Allison (2013:131-132) interpreted a leaning city sea wall at Early Islamic Ayla as earthquake damage most likely caused by the 1033 CE earthquake. In this reading, the original city wall was tilted during the earthquake, but did not collapse completely, leaving a damaged wall that required stabilization. Allison (2013) speculated that the wall may have tilted due to liquefaction. The revetment wall was therefore interpreted as a later repair, built against the leaning city wall in order to support it after the earthquake. Its position beside the tilted wall suggests that it was not part of the original city-wall construction, but a response to damage already visible in the standing architecture (Allison, 2013:131-132).

Allison favored attribution of the leaning city wall and its associated revetment repair to the 1033 CE earthquake because the archaeological dating evidence places the repair after the mid-8th century earthquakes but before the final destruction of Ayla in the earthquake and probable tsunami in 1068 CE. Pottery recovered from inside the revetment wall was interpreted as consistent with a repair in the mid-10th century CE or later (Allison, 2013:124-131). Radiocarbon evidence also informed this interpretation. A charcoal fragment recovered within the revetment wall yielded a calibrated date of 777-981 CE (2σ). Allison therefore regarded the 1033 CE earthquake as the most plausible historical candidate, particularly because it falls within the Fatimid period, a time when Ayla was still inhabited and functional enough to justify a major structural repair (Allison, 2013:131-132). This contrasts with the historical earthquake of 1068 CE, which historical accounts describe as catastrophic for Ayla, with the city effectively destroyed and nearly all inhabitants killed. In such a scenario, there would have been little practical capacity or incentive to undertake a substantial engineering repair such as the construction of the revetment wall afterward. Allison also notes that Ayla had already suffered political and military disruption during the Fatimid period, including the sack of the city in 1024 CE, but the settlement evidently remained occupied and economically viable afterward, making a post-1033, pre-1068 repair scenario archaeologically plausible (Allison, 2013:74-76, 131-132). al-Tarazi and Khorjenkov (2007) attributed the tilted wall evidence to the 1068 CE earthquake(s). the

Fig. 3.9 - Photograph of the visible northwestern edge of the leaning city wall and revetment wall as exposed in the southeastern balk of Trench AY1. Pottery was collected from beneath the revetment wall via the small probe visible on the right side of the photo. View toward the southeast. - click on image to open in a new tab - Allison (2013)