Fig 1.
A: Geographical location of El Castillejo (map generated by the authors using QGis 3.26; Basemap: MDS05 2020 CC-BY 4.0 ign.es; seismic faults from the QAFI database [17]); B: Drone view of the archaeological site from SW. The village of Guájar Faragüit lies in the background. Photograph by P. Forlin.
Fig 2.
Examples of seismic damage and restoration observed at El Castillejo: A-B: shear cracks in Building 4; C: horizontally-shifted block of rammed earth in Building 3; D: tilted block of the outer wall; E: broken and displaced wall in Building 30; F: post-earthquake restoration of the western gatehouse and G: Building 10. Photographs by P. Forlin.
Fig 3.
The distribution of structural deformation and damage grouped by buildings (see key for the types of damage).
The dot red lines indicate previously excavated areas. Trenches discussed here are shown as red circles (map generated by P. Forlin using QGis 3.26).
Fig 4.
A: Trench 1 under excavation. Note the burning layers 103–108 sealed by debris and rubble 106 and 120; B: Close-up of the in situ carbonised beams; C: Stratigraphic section S; D: Stratigraphic section E.
Fig 5.
A: Trench 2 under excavation looking NE. Note wall 206 showing penetrative fractures associated with fallen blocks of rammed earth; B: Close up of debris and fallen blocks 202 and 203; C: the occupation surface 208 looking north; D: the occupation surface 210 looking north; E: Stratigraphic section S; F: Stratigraphic section W.
Fig 6.
A: Trench 3 looking N; B: Stratigraphic section. 304 corresponds with the rammed earth block and loose horizon.
Fig 7.
Plot of the radiocarbon dates obtained from the analysed samples (image by G. Capuzzo).
Table 1.
Radiocarbon dates and Bayesian statistics.
Data provided by the SUERC radiocarbon laboratory of the University of Glasgow.
Table 2.
The uncertainty associated with the luminescence age (test year, 2019) is given at the 68% level of confidence, calculated following a procedure of error propagation [41]. The luminescence age corresponds to the quotient of the paleodose, P, and the dose rate, Dr; those shown were rounded to the nearest 5 years. The two ages calculated for sample 438–2 take account of different histories of use of the brick following manufacture (A and B), as discussed in the main text. The overdispersion in P, σB, is an indicator of beta dose rate heterogeneity in the case of heated samples, and the number of determinations (aliquots), n, is given in the adjacent column. In calculating the annual dose rate, Dr, the beta dose rate Dβ includes a small contribution (0.035 mGy/a) from radiation emitted by lithogenic radionuclides within the quartz grains. The average moisture content of ceramic and environmental materials was assumed to be 3±1% for sample 348–2 and 10±2% for sample 348–5, by weight.
Fig 8.
A: Seismicity in Andalusia, the red area shows the archaeological site at El Castillejo, the red lines are possible active faults in the vicinity (generated by K. Reicherter using QGis 3.26; Basemap: MDT200 2015 CC-BY 4.0 ign.es; historical earthquakes from [23], faults from the QAFI database [17]) B: recent earthquakes around El Castellejo (star) (K. Reicherter’s own compilation; Basemap Open Street Map); C: Palaeo-shake map modelled for the missing El Castellejo earthquake, based on M6 and 10 km hypocentral depth [35] (generated using the ShakeMap map code by Jens Skapski; Basemap Open Street Map [http://usgs.github.io/shakemap/index.html]).