The Tsunami Caused by the Prehistoric Eruption of Thera

Thera in the Mediterranean erupted violently about 3500 years B.P. and resulted in a depression in the sea. A huge amount of pumice would be ejected by this eruption. Some pumice is found on the island itself and at the bottom of the surrounding sea, and some pumice would be transported to the other places by the tsunami caused by the eruption. In order to surmise the violence of the eruption, one must consider the tsunami.

The source area of the tsunami generated by the Thera eruption must remain unkown, since one can not clarify the processes and the magnitude of the eruption. The original topography of Thera Island before the eruption is not known and, therefore, it is not clear whether the tsunami emerged through the north-western and the south-western openings of the assumed inner bay or from all the peripheral sea coasts. In this paper the assumption that the source area was an enveloping circle of the present Thera Islands and that the original height of the tsunami was uniform along the circle will be adopted in drawing the refraction diagrams of the tsunami. For this assumption, the author has to show a realistic model: a tremendous amount of the ejecta and lithic fragments from the central crater fell back or rushed down along the slopes of the volcano in all directions and reached the shores, causing progressive water barricades.

In this model, the initial motions of the tsunami should be upwards in all directions around the source: this result is analogous to the observations along the Sunda Straits in the 1883 eruption of Krakatoa.

To estimate the original height of the tsunami at the source on Thera, one may consider the following two facts.

One is the evidence on the island of Anaphi: Marinos & Melidonis (1971) found the beds of pumice in three localities on the island, one of them on the western coast and two on the north-eastern coast. The most appropriate of the layers for interpretation of the tsunami is on the western corner (that facing Thera) 350 meters distant from the coast. There, according to Marinos & Melidonis, at an altitude of about 40 - 50 meters, in two surface cavities, a layer of white pumice of Thera type is covered by soil and alluvial stones. The material of this layer consists of rough grained pumice with some bigger rounded pieces of pumice about 10 - 15 cm in diameter. The extent of this layer is limited and its thickness is about 1 meter. A bathymetric sketch map between Thera and Crete including Anaphi is shown in Fig. 1 on reference to Chart No. 1091.

The other evidence is found at the north of Jaffa - Tel Aviv, being about 1000 km distant from Thera, in the eastern Mediterranean. According to Pfannenstiel (1960), layers of sea-borne pumice are found in sediments of a postglacial terrace 5 meters above the sea-level, north of Jaffa - Tel Aviv and these layers were transported by the tsunami of the Minoan eruption of Thera. Calculating eustatic rise of the sea-level by about 2 meters since the Minoan age, the height of the tsunami at Tel Aviv should have been about 7 meters. A bathymetric sketch map between Thera and Tel Aviv is shown in Fig. 2 on reference to Chart No. 183.

Velocities of tsunamis depend on sea depths, and may be expressed by √gH, where g and H denote the gravity acceleration and the depth of the sea respectively. The refraction diagrams starting from Thera for Crete and Tel Aviv are drawn as shown in Figs. 3 and 4 respectively.

Heights of tsunami are governed by the Green's law which states

H₀ = H₁ ( b₁/b₀ )^½  ( h₁/h₀ )^¼

where H, b, and h denote height of tsunami, width between two adjacent trajectories, and depth of the sea respectively, and suffix O denotes the origin area and suffix 1 the area concerned. In the present case, it is assumed that h₀ is nearly equal to h₁, both being near shores. In addition to the loss of amplitude due to the wave's spreading out, there is also a certain loss from absorption due to frictional dissipation of the energy of tsunamis. In the following discussion, the latter will be ignored because of its small magnitude.

Applying Green's law to the two pieces of evidence, one may estimate the probable original height of the tsunami at Thera:

Island of Anaphi

The original height of the tsunami at Thera might be preserved unchanged at the western coast of Anaphi, i.e. 50 meters, according to Green's law as shown in Fig. 3. The tsunami arrived at the western shore of Anaphi from Thera in 10 min.

 Jaffa - Tel Aviv

The observed height of the tsunami at Jaffa - Tel Aviv (7 meters) may be related to that of the source at Thera by that of the Kasos Straits east of Crete.

The width between the adjacent trajectories at Thera should have expanded about 8 times at the Kasos Straits D₁, D₂, and further expanded about 20 times from the strait to Jaffa - Tel Aviv J₁ J₂ as shown in Fig. 4. In total the width should have expanded about 160 times between Thera and Jaffa - Tel Aviv. The tsunami covered the distance in about 105 min. By Green's law, one may estimate the original height of the tsunami at Thera at 89 meters.

The two estimations of the original height of the tsunami, 50 meters obtained from the evidence at Anaphi and 89 meters from the evidence at Jaffa - Tel Aviv, are averaged by weighting twice for the former because the former estimation may be more reliable: The most probable height of the tsunami at the source area becomes 63 meters.

Marinatos (1939) found a thick layer of Thera pumice at Amnisos on the northern coast of Crete, transported by the tsunami. The height of the tsunami there will be determined by Green's law using the above original height: as shown in Fig. 3, the width between the adjacent trajectories T₃ C₁ and T₄ C₂ should have expanded about 33 times. One may conclude that the tsunami took about 25 min. from Thera to Herakleion on Crete, and would have diminished to about 11 meters in height during the transit.

----------------------------------

--------------------------------