Passchier et al. (2021) identified two intervals in the JW01 aqueduct that may record earthquake damage. The first occurs in layer L48 of sequence JJT2, where carbonate thin sections show deformation twins, broken calcite crystals, and fractures filled with micrite, indicating a sudden mechanical shock (e.g. a "falling stone") during active use of the aqueduct. These features are overlain by undeformed micrite (interpreted as summertime deposition) and were immediately followed by the application of a new plaster layer (P3). A second possible earthquake event occurs in layer L50 of sequence JJT3, where a "wide sparite-dominated layer" and fractured carbonate at site JWP128 may suggest a later phase of structural damage preceding abandonment.

These observations come from microfabric analysis of polished slabs and thin sections sampled at two locations (JWP128 and JWP114). Annual layering was established from rhythmic micrite–sparite alternation and stable isotope cycles, producing a 59-year sequence. Radiocarbon dates from plaster layers P1–P3 place the use of the aqueduct between the 1st century BCE/1st century CE and the early 3rd century CE. Within this sequence, the damage recorded in L48 ocurred after roughly 48 years of operation, while the later disturbance in layer L50 occurred near the end of the aqueducts’s lifespan. AMS analysis of ¹⁴C in charcoal from plaster layer P3 yielded a calibrated 1σ date for the L48 event between 22 and 311 CE. Counting approximately 48 years of deposition between plaster P2 (dated 4–130 CE) and layer L48 could conceivably narrow that window to 52–178 CE. The 2nd event (layer L50 in sequence JJT3) appears to have struck soon after (a few years to a decade) the first event (L48).