Minoan Tephra in Lake Sediments in Western Turkey: Dating the Eruption and Assessing the Atmospheric Dispersal of the Ash

Proxy evidence from tree rings and ice cores suggests a 17th century BC date for a large volcanic eruption, but evidence linking the proxy record with the Bronze Age eruption of Santorini is lacking. The recovery of Minoan tephra in sediment cores from lakes in western Turkey, hundreds of kilometres from the volcano, provides the opportunity to date the Minoan ash layer using uncontaminated material, and to add new information on the atmospheric dispersal of the ash.

Much of the information on the distribution of the Santorini Minoan tephra has come from the presence or absence of the ash in deep-sea sediment cores. Deep-sea core studies suggest an easterly to south-easterly pattern of ash distribution, and a dense rock equivalent (DRE) volume of about 13 km³ for the ash. The discrepancy between the calculated DRE volume of the ash in deep-sea sediments and the volume of the caldera at Santorini led to the suggestion that the Bronze Age eruption was smaller than previously thought (13-19 km³), and that an earlier eruption was responsible for producing much of the present caldera.

More recently non-marine deposits of the ash on Kos, Rhodes, and in western Turkey provide evidence that a major axis of atmospheric dispersal was to the east or north-east. They also suggest that a considerably greater volume of the ash can be expected as more terrestrial deposits are discovered. This lends support to the tree-ring evidence for a two-year cool period attributed to the eruption.

In addition, the presence of the ash layer in sediment cores from lakes in western Turkey provides a tephrochronological marker that will be useful in correlating pollen and stratigraphy studies in the region.

INTRODUCTION

Recent discoveries of Minoan tephra deposits in Turkey provide a new opportunity to evaluate several aspects of the Bronze Age eruption, including its dating, its impact on neighbouring land areas, and the dispersal of the volcanic ash cloud. In this paper I will outline the work currently in progress in western Turkey, and provide new information on the date of the eruption.

While conducting coring for palynological investigations in conjunction with archaeological research at the ancient urban site of Sardis, several cores were taken from Gölcük, a small montane lake in the Bozdağ Range in western Turkey (Fig. 1).

One of the lake sediment cores penetrated a total of 12.8 m of sediment. The sediment core, a graphic log for which is shown in Fig. 2, contained lacustrine mud in the upper 825 cm; the next 1.90 m of the core (from 825-1020 cm depth) was made up of alternating layers of peat and silty lacustrine mud. About midway through this stratigraphic unit was a 12 cm-thick layer of volcanic ash from the Minoan eruption of Santorini (Sullivan 1988). The lake is located 320 km north-east of Santorini.

The contact between the underlying peat and the ash layer is sharp, indicating that the ash was deposited rapidly. The contact with the overlying peat is gradual, suggesting that bioturbation has mixed the ash upward, and/or that ash subsequently delivered to the lake by stream flow and sheetwash from the surrounding slopes was mixed with other sediment en route to the lake.

Peat material from above and below the ash yielded radiocarbon dates of 3110 ± 160 and 7400 ± 120 BP respectively. The ash layer and its origin were not immediately recognized, so no material from immediately below the ash was sampled for radiocarbon dating. By the time the origin of the ash was recognized, the core had been sufficiently disturbed by palynological and sedimentological sampling that a radiocarbon sample was not taken, since any date obtained from the disturbed core would have the potential to be unreliable.

The recovery of Santorini Minoan tephra in the Gölcük core was the first in Turkey. Since then, a second deposit of the Minoan tephra has been identified in a sediment core taken from Köyce iz (see Fig. 3), a small coastal lake in south-west Turkey (Van Zeist et al. 1975).

In their 1975 publication, Van Zeist et al. described a 9-cm-thick layer of grey, sandy material at a depth of 414-423 cm in their core. The layer was about 15 cm below a dated layer yielding a radiocarbon date of 3270 BP, reminiscent of the stratigraphic position of the Minoan ash layer in the Gölcük core. Professor Van Zeist kindly provided a sample of the grey, sandy material. The index of refraction (1.507 ± 0.003) and major element glass chemistry, determined by electron microprobe analysis, are consistent with those of the Minoan tephra. The major element glass chemistry is shown in Table 1.

The recovery of the ash in lakes in western Turkey, together with discoveries of the ash on the islands of Rhodes and Kos, implies a somewhat different pattern of dispersal than that suggested by the deep-sea sediment record. This will be discussed further in a subsequent section. However, the recovery of ash from lake sediments also has implications for radiocarbon dating the Minoan eruption.

TABLE 1. Major element chemistry of the glass shard fraction of the Köyce iz tephra determined by electron microprobe.

Average values for eight sample runs. Expressed in % by weight for each oxide.

DATING THE MINOAN ERUPTION

A universally accepted date for the Minoan eruption of Santorini remains elusive. Despite decades of research, neither archaeologists nor earth scientists have been able to convincingly establish the date of the eruption. A recent spate of articles dealing with the date of the eruption illustrates three of the main difficulties in establishing the Santorini chronology.

