammetry, Polytechnic University of Valencia, Camino de Vera, s/n, 46022 Valencia, Spain b Department of Applied Physics, Polytechnic University of Valencia, Camino de Vera, s/n, 46022 Valencia, Spain c Department of Architectural Constructions, Polytechnic University of Valencia, Camino de Vera, s/n, 46022 Valencia, Spain Received 7 June 2006; accepted 11 October 2006
Abstract This paper describes GPR (ground penetrating radar) surveys performed inside the crypt of the San Miguel de los Reyes Monastery (1546e1835) in order to detect the exact location of its founders’ remains, the Dukes of Calabria (16th century). This Monastery was erected to house their family mausoleum and the bodies of the founders were buried near the high altar of the church (1645). However, in the 18th century, the tombs were exhumed to provide them with a worthier burial site: the crypt below the high altar. There is no documentation specifying the exact location of the tombs inside the crypt. Therefore, in order to reveal the exact location of the tombs the GPR survey was conducted inside the crypt. In our specific study, the available historical documentation led us to suppose that the Dukes of Calabria’s remains were inside their mausoleums. However, after having performed the GPR analysis, we discovered that the mausoleums were solid and not hollow. The project required data collection on four areas in the crypt: the altar crypt, the Fernando de Arago´n mausoleum, the Germana de Foix mausoleum and the floor between the two mausoleums and the altar. In this study, we have processed the GPR records in three different ways: the radargrams were processed in a standard manner, a detailed spectral analysis of all anomalous areas was carried out, and finally a 3D representation was generated. After this complete analysis we concluded that the bodies were not located inside their mausoleums, because they were shown to be solid. Besides, a burial site was located in the crypt subsurface near the Germana de Foix mausoleum, in which four different elements could be identified. Two of them may well be the tombs of the Dukes of Calabria and the other two the tombs of the Germana de Foix sisters. The results obtained in this survey are a good example of GPR application as an efficient and respectful tool for use in Cultural Heritage restoration studies, providing it with a very useful technique for similar projects such as those carried out in the restoration of historical buildings and those in which the elements to be examined are beneath a shallow coating of material. Ó 2007 Elsevier Masson SAS. All rights reserved. Keywords: Ground Penetrating Radar (GPR); Cultural heritage; Restoration and conservation; Hieronymite monastery; Burial element location; Spectral analysis; 3D representation
* Corresponding author. Tel.: þ34 963877557; fax: þ34 963877169. E-mail addresses: [email protected] (F. Garcı´a Garcı´a), [email protected] (M. Ramı´rez Blanco), [email protected] (I. Rodrı´guez Abad), rmsala@fis. upv.es (R. Martı´nez Sala), [email protected] (I. Tort Ausina), [email protected] (J. Benlloch Marco), [email protected] (J.L. Montalva´ Conesa). 1 Tel.: þ34 963877456; fax: þ34 963877129. 2 Tel.: þ34 963877557; fax: þ34 963877169. 3 Tel.: þ34 963877526; fax: þ34 963877975. 4 Tel.: þ34 963877121; fax: þ34 963879129. 1296-2074/$ - see front matter Ó 2007 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.culher.2006.10.005
88
F. Garcı´a Garcı´a et al. / Journal of Cultural Heritage 8 (2007) 87e92
1. Introduction Ground penetrating radar (GPR) surveys have proven to be an increasingly useful and non-destructive tool for the identification of burial structures, since with this technique we are able to identify the spatial position of structures of interest such as tombs, cavities and buried elements [1e5]. This paper describes GPR surveys performed inside the San Miguel de los Reyes Monastery (Valencia, Spain). The aim of this work was to detect the exact location of its founders’ remains, the Dukes of Calabria, Fernando de Arago´n (1488e1550) and Germana de Foix (1488e1536), who founded the Monastery to house their family mausoleum. There is evidence of buildings from the Roman period and recent archaeological excavations also found evidence of an 11th century Muslim hamlet called La Rascanya, which came into the possession of the abbots of Santa Maria de la Valldigna after the Christian conquest in 1238. The abbots founded the humble Monastery of San Bernat de Rascanya in 1381 [6] and in 1546 it was chosen by the Dukes to found a Monastery under the order of St. Jerome that they could use as a family mausoleum [7]. The building was used as a Monastery until Mendizabal’s disentailment in 1835, when it became Government property. After a period of neglect the building was restored and adapted to become the new location of the Valencian