The General Picture of the Climate during the time frame of the Urban Mind project

Below follows a summary of a first screening of available literature on climate patterns in the eastern Mediterranean-Middle East region, covering the last ca 12000 years. The review aims to give a general overview of available data but does not include a critical evaluation of quality and uncertainties inherent in the records or a deep analysis of local to regional patterns. This is a task that will form part of the ongoing research. All sites mentioned are indicated in the map (Fig. 1).

Younger Dryas (YD) (c. 12.7 – 11.5 cal ka BP, c. 10750–9550 BCE)
The Northern Hemispheric cold event occurring during the deglaciation seems to have been a period of aridity and cool temperatures in most of the eastern Mediterranean region, as deduced mainly from pollen studies and lake salinity levels (Rossignol- Strick, 1995, Stevens et al., 2001, Snyder et al., 2001). In a review of paleoclimate records from the Levant, Robinson et al. (2006) suggested that YD was an extremely arid period with regionally colder temperatures in comparison to the Holocene. In the marine environment (Aegean Sea) cold conditions are evident from low numbers of foraminifera species with an affinity for warm temperatures (e.g. Rohling et al., 2002). In the Marmara Sea, however, an anomalous increase in temperature is observed during the YD overall cold event, which is explained by a possible influx of slightly warmer Mediterranean water (Sperling et al., 2003).

The emerging picture of cold and arid conditions during the Younger Dryas chronozone does not seem to be well recorded in some palynological records from northern and north-western Greece (Lawson et al., 2004, 2005). The expected expansion of temperate taxa (e.g. Pinus and deciduous Quercus) is not well established in these records, however, the authors ascribe this to micro-climate effects such as topographical setting and not as an indication of a less prominent YD event (Lawson et al., 2004, 2005). This is supported by other proxies indicating cold temperatures and aridity from adjacent areas (e.g. Combourieu-Nebout et al., 1998 and Wilson et al., 2008). Combourieu-Nebout et al. (1998) pointed out an increase in pollen from Artemisia and Ephedra in a core from the Adriatic Sea as an indicator of semi-desert conditions. Further, Wilson et al. (2008) infer cold and arid conditions in NW Greece from a diatom record. A hiatus in stalagmite formation in Poleva cave in Romania (Constantin et al., 2007) during YD may also indicate cold conditions.

Early Holocene (c. 10.5 – 6 cal ka BP, c. 8550–4050 BCE)
A general warming trend marks the onset of the Holocene. A rapid expansion of the taxa Quercus (oak), indicates good moisture conditions, and of Pistacia, suggesting higher temperatures in the eastern Mediterranean region (Rossignol-Strick, 1995). Further, an alleviated climate situation is indicated by e.g. pollen and diatom records from north-western Greece (Lawson et al., 2004, 2005; Wilson et al., 2008) and western Iran (Stevens et al., 2001). The Anatolian peninsula experiences more humid conditions, between 8650 – 6950 BCE (10600 – 8900 cal yrs BP), both in comparison to the present day situation, and to the conditions at YD, (Roberts et al., 2001; Eastwood et al., 2007).

Rising lake levels in the Dead Sea at c. 8050 BCE (c. 10,000 cal yrs BP) indicating more moisture (Robinson et al., 2006), is followed by a rapid drop in lake levels around 6150 BCE (8,100 cal yrs BP) suggesting a return to more arid conditions at this time, eventually reflecting the wide-spread so called 8.2 cold event.

In the Negev desert precipitation increased in early Holocene, as deduced from changes in the stable carbon isotope composition of shells, interpreted to reflect changes in vegetation composition (Goodfriend, 1999). In Jordan Hunt et al. (2004) found evidence for a general wetter early Holocene but interrupted by decreasing rainfall at around 8,000 (unclear whether calibrated age or not).

Warming is also manifested in marine records (Emeis et al. 2000; Sangiorgi et al., 2003; Sperling et al., 2003), where, for example, the Marmara Sea displays a warm period between 7050 and 3750 BCE (9 and 5.7 ka (cal.)) (Sperling et al., 2003).

Mid-Holocene (c. 6 – 3 cal ka BP, c. 4050–1050 BCE)
The mid-Holocene in the eastern Mediterranean is a period with contrasting paleoclimate data. More arid conditions in parts of the region concur with more humid phases at other places (cf. Staubwasser & Weiss, 2006). Roberts et al. (2004) suggested an overall expansion of tree taxa in the region during the mid-Holocene possibly reaching its high around 6 kyr 14C BP (calibrated by Finné using oxcal to c. 4.9 cal kyr BP/ 2950 BCE). After this xerophytic (drought resistant) taxa expanded probably as a response both to climate change and anthropogenic impact (Roberts et al., 2004).