The first problem has to do with the inexactitude of the traditional pottery-derived dating series used by archaeologists. Until recently, the ceramic chronology was used to suggest a mid-to late 16th century BC date for the eruption (Warren 1984). Recently, the reliability of this chronology has been questioned, and 'revisionist' chronology suggests a date as much as one hundred years earlier for the eruption (Betancourt 1987; Cadogan 1987).

The second difficulty has to do with absolute dating techniques based on supposed proxy evidence of the eruption. Evidence from ice cores from the Greenland ice cap and tree-ring analysis from western North America and Northern Ireland have been used to assign absolute dates to the Santorini eruption. 

Hammer et al. (1987, 1988a) report that a sulphuric acid peak in an ice core from the Dye 3 site in Greenland, dated to 1644 ± 20 BC, is the result of acid fallout from the Santorini eruption. And, LaMarche and Hirschboeck (1984) and Baillie and Munro (1988) attribute frost rings and low-growth events dating to 1628-1627 BC to the eruption. The rationale is simple and compelling: the Santorini eruption represents the only known natural event for several hundred years of sufficient magnitude and the proper physical characteristics to release enough sulphur oxides and volcanic ash to produce the acid fallout in Greenland and to lower mid-latitude Northern Hemisphere temperatures for two years (cf. Mass and Portman 1989).

The difficulty here is that the Santorini eruption is implicated by default. It is not distinct physical evidence that incontrovertibly links the eruption to the evidence. Rather, chronological coincidence is the compelling argument for attributing the proxy evidence to the Minoan eruption of Santorini. Those who might be hostile to the mid-17th century date, or those who are merely cautious, can point to the fact that while the Santorini eruption is the only known eruption in the appropriate time interval, that does not preclude the possibility that a heretofore unknown eruption could be responsible, as Cadogan (1988) points out.

The final difficulty has to do with the results of radiocarbon dating. One might assume that radiocarbon dating would have provided incontestable evidence for the date of the eruption. However, this has not been the case. Poorly chosen samples from long-lived organic material, small samples that yielded large uncertainties, and the possibility of volcanogenic contamination of samples have produced a wide range of radiocarbon dates for the eruption (Pichler and Friedrich 1976; Hammer et al. 1987, 1988a).

Attempts have been made to resolve some of the radiocarbon date ambiguity by selecting short-lived organic material for dating (Hammer et al. 1987, 1988a), but even this has produced controversial results (Cadogan 1987, 1988; Manning 1988; Hammer et al. 1988b).

The presence of the Minoan tephra in the core from Gölcük provides a new opportunity to attempt to date the eruption with radiocarbon. The lake is located in a basin whose watershed contains no carbonate rocks. It is located far enough from the volcano that volcanogenic contamination is not a problem. And the ash lies atop a layer of peat that can provide material for radiocarbon dating.

It has been found that layers of volcanic ash deposited in uncompacted lake muds may break up and sink into the sediments soon after deposition (Anderson et al. 1984). This could result in the emplacement of an ash layer in older sediments. However, since the Gölcük ash layer is underlain by fibrous peat, and not by lacustrine muds, there should be no problem of ash sinking into uncompacted sediments (Anderson et al. 1984, 506). Furthermore, where sinking of ash layers was detected, the contact between the ash and the underlying sediment is irregular; the contact between the Minoan ash and the underlying fibrous peat was abrupt and regular in the Gölcük core.

New sediment cores will be obtained from Gölcük in late June, 1989. These cores will be used to obtain samples for radiocarbon age determination.

ASH DISPERSAL

Much of the information regarding the magnitude of the Bronze Age Santorini eruption, and the dispersal of the volcanic ash and its effects have come from deep-sea sediment cores. New evidence derived from terrestrial and lacustrine deposits of the ash permit a re-evaluation of the dispersal of the ash, and its impact on western Turkey. Furthermore, these deposits indicate that some earlier estimates of volume of material ejected during the eruption are too low.

Several maps showing the distribution of Minoan tephra recovered in deep-sea cores have been published. The axis of distribution shown on these maps has undergone a progressive shift to the east from the earliest map (Ninkovich and Heezen 1965) to the later maps (Watkins et al. 1978; McCoy 1980). This shift in the distribution pattern is based on the accumulation of an increasing number of cores.

In addition, discoveries of the ash on the Aegean islands of Kos (Keller 1980) and Rhodes (Doumas and Papazoglou 1980; Doumas 1986) provided evidence that the dispersal of the ash was more toward the east than would be indicated by the deep-sea cores. The discovery of Minoan tephra in lakes east and north-east of Santorini reinforce this conclusion.

It seems likely that the south-easterly distribution of at least the distal ash shown in the distribution maps based on deep-sea cores, that is, the ash recovered from south-east of Karpathos, may be largely due to dispersal by marine currents rather than due to atmospheric dispersal. If this is so, continued research into the terrestrial deposits of the Minoan ash may hold great potential for determining the atmospheric dispersal of the ash, and its impact on Bronze Age peoples in the eastern Aegean.

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

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