Library and nowadays Valencia City Council is in charge of it. When the Monastery’s church was finished in 1645, the bodies of the Dukes were interred near the high altar. However, later on, in the 18th century, the tombs were exhumed and the hieronymite monks decided to provide a worthier burial site for them and for the two Germana de Foix sisters. The place chosen was the crypt situated just below the high altar, but there is no documentation showing the exact location of the tombs inside the crypt [8]. Therefore, the main goal of our investigation was to carry out the GPR analysis of the crypt to find out if the Dukes’ bodies were located inside their mausoleums, as well as discovering any possible elements that might exist in the subsurface such as funeral remains, before taking any action which could damage them. 2. GPR survey The GPR study of San Miguel de los Reyes Crypt was performed with a SIR-3000 control unit manufactured by GSSI. Three antennas deployed separately and operating at 1.5 GHz, 900 MHz and 400 MHz, were used in the survey. Typically, the lower the frequency, the greater the penetration depth, but a low frequency implies a decrease in resolution. Therefore, to obtain a sufficient depth of penetration the 400 MHz centre frequency antenna was deployed to prospect the crypt floor and the higher frequency antennas were used at the mausoleums to give good resolution at shallow depth. 2.1. Experimental data collection and processing The project required the collection of data from four areas, in which 80 profiles were collected (Fig. 1, Table 1). The first
studied area is located in the altar crypt (five profiles: A1eA5); the second is located in the Fernando de Arago´n mausoleum (six profiles: DC1eDC6) and the third in the Germana de Foix mausoleum (six profiles: RG1eRG6). Due to the fact that the first reconnaissance survey carried out in the mausoleums and in the altar crypt showed that it was improbable that the bodies were inside them, a fourth area was also analysed, the crypt floor. It covers the area (5 5 m) between the two mausoleums and the altar, the GPR profiles were performed as follows: longitudinal profiles (P1eP6), transversal profiles (P7eP11) and diagonal profiles (D1eD9). In addition, three profiles were carried out in the walls of the crypt (M1e M3) near the Fernando de Arago´n mausoleum. In the mausoleums and the altar the 1.5 GHz and 900 MHz antennas were moved on the surface with 20 ns time window. However, the 400 MHz centre frequency antenna was deployed to prospect the crypt floor with different time windows: 50, 60, and 80 ns. Subsequent to field acquisition, GPR data were processed using RADAN-NT software (GSSI). The raw data acquired showed directly the presence of some interesting anomalies, however, standard data processing steps were applied to obtain the best imaging. At first, data were horizontally scaled, to be able to compare the radargrams in the X direction. Due to the rapid attenuation of the high frequency GPR signal, it was also necessary to apply some type of gaining function to enhance late reflections. Filtering the GPR data limits the frequency range, suppresses unwanted reverberations, and may suppress noise. A number of band pass filters and Hilbert Transform were applied as necessary to the individual data in order to improve the anomaly identification [9,10]. The dielectric permittivity (3) of the matrix material at the crypt floor was assumed to be 13, since similar 3 values for this kind of clay filled subsurface have been reported by several authors [9e11]. The composition of the subsurface and, hence, its 3 could be confirmed thanks to drilling surveys that were carried out in the adjoining rooms. The dielectric permittivity of the mausoleums and altar were calculated through the in situ measurements of the marble tiles, leading us to estimate the 3 to be 9.5. All the time windows were converted to depth using these dielectric permittivity values in all profiles. To convert time scale to depth, the subsurface electromagnetic wave velocity was obtained from 3 [9]. The average velocity of the GPR wave front in the crypt floor was estimated to be 8 cm/ns whereas in the mausoleums and altar 10 cm/ns. 2.2. Results and discussion The survey performed in the altar and the mausoleums revealed that no bodies were located inside them. Except for the radargrams corresponding to the DC4 and RG4 profiles, all the radargrams obtained in the mausoleums presented similar features and no anomalies were detected. Fig. 2 shows the GPR sections for the DC4 profile, which presents an anomaly, and the DC3 profile, which can be considered to be representative for the areas of the mausoleums
F. Garcı´a Garcı´a et al. / Journal of Cultural Heritage 8 (2007) 87e92
89
Fig. 1. Scheme of field acquisition GPR profiles taken from the San Miguel de los Reyes Crypt.