In western Iran the mid-Holocene period is characterised by more Quercus pollen registered in both the Lake Zeribar and Lake Mirabar sediment records, indicating restricted summer droughts (Stevens et al., 2001, 2006). Both records however indicate a drought event around 3450 BCE (5,400 cal yrs BP). This dry event is observed as a prolonged dry period in southwest Turkey between c. 3550 and 1550 BCE (c. 5,500 cal yrs BP and 3,500 cal yrs BP) (Eastwood et al. 2007). A drier hydroclimatic regime in central Turkey is indicated by records from Eski Acigöl (Roberts et al., 2001), showing a decline of mesophilic tree taxa in the area (Roberts et al., 2001).

In the Negev desert the mid-Holocene seems to have been wetter than present but drier than the early Holocene (Goodfriend, 1999). At the Dead Sea just north of the Negev Migowski et al. (2006) identified a period with low lake levels between 6150 and 3650 BCE (8,100 and 5,600 cal yrs BP) corresponding to a dry phase. This dry period was replaced by a wetter phase between 3650 and 1550 BCE (5.6 and 3.5 cal kyr BP) with a subsequent rising lake level (Migowski et al., 2006).

After c. 5550 BCE (c. 7.5 cal kyr BP) a rapid increase in regional aridity is suggested by a higher input of aeolian sediments coinciding with an increase in salinity evident 8(15) in a core from the Red Sea (Arz et al., 2003). In the Marmara Sea record (Sperling et al., 2003) a cooling trend during the mid-Holocene is evident. A winter season cooling around 4050 BCE (6000 cal yrs BP) of the Adriatic Sea is suggested by Sangiorigi et al. (2003) and a cold spell around the same time is proposed by Rohling et al. (2002).

Late Holocene (c. 3 cal ka BP – present, c. 1050 BCE – present)
Records from the late Holocene may be subject to anthropogenic influence which could hamper climate interpretations. This is especially true for palynological studies, as mentioned above. This period is initially signified by a general cooling following a Holocene climatic optimum around 2050 CE (4000 cal yrs BP) (Johnsen et al., 2001). Cores from the Adriatic and the Aegean Seas both indicate cooling around 1050 BCE (3 kyr cal BP) (Rohling et al., 2002; Sangiorig et al., 2003).

From palynological data and lake level changes, Neumann et al. (2007) identified the following pattern of dry and wet phases in the Dead Sea area: 1500–250 BCE was mainly a dry period. This period was interrupted by a short wet spell around 1250 BCE. A longer wet period replaced the arid conditions between c. 100 BCE and 500 CE with an interruption by a shorter period of aridity at 250 CE. The time from 500 CE to present seems to have been mainly dry with a shorter wet spell at c. 1300 CE.

Migowski et al.’s (2006) lake level reconstruction also indicates low lake levels in the Dead Sea from c. 1550 BCE (c. 3.5 kyr cal BP). From the region west of the Levant mainly arid conditions seem to have prevailed. Zalat and Vildary (2007) suggested that the northern Nile Delta was characterised by aridity, however, alternated with short moist episodes during the late Holocene.

There are several dendrochronological records from Anatolia and the north Aeagean used for reconstructing spring precipitation on annual and sub-decadal scales. These reconstructions indicate that persistent drought events have not occurred during the last millennium of the Holocene (e.g. Touchan & Hughes, 1999; Griggs et al., 2007). Touchan et al.’s (2007) study in southwest Anatolia suggests that the three 70-year periods with the most humid May-June precipitation are: CE 1518–1587, 1098–1167 and 1743–1812. Their reconstruction indicated that the following periods were the driest 70-year periods: CE 1195–1264, 1434–1503 and 1591–1660 (Touchan et al., 2007). From lake levels in Lake Nar Gölü in central Turkey Jones et al. (2006) inferred two dry periods at: CE 300–500 and 1400–1950. The authors also concluded that two wetter intervals occurred at: CE 560–750 and 1000–1350) (Jones et al., 2006). In the eastern most part of the region covered by the Urban Mind project Aral Sea lake levels for the last 1600 years have been reconstructed indicating low stands at: CE 400, 1195–1355 and 1780– present (Austin et al., 2007).