The future climate of the Mediterranean

Abstract

The main objective of the thesis was to model future climate scenarios for the Mediterranean region particularly C02, NH4 and N20 concentrations, global mean temperatures, global mean sea level rise, precipitation and pressure. For modelling MAG ICC and SCENGEN, which are readily available, coupled computer models, were used. The models were run once the required properties were chosen as well as the particular regions studied. The runs undertaken showed that greenhouse gas concentration predictions exhibit a marked rise for C02 and N20 whilst those for NH4 exhibit a decline once a maximum concentration is reached. The main results obtained regarding temperature were that with increased greenhouse gas concentrations, particularly C02, temperature will definitely increase. Also, the greater the climate sensitivity being considered, the greater the resultant increase. Global mean sea level will also experience a rise, mainly as a result of increased temperature. This rise in global temperature causes phenomena such as the melting of polar ice caps, which in turn lead to increased sea level. Precipitation predictions have a great uncertainty due to model errors, therefore predictions made are more tentative. These depend largely on the region being considered and the scenarios being used. The areas with a predicted decrease in precipitation will suffer periods of dry spells, while others will experience an increase in mean precipitation. Also, it was noted that whilst precipitation may decrease overall annually, there will be increased precipitation over a shorter period, which may lead to an increased risk of flooding. Analysis of probability of increase in temperature yield a result of one which leads to the conclusion that the three regions studied in the Mediterranean will experience warming. Analysing the inter-model SNR for temperature, all results obtained, for all three regions and using both scenarios, are greater than 4. This means that results and predictions made are robust. Finally, the annual change in temperature variability expressed as the inter-model signal-to-noise ratio was considered. The results obtained for the annual inter-model signal-to-noise ratio for change in temperature variability, are all less than unity or approximately equal to it, illustrate a great deal of uncertainty for predicted changes in temperature variability.