without singularities. The clearly observable anomaly in Fig. 2(a) could be related to the existence of a singular structure, but not to a burial site because its dimensions are not big enough to correspond to a tomb. Each material type reacts to GPR radiation in a unique manner, resulting in a unique variation within the operating frequency band. The medium acts as a vertical filter during all the data collection, reducing significantly the frequency band width and the central antenna frequency peak is perturbed. Therefore, signal analysis could then be used as a complementary identification tool for object recognition. With this aim, the spectral analysis has been applied to all radargrams traces, but due to the large amount of traces that made up a radargram (350 traces/m) a stacking factor of 20 was applied to allow us to compare them to one another. In Fig. 2(c) two spectra of two set of traces are depicted; each one of them is an example of the spectral response in different cases. A very big peak of energy was obtained over
500 MHz when an anomaly is present (S1 spectrum); nevertheless, this peak does not occur when there is no anomaly (S2 spectrum). The results of the measurements performed on the altar crypt were very similar to those performed on the mausoleums. The radargrams corresponding to the profiles A3, A4 and A5 showed that this part was solid and no anomalies were detected. However, a singularity was found in the radargrams of the profiles A1 and A2. The dimensions of this singularity (0.4 0.4 m) and its location suggests that it may be an urn to protect relics. The thorough analysis, performed on the crypt subsurface, showed the presence of anomalies, probably due to the presence of the Dukes’ coffins under the crypt pavement. This anomalous area includes a surface limited by the P2eP4 lines and the P7eP9 lines. As a representative example of the anomaly the D1 profile is shown (Fig. 3). In this radargram the contact between the
F. Garcı´a Garcı´a et al. / Journal of Cultural Heritage 8 (2007) 87e92
(a)
Table 1 Profile length of the San Miguel de los Reyes Crypt
Distance (m) 0
Length (m)
Profile
Length (m)
Profile
Length (m)
Profile
Length (m)
P1 P2 P3 P4 P5 P6 P7 P8 P9 P10
4.7 4.7 4.7 4.7 4.7 4.0 4.0 4.0 4.0 4.0
P11 D1 D2 D3 D4 D5 D6 D7 D8 D9
4.0 6.7 6.7 4.6 2.7 1.2 5.4 4.0 2.5 1.1
RG1 RG2 RG3 RG4 RG5 RG6 DC1 DC2 DC3 DC4
1.4 1.2 1.2 1.4 0.8 1.2 1.4 1.2 1.2 1.4
DC5 DC6 A1 A2 A3 A4 A5 M1 M2 M3
0.8 1.2 1.7 0.5 1.2 1.2 1.2 0.8 0.8 0.8
1
2
3
4
5
6
0
0
0.5
S4
S3
20
1.0 1.5
40
2.0
Time Window (ns)
Profile
Depth (m)
90
60
2.5
(b) 5000 Spectrum of the aria without burial anomaly
(a)
Amplitude
4000
Distance (m) 0.0
0.4
0.8
0.0
1.2 0
S1
S4
Burial anomaly Spectrum
3000 2000 1000
S3
Depth (m)
0 5
10
Distance (m) 0.4
0.0
1.2 0
5 0.4 10
Time Window (ns)
Depth (m)
0.8
S2
0.0
15
0.8
8000 6000
Amplitude
600M
800M
1G
1G
Fig. 3. (a) GPR section of the D1 profile with 400 MHz centre frequency antenna acquired after processing steps with an 80 ns acquisition time. (b) Spectrum analysis of two traces sets (S2 and S3) of the D1 profile.
15
0.8
(c)
400M
Frequency (Hz)
0.4
(b)
200M
Spectrum of the anomaly Spectrum of the massive element
S1
4000
S2
2000 0 500,0M
1,0G
1,5G
2,0G
2,5G
Frequency (Hz) Fig. 2. Comparison of the GPR sections of the DC4 profile (a) and the DC3 profile (b) acquired with the 1.5 GHz centre frequency antenna after processing and recorded with a 20 ns time window. (c) Spectrum analysis of two traces sets (S1 and S2) corresponding to the DC4 and DC3 profiles.
top of what we suppose to be the coffins and the surrounding material is showed to be at 1.3 m depth (16 ns), and it is 4.5 m long. If GPR vertical sections are analysed, as the anomalies related to the coffins are present, we can observe that the coffins exhibit distinct responses moving from the left to the right side of the GPR profile. Particularly from 3.5 to 5.5 m the amplitude of the reflection is reduced. This is probably due to different size and quality of the coffins. In the spectral study of these radargrams we observed that the frequency pattern of the radargrams sectors in which the coffins were not present showed a similar trend, that was, a displacement to the low frequencies of the band width (from 100 to 500 MHz) and no outstanding amplitude peak could be defined (S3 spectrum, Fig. 3(b)). Nevertheless, the spectra behaviour when we are considering traces that correspond to the coffins is completely different. In this case, the frequency band width is also reduced (from 100 to 600 MHz); but we find a high amplitude peak around 250 MHz and three minor peaks at 275, 310 and 350 MHz (S4 spectrum, Fig. 3(b)). This absence of correlation between the spectra can be considered as a pattern which characterises the spectral response of the coffins. Three-dimensional representations of the crypt subsurface were made up, by using the RADAN-NT software, to assist the spatial correlation of radargrams and coffins responses. At first the entire dataset has been displayed at once as a solid cube. To improve the imaging of the anomalies a cut-out display of the 3D data cube was generated (Fig. 4(a)) and after
F. Garcı´a Garcı´a et al. / Journal of Cultural Heritage 8 (2007) 87e92
91
Fig. 4. Three-dimensional cut-out display of the crypt subsurface obtained from the longitudinal profiles (P1eP5) acquired with the 400 MHz centre frequency antenna.
