Aerial photograph of Caesarea and its harbour. The dark patches at sea are submerged rocks. Note that most of the rocky outcrops, other than the submerged relics of the Herodian breakwaters, are truncated from the west along a straight, N-S trending linement.

(Photo: Courtesy Ofek Ltd)

Ancient harbors in the eastern Mediterranean Sea are excellent sites for measuring historic coastal earthquakes and sea level variations, as port installations can be dated very accurately by archaeological methods. Relics of breakwaters, docks, and quays provide evidence of the contemporary sea level and the tidal range is less than 50 cm. Caesarea is a remarkable site for the measurement of historic earthquake displacements and sea level changes, due to the intensive archaeological research which has been carried out for many years by Avner Raban, in collaboration with Yossi Patrich from CMS and Yosef Porat from the Israel Antiquities Authority.

The port of Caesarea was a large, open sea harbor that was built by King Herod the Great (37-4 BCE) between 22 and 10 BCE. It was active during the Roman, Byzantine, Early Moslem, and Crusader periods, until the city was razed in 1265 by Sultan Baibars I (1223-1277), during the final throes of the Crusader Kingdom. Archaeological markers of ancient sea levels, such as harbor installations and loading platforms, abound in Caesarea. Some indicate a similarity between the present sea level and that of Roman times, correct to ±30 cm, whereas others reveal a marked difference. These apparently conflicting data have led to controversial interpretations of the recent geological history of the Caesarea region. The new data that we have obtained, serves to unite the opposing interpretations into a comprehensive geological model.

As far as the coastal geology is concerned, the archaeological findings at Caesarea present conflicting indications of the existing structural patterns between 2,000 and 750 years ago. Some evidence suggests that geological stability prevailed in Caesarea from Herodian times to the present, while other discoveries indicate geological activity. Avner Raban and Alex Flinder, have shown that the large piscina, the rectangular pond quarried into the hard sandstone some 500 m south of the southern breakwater could, today, be brought to an operational state, as its water circulation system correlates to the present sea level. Similarly, Raban has shown that sluice channels, which were cut into the sandstone at the base of the southern breakwater to generate currents and flush the harbor, are also positioned at the current sea level. It should be emphasized that the findings suggesting structural and oceanographic stability, are invariably located on land, or at very shallow marine depths.

On the other hand, additional archaeological evidence suggests that subsidence of 5-8 m has taken place in Caesarea since Roman times. Avner Raban in collaboration with David Neev and colleagues from the Geological Survey, has revealed segments of the breakwaters which were built above sea level, but are currently submerged in water depths of 6-10 m. A loading platform was discovered at a depth of approximately 6 m on the southern breakwater, northwest of the citadel. There is no doubt that the platform was originally built slightly above sea level. Unlike the citadel and the near-shore segments of the breakwater, which were built on hard sandstone and remained at their original elevation, the submerged landing platform was built on unconsolidated sand. The present location of the platform suggests at least 6 m of subsidence. Additional evidence of subsidence was reported by Avner Raban from a site near the entrance to the harbor, where an intact segment of the original breakwater was preserved, from its foundation to its paved top. There, the breakwater is approximately 3 m high. Presumably about 1 m of the structure extended above sea level. At present the paved top of the breakwater is at a depth of 4 m, suggesting a subsidence of approximately 5 m.

The Coastal Faults of Caesarea

The transition zone from land to sea in Caesarea is readily traceable in aerial photographs. It can be seen as a straight line bounding submerged outcrops of sandstone. From the distribution of the outcrops and other geological data, David Neev and his colleagues inferred that active geological faults occur off the shores of Caesarea and that tectonic offsets along these faults could possibly trigger earthquakes (earthquakes are the result of intermittent displacement along a geological fault). The damage from structural tremors is twofold: the damage of the actual motion, which would severely effect any structure transecting the active segment of a fault; and vibrations produced by the seismic waves generated by the motion, which travel the terrain adjacent to the focus of the earthquake and violently shake and shatter artificial structures.

