A series of investigations over the past decades have identified and analyzed a roughly 2000-year-old laterally extensive tsunami deposit preserved, at a minimum, in the shallow (~10-20 m depth) offshore shelf adjacent to the ancient harbor of Caesarea. Underwater archaeological excavations targeting this offshore deposit were first undertaken in Area W by Reinhardt et al. (2006) and later revisited by Goodman-Tchernov and Austin (2015). Sediment cores collected at various locations on the offshore shelf by Goodman-Tchernov et al. (2009) recovered the same deposit, which they designated Event 2. Building on this work, Goodman-Tchernov and Austin (2015) conducted a 3D seismic CHIRP survey that identified a subsurface reflector, Reflector B, corresponding stratigraphically to Event 2. A structure map of this reflector enabled reconstruction of the harbor’s configuration at the time of the tsunami, suggesting that backwash flow was funneled through one or two concentrated channels at the harbor mouth. This hydraulic pattern implies that the harbor was more or less intact when the tsunami hit.

Reinhardt et al. (2006:1061) dated the tsunami deposit in Area W to between the 1st century BCE and the 2nd century CE by combining radiocarbon analysis of articulated Glycymeris shells with optically stimulated luminescence (OSL) dating. Researchers were able to further constrain this time window using historical sources and sedimentological studies from the inner harbor. Construction of Caesarea’s ancient harbor was initiated by King Herod the Great around 21 BCE and was completed by around 10 BCE (Reinhardt and Raban, 1999:811). Citing Holum et al. (1988), Reinhardt and Raban (1999:811) claim that the harbor “was described by Josephus Flavius (a Jewish historian) in A.D. 72–75.” While Josephus did not write anything specific about the state of the harbor during that period, he did describe the harbor’s appearance and configuration immediately after it's construction. He did this using sources as inital construction was completed almost 50 years before Josephus was born but, considering his familiarity with Caesarea, he may have also constructed those descriptions based on observation. Thus, it seems likely that if the harbor had deteriorated significantly from its original condition, he would have noted it. This then provides a fairly reliable terminus post quem of the early 70s CE. For the terminus ante quem, Goodman-Tchernov and Austin (2015:445) report that “the overall state of the harbor had significantly deteriorated by the end of the 2nd century CE, and probably even earlier, according to radiocarbon-dated sedimentological evidence showing a shift from a low-energy, harbor environment to an open-water exposed, unprotected environment during that period (Reinhardt and Raban, 1999; Reinhardt et al., 1994).” Thus, the laterally extensive Event 2 tsunami — which would have caused significant damage to the harbor — probably struck between the early 70s CE and about 200 CE. The tsunami deposit is apparently also present within Caesarea’s inner harbor. Reinhardt and Raban (1999:812) described a “high-energy water deposit” (L4) inside the harbor containing “rounded and encrusted cobbles,” ship ballast, and a “shell-rich matrix,” all indicators of intense wave activity. Goodman-Tchernov et al. (2009) interpreted the inner-harbor L4 deposit as a proximal facies of the same tsunami deposit documented on the adjacent shelf.

Although some researchers have suggested that the 1st or 2nd century CE harbor deterioration was largely due to the tsunami, it is not yet resolved whether the tsunami was primarily or wholly responsible for the harbor’s demise, or whether it was only one of several contributing factors. Reinhardt and Raban (1999:812), for example, initially attributed the L4 “high-energy water deposit” to harbor degradation without invoking a tsunami as the cause. They also reported that during the 1993 excavation season, a “late first century A.D. shipwreck” was discovered resting atop a submerged breakwater. The ship was dated by its associated lead ingots to “A.D. 83–96,” and its presence on a submerged structure indicated to them that “this portion of the structure must have been submerged to allow a ship to run-up and founder on top” (Raban, 1999: Fig. 3B). Such subsidence of the breakwater could have resulted from a variety of processes. Another diagnostic harbor structure was also reported on by Reinhardt and Raban (1999:812). They observed that in “many instances,” the large caissons that comprise the harbor moles can be observed to tilt by 15°–20°. Such tilting, they suggest, could result from current scour during large-scale storms or tsunamis, liquefaction triggered by earthquakes, or differential settlement — particularly in cases where caissons were founded on unconsolidated sand rather than firm kurkar ridges. Further potential evidence for earthquake and/or tsunami activity was uncovered by Fritsch and Ben-Dor (1961). In their excavations, they encountered a “large wooden beam” beneath which a “second-century Roman amphora” was found, where the the beam and masonry were seemingly "catapulted into the sea at the same time.” They suggested that this destruction may have had a seismic origin. Other finds from the same layer included “numerous sherds of second-century amphorae, corroded bronze coins, ivory hairpins, and colorful bits of glass.”

Based on their structure map for Reflector B, Goodman-Tchernov and Austin (2015:451) argue that “the incoming wave must have encountered an intact and standing outer harbor mole, which would have forced abrupt shoaling of the incoming wave, scouring deeply the area immediately outside the harbor, while also breaching the tops of man-made features.” They add that “during subsequent retreat of the wave, that outflowing water would have concentrated through the harbor mouth, between the [perpendicular to the shoreline] reinforced moles, preferentially scouring and eroding the region immediately outside the harbor entrance and depositing larger deposits farther offshore, as is evident in the ~80 cm 2nd century deposit in Area W (Reinhardt et al., 2006).” In fact, Goodman-Tchernov et al. (2009:945) noted the presence of a “pronounced disconformity” in Area W where “at least 2 ka worth of deposits” below the Event 2 tsunami layer were “missing (Reinhardt et al., 2006).” This, they suggested, could have been the “result of erosion and scouring" during "the Roman period tsunami (Event 2).”

Despite all the offshore evidence, Galli et al. (2021) reports that "no tsunami deposits that can be attributed to the second century CE were reported from land excavations in Caesarea and around."

Although some scholars have proposed a local neotectonic explanation for the submergence of the harbor structures and the harbor’s deterioration by the end of the 2nd century CE (if not earlier), Dey et al. (2014) observed that the coastline remained stable for roughly 2000 years, with sea-level fluctuations not exceeding ±50 cm and no evidence of vertical displacement in the harbor structures. Galli et al. (2021: 8–12) likewise concluded that both the coastline and sea level have remained stable since “at least the Hellenistic period.”

Seismic and Core cross sections showing interpreted tsunamigenic strata - To the left are individual cores from Caesarea and to the right are composite cores from Caesarea and Jisr el Zarka (~1.5-4.5 km. from Caesarea. Click on either image to open in a new tab.