filtering and applying different gaining functions the anomaly was isolated (Fig. 4(b)). After the analysis of the 3D cut-out display we concluded that the coffins are situated parallel to the mausoleums, but the group of coffins is horizontally offset to the section corresponding to the Germana de Foix mausoleum. The coffins were not placed at the same depth, that is, it can be identified four burial elements at different depths, from 1.3 to 1.6 m. We assume that the four burial elements correspond to the tombs of the Dukes of Calabria and those of the Germana de Foix sisters. Finally, none of the profiles carried out on the walls of the crypt (M1eM3) showed any interesting features, and indicated that these walls are solid. 3. Conclusions In our specific study, the available historical documentation led us to suppose that the remains of the Dukes were inside their mausoleums. After performing the GPR analysis, however, we were able to conclude that the mausoleums were solid. Moreover, although a very small cavity or singular zone was found there, it could not be attributed to the coffins due to its dimensions. Finally, anomalies that could be attributed to the Dukes burial were found in the crypt floor profiles.
Spectral analysis, as a complementary study, confirmed the information obtained from the visual interpretation of the radargrams. It has been useful to corroborate the presence of anomalies in a definitive way in the subsurface of the Dukes of Calabria’s crypt. As it is shown in this study spectral analysis is a potentially suitable tool to deal with object recognition. The 3D cut-out display showed itself to be an useful and effective tool in locating burial sites in the crypt subsurface that may correspond to those of the Dukes. Although in the radargrams the presence of what are probably four coffins was already observed, the 3D representation was conclusive in pinpointing their orientation and dimensions. The results obtained in this survey are a good example of GPR application as an efficient and respectful tool in Cultural Heritage restoration studies. The GPR technique helps to focus excavation work avoiding any destructive actions in mausoleums and minimizing them in other areas of the crypt.
References [1] L.B. Conyers, Innovative ground-penetrating radar methods for archaeological mapping, Archaeological Prospection 13 (2006) 139e141.
92
F. Garcı´a Garcı´a et al. / Journal of Cultural Heritage 8 (2007) 87e92
[2] G. Barone, C. Branca, S. Gresta, S. Imposa, A. Leone, D. Majolino, Geoarcheometric and geophysical methodologies applied to the study of cultural heritage: ‘‘St. Agata la Vetere’’ in Catania (Sicily, Italy), Journal of Cultural Heritage 5 (2004) 263e271. [3] M. De la Vega, A. Osella, E. Lascano, J.M. Carcione, Ground penetrating radar and geo-electrical simulations of data from the Floridablanca archaeological site, Archaeological Prospection 12 (2005) 19e30. [4] D. Ranalli, M. Scozzafava, M. Tallini, Ground penetrating radar investigations for the restoration of historic buildings: the case study of the Collemaggio Basilica (L’Aquila, Italy), Journal of Cultural Heritage 5 (2004) 91e99. [5] F. Garcı´a, I. Rodrı´guez, I. Tort, J.L. Montalva´, M. Ramı´erz, R. Capuz, R. Martı´nez, J. Benlloch, GPR technique as a tool for archaeological prospecting: exact detection of the Dukes of Calabria tombs (16th century) at San Miguel de los Reyes hieronymite monastery (Valencia, Spain), in: S. Piro (Ed.), Proceedings of the 6th International Conference on Archaeological Prospection, Rome, Italy, September 14e17 (2005), pp. 197e200.
[6] J. Teixidor, Antigu¨edades de Valencia, Observaciones crı´ticas donde con instrumentos aute´nticos se destruye lo fabuloso, dejando en su debida estabilidad lo bien fundado, Ed. Roque Chaba´s y Llore´ns, Valencia, 1895. [7] Marque´s de Cruilles, Do~na Germana de Foix, u´ltima Reina de Arago´n y su e´poca, Estudio histo´rico (manuscritos), Biblioteca Serrano Morales del Ayuntamiento de Valencia, Valencia, 1981. [8] L. Arciniega, El Monasterio de San Miguel de los Reyes, Ed. Biblioteca Valenciana, Generalitat Valenciana, Valencia, 2001. [9] V. Pe´rez, Radar del subsuelo, Evaluacio´n en arqueologı´a y en patrimonio histo´rico-artı´stico, PhD thesis, Universidad Polite´cnica de Catalu~na, Barcelona, 2001. [10] F. Garcı´a, Aplicaciones de la te´cnica geofı´sica de prospeccio´n por georadar en ingenierı´a civil y glaciologı´a, PhD thesis, Universidad Polite´cnica de Catalu~na, Barcelona, 1997. [11] J. Davis, A. Annan, Ground-penetrating radar for high resolution mapping of soil and rock stratigraphy, Geophysics Prospection 37 (1989) 531e551.
