Climate modelling sensitivity experiments for the Messinian Salinity Crisis

The repeated isolation of the Atlantic Ocean during the Messinian led to at least the partial desiccation of the Mediterranean Sea and to the deposition of massive evaporites (Messinian Salinity Crisis). Overall, proxy data do not indicate dramatic climatic changes in the Mediterranean region for the time of the Messinian Salinity Crisis but a constant climate seems to be unrealistic at first sight. To analyse the climate response to environmental changes in the Mediterranean region during the Messinian Salinity Crisis, we perform climate modelling sensitivity experiments. As a reference base, we use a Tortonian simulation (TORT) and simulate three Messinian sensitivity scenarios: (1) the Mediterranean Sea surface 1500 m lower than the normal sea level (MSC LOWSEA); (2) the Mediterranean basin completely desiccated, the resulting land surface 1500 m lower and covered by grassland vegetation (MSC LOWGRASS): and (3) the Mediterranean Sea simply replaced by grassland vegetation (MSC GRASS). All sensitivity experiments represent extreme scenarios for analysis of general patterns of Messinian climate change. In MSC LOWSEA, mean annual temperatures in the Mediterranean basin increase by +2 degrees C, and surrounding regions get cooler by up to -3 degrees C as compared to TORT. The lowering of the Mediterranean Sea level increases the annual precipitation in the south-western part of Europe and in northwestern Africa (up to + 1000 mm/a). In MSC LOWGRASS as compared to TORT, the Mediterranean basin strongly heats up by +7 degrees C, while temperatures in the surrounding areas do not change more than +/- 2 degrees C. The precipitation in the Mediterranean basin significantly decreases (up to -600 mm/a). Precipitation in the areas surrounding the Mediterranean basin does not change substantially in MSC LOWGRASS, but rainfall in W-Africa increases up to +600 mm/a. In MSC GRASS as compared to TORT, temperatures in the Mediterranean region increase by +4 degrees C. This is twice as high as the temperature increase which resulted from lowering the Mediterranean Sea surface in MSC LOWSEA. Annual rainfall over the Mediterranean basin in MSC GRASS is -400 mm/a lower than the rainfall in TORT. Proxy-based climate reconstructions for the time of the Messinian Salinity Crisis indicate no dramatic climatic changes around the Mediterranean. In contrast to this, for eastern N-Africa and Eastern Europe climate changes are indicated. The MSC LOWSEA and MSC LOWGRASS sensitivity experiments show that the strong environmental changes in the Mediterranean realm during the MSC did not necessarily cause strong climatic changes at the margins of the Mediterranean basin. Therefore, the present modelling studies are in agreement with climate proxy data from the Mediterranean region. Additionally, our results principally agree with the climate changes inferred from the fossil record of Central and Eastern Europe. For eastern N-Africa, our modelling studies contradict proxy data. Based on our results, weak climatic changes indicated by the fossil record at the Mediterranean coastlines are realistic. (C) 2010 Elsevier B.V. All rights reserved.