Earthquakes are a common phenomenon in Israel. Most occur in the region of the Dead Sea Rift, but, as shown in the catalog of historic tremors of the Levant, prepared by David Amiran et al, of the Hebrew University, Jerusalem, Mediterranean coastal cities have been repeatedly damaged by earthquakes. It seems that these earthquakes derived from displacements along local active faults in the coastal plain and the continental margin. Along the coastal plain of Israel, land-based evidence of faulting from recent geological periods has been considered unreliable, due to its similarity to widespread relics from a millennia of human activity. Ancient quarries, excavations and drainage works appear in places, remarkably similar to faulted escarpments. This structural ambiguity does not interfere with the exploration in the shallow continental shelf, where the geological regime of the coastal plain still prevails and ancient artificial interference is very unlikely.

In order to reconcile both the equivocal geological data and the conflicting archaeological data regarding the active boundary between the geologically stable coast and the subsiding continental shelf, we conducted a detailed geophysical survey on the shallow continental shelf off Caesarea and made a series of submarine geological observations. The survey encountered two distinct types of submarine terrains, one comprising unconsolidated sediments, mostly sand, and the other - rocky outcrops of sandstone. The rocky terrain forms a series of low ridges, and the sediment-filled areas form troughs between the ridges. Both the troughs and the ridges are aligned in belts which run nearly parallel to the shore. One of these rocky belts builds the coast of Caesarea, where the rocks are partly exposed and partly submerged. Other rocky ridges are found farther out to sea and also on land. They represent the changing position of the seashore during the last two million years. The variable coastlines are the derivative of sea level oscillations due to climatic changes. The present morphology of the Caesarea coast is shaped primarily by its rocky foundation, however, ancient quarrying and construction have also been contributing factors.

The survey repeatedly traversed the submerged western edge of the coastal sandstone ridge, looking for evidence for structural offsets at the transition zone between the rocky ridge and the sediment-filled trough. At the western edge of the coastal sandstone ridge, we encountered a system of faults, trending approximately N-S, nearly parallel to the general orientation of the present coast, downthrowing the western flank of the coastal ridge. The faults seem to offset the top of the sandstone by 1-3 m, truncating the entire coastal sandstone ridge of Caesarea.

Our submarine observations of the sandstone escarpments complement previous observations reported by Raban, Neev and their colleagues, which support the interpretation that the coastal sandstone ridge is transected by a series of faults. As the geophysical surveys uncovered the faults that transect the Herodian breakwaters and the traces of these faults are present at the boundary between the stable and submerged sections of the breakwaters, it seems that these faults were active after the construction of the harbor of Caesarea, thus downthrowing the western block. The vertical offset of the faults at the harbor was intensified by the heavy load of the breakwaters, but the initial motion was apparently triggered by earthquakes.

The geophysical survey off Caesarea encountered a series of coast-parallel faults further offshore. A N-S trending fault was encountered in the trough, nearly 500 m west of the sandstone ridge and west of the breakwaters. This fault shows displacements of 4-6 m, but since it does not transect artificial structures, its dating is ambiguous. Other faults were found to transect the distal sandstone ridge. They are associated with escarpments at the seafloor and also offset sandstone layers. The sandstone was dated to not older than 2,000,000 years. Logically, the faults offsetting it must be younger. We do not have more specific age constraints for these faults as yet. We also attempted to find evidence for young geological faults on land, but the indications encountered were not beyond the doubt of an artifact, and were considered inconspicuous. Thus the fault that offsets the breakwaters of Caesarea is not a unique feature. There is ample evidence for tectonic activity off the shores of Caesarea in the Quaternary period (the last 2,000,000 years). The authors of the ancient texts describing enormous damages caused in Caesarea by tremors may have exaggerated, but in all likelihood founded their reports on real events.

This study was made possible through the diligent efforts of Ilana Perecman, a student at the Department of Maritime Civilizations and the competent assistance of the marine technicians at the National Institute of Oceanography in Haifa. The generous financial support of The Caesarea - Edmond Benjamin de Rothschild Foundation, is gratefully acknowledged.

Yossi Mart

(a) A seismic reflection profile of the transtionzone between the coastal sandstone ridge and the sand-filled, submerged trough. The profile is produced by transmitting and recieving a series of short bursts of sound. The sound is reflected from the seafloor and buried strata, to image the submarine stratification. (b) Interpretation of the profile shows that a fault has displaced the westren flank of the sandstone ridge downwards.


Close Window