Case study
GPR technique as a tool for cultural heritage restoration: San Miguel de los Reyes Hieronymite Monastery, 16th century (Valencia, Spain) Francisco Garcı´a Garcı´a a,*, Manuel Ramı´rez Blanco c,1, Isabel Rodrı´guez Abad a,2, Rosa Martı´nez Sala b,3, Isabel Tort Ausina b,4, Javier Benlloch Marco c,1, Jose Luis Montalva´ Conesa b,4 a
Department of Cartographic Engineering, Geodesy and Photogrammetry, Polytechnic University of Valencia, Camino de Vera, s/n, 46022 Valencia, Spain b Department of Applied Physics, Polytechnic University of Valencia, Camino de Vera, s/n, 46022 Valencia, Spain c Department of Architectural Constructions, Polytechnic University of Valencia, Camino de Vera, s/n, 46022 Valencia, Spain Received 7 June 2006; accepted 11 October 2006
Abstract This paper describes GPR (ground penetrating radar) surveys performed inside the crypt of the San Miguel de los Reyes Monastery (1546e1835) in order to detect the exact location of its founders’ remains, the Dukes of Calabria (16th century). This Monastery was erected to house their family mausoleum and the bodies of the founders were buried near the high altar of the church (1645). However, in the 18th century, the tombs were exhumed to provide them with a worthier burial site: the crypt below the high altar. There is no documentation specifying the exact location of the tombs inside the crypt. Therefore, in order to reveal the exact location of the tombs the GPR survey was conducted inside the crypt. In our specific study, the available historical documentation led us to suppose that the Dukes of Calabria’s remains were inside their mausoleums. However, after having performed the GPR analysis, we discovered that the mausoleums were solid and not hollow. The project required data collection on four areas in the crypt: the altar crypt, the Fernando de Arago´n mausoleum, the Germana de Foix mausoleum and the floor between the two mausoleums and the altar. In this study, we have processed the GPR records in three different ways: the radargrams were processed in a standard manner, a detailed spectral analysis of all anomalous areas was carried out, and finally a 3D representation was generated. After this complete analysis we concluded that the bodies were not located inside their mausoleums, because they were shown to be solid. Besides, a burial site was located in the crypt subsurface near the Germana de Foix mausoleum, in which four different elements could be identified. Two of them may well be the tombs of the Dukes of Calabria and the other two the tombs of the Germana de Foix sisters. The results obtained in this survey are a good example of GPR application as an efficient and respectful tool for use in Cultural Heritage restoration studies, providing it with a very useful technique for similar projects such as those carried out in the restoration of historical buildings and those in which the elements to be examined are beneath a shallow coating of material. Ó 2007 Elsevier Masson SAS. All rights reserved. Keywords: Ground Penetrating Radar (GPR); Cultural heritage; Restoration and conservation; Hieronymite monastery; Burial element location; Spectral analysis; 3D representation
* Corresponding author. Tel.: þ34 963877557; fax: þ34 963877169. E-mail addresses: [email protected] (F. Garcı´a Garcı´a), [email protected] (M. Ramı´rez Blanco), [email protected] (I. Rodrı´guez Abad), rmsala@fis. upv.es (R. Martı´nez Sala), [email protected] (I. Tort Ausina), [email protected] (J. Benlloch Marco), [email protected] (J.L. Montalva´ Conesa). 1 Tel.: þ34 963877456; fax: þ34 963877129. 2 Tel.: þ34 963877557; fax: þ34 963877169. 3 Tel.: þ34 963877526; fax: þ34 963877975. 4 Tel.: þ34 963877121; fax: þ34 963879129. 1296-2074/$ - see front matter Ó 2007 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.culher.2006.10.005
88
F. Garcı´a Garcı´a et al. / Journal of Cultural Heritage 8 (2007) 87e92
1. Introduction Ground penetrating radar (GPR) surveys have proven to be an increasingly useful and non-destructive tool for the identification of burial structures, since with this technique we are able to identify the spatial position of structures of interest such as tombs, cavities and buried elements [1e5]. This paper describes GPR surveys performed inside the San Miguel de los Reyes Monastery (Valencia, Spain). The aim of this work was to detect the exact location of its founders’ remains, the Dukes of Calabria, Fernando de Arago´n (1488e1550) and Germana de Foix (1488e1536), who founded the Monastery to house their family mausoleum. There is evidence of buildings from the Roman period and recent archaeological excavations also found evidence of an 11th century Muslim hamlet called La Rascanya, which came into the possession of the abbots of Santa Maria de la Valldigna after the Christian conquest in 1238. The abbots founded the humble Monastery of San Bernat de Rascanya in 1381 [6] and in 1546 it was chosen by the Dukes to found a Monastery under the order of St. Jerome that they could use as a family mausoleum [7]. The building was used as a Monastery until Mendizabal’s disentailment in 1835, when it became Government property. After a period of neglect the building was restored and adapted to become the new location of the Valencian Library and nowadays Valencia City Council is in charge of it. When the Monastery’s church was finished in 1645, the bodies of the Dukes were interred near the high altar. However, later on, in the 18th century, the tombs were exhumed and the hieronymite monks decided to provide a worthier burial site for them and for the two Germana de Foix sisters. The place chosen was the crypt situated just below the high altar, but there is no documentation showing the exact location of the tombs inside the crypt [8]. Therefore, the main goal of our investigation was to carry out the GPR analysis of the crypt to find out if the Dukes’ bodies were located inside their mausoleums, as well as discovering any possible elements that might exist in the subsurface such as funeral remains, before taking any action which could damage them. 2. GPR survey The GPR study of San Miguel de los Reyes Crypt was performed with a SIR-3000 control unit manufactured by GSSI. Three antennas deployed separately and operating at 1.5 GHz, 900 MHz and 400 MHz, were used in the survey. Typically, the lower the frequency, the greater the penetration depth, but a low frequency implies a decrease in resolution. Therefore, to obtain a sufficient depth of penetration the 400 MHz centre frequency antenna was deployed to prospect the crypt floor and the higher frequency antennas were used at the mausoleums to give good resolution at shallow depth. 2.1. Experimental data collection and processing The project required the collection of data from four areas, in which 80 profiles were collected (Fig. 1, Table 1). The first
studied area is located in the altar crypt (five profiles: A1eA5); the second is located in the Fernando de Arago´n mausoleum (six profiles: DC1eDC6) and the third in the Germana de Foix mausoleum (six profiles: RG1eRG6). Due to the fact that the first reconnaissance survey carried out in the mausoleums and in the altar crypt showed that it was improbable that the bodies were inside them, a fourth area was also analysed, the crypt floor. It covers the area (5 5 m) between the two mausoleums and the altar, the GPR profiles were performed as follows: longitudinal profiles (P1eP6), transversal profiles (P7eP11) and diagonal profiles (D1eD9). In addition, three profiles were carried out in the walls of the crypt (M1e M3) near the Fernando de Arago´n mausoleum. In the mausoleums and the altar the 1.5 GHz and 900 MHz antennas were moved on the surface with 20 ns time window. However, the 400 MHz centre frequency antenna was deployed to prospect the crypt floor with different time windows: 50, 60, and 80 ns. Subsequent to field acquisition, GPR data were processed using RADAN-NT software (GSSI). The raw data acquired showed directly the presence of some interesting anomalies, however, standard data processing steps were applied to obtain the best imaging. At first, data were horizontally scaled, to be able to compare the radargrams in the X direction. Due to the rapid attenuation of the high frequency GPR signal, it was also necessary to apply some type of gaining function to enhance late reflections. Filtering the GPR data limits the frequency range, suppresses unwanted reverberations, and may suppress noise. A number of band pass filters and Hilbert Transform were applied as necessary to the individual data in order to improve the anomaly identification [9,10]. The dielectric permittivity (3) of the matrix material at the crypt floor was assumed to be 13, since similar 3 values for this kind of clay filled subsurface have been reported by several authors [9e11]. The composition of the subsurface and, hence, its 3 could be confirmed thanks to drilling surveys that were carried out in the adjoining rooms. The dielectric permittivity of the mausoleums and altar were calculated through the in situ measurements of the marble tiles, leading us to estimate the 3 to be 9.5. All the time windows were converted to depth using these dielectric permittivity values in all profiles. To convert time scale to depth, the subsurface electromagnetic wave velocity was obtained from 3 [9]. The average velocity of the GPR wave front in the crypt floor was estimated to be 8 cm/ns whereas in the mausoleums and altar 10 cm/ns. 2.2. Results and discussion The survey performed in the altar and the mausoleums revealed that no bodies were located inside them. Except for the radargrams corresponding to the DC4 and RG4 profiles, all the radargrams obtained in the mausoleums presented similar features and no anomalies were detected. Fig. 2 shows the GPR sections for the DC4 profile, which presents an anomaly, and the DC3 profile, which can be considered to be representative for the areas of the mausoleums
F. Garcı´a Garcı´a et al. / Journal of Cultural Heritage 8 (2007) 87e92
89
Fig. 1. Scheme of field acquisition GPR profiles taken from the San Miguel de los Reyes Crypt.
without singularities. The clearly observable anomaly in Fig. 2(a) could be related to the existence of a singular structure, but not to a burial site because its dimensions are not big enough to correspond to a tomb. Each material type reacts to GPR radiation in a unique manner, resulting in a unique variation within the operating frequency band. The medium acts as a vertical filter during all the data collection, reducing significantly the frequency band width and the central antenna frequency peak is perturbed. Therefore, signal analysis could then be used as a complementary identification tool for object recognition. With this aim, the spectral analysis has been applied to all radargrams traces, but due to the large amount of traces that made up a radargram (350 traces/m) a stacking factor of 20 was applied to allow us to compare them to one another. In Fig. 2(c) two spectra of two set of traces are depicted; each one of them is an example of the spectral response in different cases. A very big peak of energy was obtained over
500 MHz when an anomaly is present (S1 spectrum); nevertheless, this peak does not occur when there is no anomaly (S2 spectrum). The results of the measurements performed on the altar crypt were very similar to those performed on the mausoleums. The radargrams corresponding to the profiles A3, A4 and A5 showed that this part was solid and no anomalies were detected. However, a singularity was found in the radargrams of the profiles A1 and A2. The dimensions of this singularity (0.4 0.4 m) and its location suggests that it may be an urn to protect relics. The thorough analysis, performed on the crypt subsurface, showed the presence of anomalies, probably due to the presence of the Dukes’ coffins under the crypt pavement. This anomalous area includes a surface limited by the P2eP4 lines and the P7eP9 lines. As a representative example of the anomaly the D1 profile is shown (Fig. 3). In this radargram the contact between the
F. Garcı´a Garcı´a et al. / Journal of Cultural Heritage 8 (2007) 87e92
(a)
Table 1 Profile length of the San Miguel de los Reyes Crypt
Distance (m) 0
Length (m)
Profile
Length (m)
Profile
Length (m)
Profile
Length (m)
P1 P2 P3 P4 P5 P6 P7 P8 P9 P10
4.7 4.7 4.7 4.7 4.7 4.0 4.0 4.0 4.0 4.0
P11 D1 D2 D3 D4 D5 D6 D7 D8 D9
4.0 6.7 6.7 4.6 2.7 1.2 5.4 4.0 2.5 1.1
RG1 RG2 RG3 RG4 RG5 RG6 DC1 DC2 DC3 DC4
1.4 1.2 1.2 1.4 0.8 1.2 1.4 1.2 1.2 1.4
DC5 DC6 A1 A2 A3 A4 A5 M1 M2 M3
0.8 1.2 1.7 0.5 1.2 1.2 1.2 0.8 0.8 0.8
1
2
3
4
5
6
0
0
0.5
S4
S3
20
1.0 1.5
40
2.0
Time Window (ns)
Profile
Depth (m)
90
60
2.5
(b) 5000 Spectrum of the aria without burial anomaly
(a)
Amplitude
4000
Distance (m) 0.0
0.4
0.8
0.0
1.2 0
S1
S4
Burial anomaly Spectrum
3000 2000 1000
S3
Depth (m)
0 5
10
Distance (m) 0.4
0.0
1.2 0
5 0.4 10
Time Window (ns)
Depth (m)
0.8
S2
0.0
15
0.8
8000 6000
Amplitude
600M
800M
1G
1G
Fig. 3. (a) GPR section of the D1 profile with 400 MHz centre frequency antenna acquired after processing steps with an 80 ns acquisition time. (b) Spectrum analysis of two traces sets (S2 and S3) of the D1 profile.
15
0.8
(c)
400M
Frequency (Hz)
0.4
(b)
200M
Spectrum of the anomaly Spectrum of the massive element
S1
4000
S2
2000 0 500,0M
1,0G
1,5G
2,0G
2,5G
Frequency (Hz) Fig. 2. Comparison of the GPR sections of the DC4 profile (a) and the DC3 profile (b) acquired with the 1.5 GHz centre frequency antenna after processing and recorded with a 20 ns time window. (c) Spectrum analysis of two traces sets (S1 and S2) corresponding to the DC4 and DC3 profiles.
top of what we suppose to be the coffins and the surrounding material is showed to be at 1.3 m depth (16 ns), and it is 4.5 m long. If GPR vertical sections are analysed, as the anomalies related to the coffins are present, we can observe that the coffins exhibit distinct responses moving from the left to the right side of the GPR profile. Particularly from 3.5 to 5.5 m the amplitude of the reflection is reduced. This is probably due to different size and quality of the coffins. In the spectral study of these radargrams we observed that the frequency pattern of the radargrams sectors in which the coffins were not present showed a similar trend, that was, a displacement to the low frequencies of the band width (from 100 to 500 MHz) and no outstanding amplitude peak could be defined (S3 spectrum, Fig. 3(b)). Nevertheless, the spectra behaviour when we are considering traces that correspond to the coffins is completely different. In this case, the frequency band width is also reduced (from 100 to 600 MHz); but we find a high amplitude peak around 250 MHz and three minor peaks at 275, 310 and 350 MHz (S4 spectrum, Fig. 3(b)). This absence of correlation between the spectra can be considered as a pattern which characterises the spectral response of the coffins. Three-dimensional representations of the crypt subsurface were made up, by using the RADAN-NT software, to assist the spatial correlation of radargrams and coffins responses. At first the entire dataset has been displayed at once as a solid cube. To improve the imaging of the anomalies a cut-out display of the 3D data cube was generated (Fig. 4(a)) and after
F. Garcı´a Garcı´a et al. / Journal of Cultural Heritage 8 (2007) 87e92
91
Fig. 4. Three-dimensional cut-out display of the crypt subsurface obtained from the longitudinal profiles (P1eP5) acquired with the 400 MHz centre frequency antenna.
filtering and applying different gaining functions the anomaly was isolated (Fig. 4(b)). After the analysis of the 3D cut-out display we concluded that the coffins are situated parallel to the mausoleums, but the group of coffins is horizontally offset to the section corresponding to the Germana de Foix mausoleum. The coffins were not placed at the same depth, that is, it can be identified four burial elements at different depths, from 1.3 to 1.6 m. We assume that the four burial elements correspond to the tombs of the Dukes of Calabria and those of the Germana de Foix sisters. Finally, none of the profiles carried out on the walls of the crypt (M1eM3) showed any interesting features, and indicated that these walls are solid. 3. Conclusions In our specific study, the available historical documentation led us to suppose that the remains of the Dukes were inside their mausoleums. After performing the GPR analysis, however, we were able to conclude that the mausoleums were solid. Moreover, although a very small cavity or singular zone was found there, it could not be attributed to the coffins due to its dimensions. Finally, anomalies that could be attributed to the Dukes burial were found in the crypt floor profiles.
Spectral analysis, as a complementary study, confirmed the information obtained from the visual interpretation of the radargrams. It has been useful to corroborate the presence of anomalies in a definitive way in the subsurface of the Dukes of Calabria’s crypt. As it is shown in this study spectral analysis is a potentially suitable tool to deal with object recognition. The 3D cut-out display showed itself to be an useful and effective tool in locating burial sites in the crypt subsurface that may correspond to those of the Dukes. Although in the radargrams the presence of what are probably four coffins was already observed, the 3D representation was conclusive in pinpointing their orientation and dimensions. The results obtained in this survey are a good example of GPR application as an efficient and respectful tool in Cultural Heritage restoration studies. The GPR technique helps to focus excavation work avoiding any destructive actions in mausoleums and minimizing them in other areas of the crypt.
References [1] L.B. Conyers, Innovative ground-penetrating radar methods for archaeological mapping, Archaeological Prospection 13 (2006) 139e141.
92
F. Garcı´a Garcı´a et al. / Journal of Cultural Heritage 8 (2007) 87e92
[2] G. Barone, C. Branca, S. Gresta, S. Imposa, A. Leone, D. Majolino, Geoarcheometric and geophysical methodologies applied to the study of cultural heritage: ‘‘St. Agata la Vetere’’ in Catania (Sicily, Italy), Journal of Cultural Heritage 5 (2004) 263e271. [3] M. De la Vega, A. Osella, E. Lascano, J.M. Carcione, Ground penetrating radar and geo-electrical simulations of data from the Floridablanca archaeological site, Archaeological Prospection 12 (2005) 19e30. [4] D. Ranalli, M. Scozzafava, M. Tallini, Ground penetrating radar investigations for the restoration of historic buildings: the case study of the Collemaggio Basilica (L’Aquila, Italy), Journal of Cultural Heritage 5 (2004) 91e99. [5] F. Garcı´a, I. Rodrı´guez, I. Tort, J.L. Montalva´, M. Ramı´erz, R. Capuz, R. Martı´nez, J. Benlloch, GPR technique as a tool for archaeological prospecting: exact detection of the Dukes of Calabria tombs (16th century) at San Miguel de los Reyes hieronymite monastery (Valencia, Spain), in: S. Piro (Ed.), Proceedings of the 6th International Conference on Archaeological Prospection, Rome, Italy, September 14e17 (2005), pp. 197e200.
[6] J. Teixidor, Antigu¨edades de Valencia, Observaciones crı´ticas donde con instrumentos aute´nticos se destruye lo fabuloso, dejando en su debida estabilidad lo bien fundado, Ed. Roque Chaba´s y Llore´ns, Valencia, 1895. [7] Marque´s de Cruilles, Do~na Germana de Foix, u´ltima Reina de Arago´n y su e´poca, Estudio histo´rico (manuscritos), Biblioteca Serrano Morales del Ayuntamiento de Valencia, Valencia, 1981. [8] L. Arciniega, El Monasterio de San Miguel de los Reyes, Ed. Biblioteca Valenciana, Generalitat Valenciana, Valencia, 2001. [9] V. Pe´rez, Radar del subsuelo, Evaluacio´n en arqueologı´a y en patrimonio histo´rico-artı´stico, PhD thesis, Universidad Polite´cnica de Catalu~na, Barcelona, 2001. [10] F. Garcı´a, Aplicaciones de la te´cnica geofı´sica de prospeccio´n por georadar en ingenierı´a civil y glaciologı´a, PhD thesis, Universidad Polite´cnica de Catalu~na, Barcelona, 1997. [11] J. Davis, A. Annan, Ground-penetrating radar for high resolution mapping of soil and rock stratigraphy, Geophysics Prospection 37 (1989) 531e551.