Information Sheet on Future Climate and Impacts in the Coastal Case Studies: Gulf of Gabès

INFORMATION SHEET ON FUTURE CLIMATE: CLIMATE PROJECTIONS FOR THE CIRCE CASE STUDIES

Summary ►A set of six coupled Mediterranean Sea-atmosphere CIRCE climate models form the basis of climate and marine projections for the CIRCE case studies. ►All models indicate consistent warming by the 2050s, which is higher for Tmax than for Tmin, and for western Mediterranean case studies (Gulf of Valencia, Gulf of Oran) compared to eastern Mediterranean case studies. ►Precipitation projections are less consistent than for temperature, but a general decline is indicated for 8 of the 11 case study sites. ►Additional numbers of ‘very hot summer days’ projected for the 2050s range from 15 for the Gulf of Valencia to 5 for Alexandria. 1. CIRCE climate models As part of the CIRCE project, a set of Global day (1961-90) observations. Changes are Climate Model (GCM) and Regional Climate represented as the difference between the future Model (RCM) simulations has been performed 30-year period (2021-2050) and the baseline 30- with the key innovative attribute of including a year period (1961-90). The models indicate realistic representation of the Mediterranean Sea consistent warming over the entire region which coupled with the atmosphere. The six coupled is strongest in summer (+1.2°C to +2.5°C) and CIRCE models vary by type, spatial resolution weakest in winter. The direction of change and boundary forcing (Table 1). The spatial averaged over the entire Mediterranean area is resolution of the Mediterranean Sea ocean inconsistent for precipitation, with a summer components is between 7 and 12 km, while the increase for four models, and a summer decrease atmospheric component of the two global for two models. Climate and marine indicators models varies between about 50 and 80 km, and have been constructed for each of the coastal and that of the regional models, between 25 and 30 rural regions using the average of land-based km. All simulations use the IPCC-SRES A1B model grid boxes, and for the urban case studies emissions scenario and were run for the period using the nearest land-based grid box (checking 1950-2050. The CIRCE model ensemble has a for neighbouring inconsistencies). cold bias of about 2°C compared with present- Table 1: Summary description of the CIRCE models. Temperature (T) changes (°C) and Precipitation (P) changes (mm per season) for winter (djf) and summer (jja) are averages for all model land and sea grid boxes over the Mediterranean region Seasonal mean response over the Model Type of model, spatial resolution of the Mediterranean for 2021-2050 with acronym atmospheric component and boundary forcing respect to 1961-1990 Tdjf Tjja Pdjf Pjja CNRM Stretched grid global model zooming to 50 km over 0.78 1.42 -3 -3 the Mediterranean area INGV Global model: ~80 km 1.44 1.68 -23 -8

IPSL1 LMDZ regional model: 30 km. Forcing from an 2.01 2.46 -10 -5 (IPSLreg) earlier (non-CIRCE) IPSL GCM run IPSL2 Coupled LMDZ global and LMDZ regional model: 1.00 1.16 +6 -6 (IPSLglo) 30 km ENEA Regional model: 30 km, forcing from the CIRCE 0.97 1.41 -15 +6 INGV model. MPI Regional model, 25 km, forcing from an ECHAM5 1.38 1.52 -3 +1 GCM (non-CIRCE) run

1 Case study projections: mean temperature

Projected temperature changes are slightly larger for case studies located in the western (Gulf of Valencia and Gulf of Oran) and central (Tuscany and Puglia) Mediterranean than in the eastern Mediterranean (Figure 1). Changes in mean Tmax (maximum temperature) tend to be larger than changes in Tmin (minimum temperature), particularly in the western Mediterranean. The largest ensemble-mean change is for Tmax in the Gulf of Oran almost 2°C). The IPSL1 model system consistently indicates the greatest warming, while the ENEA (regional) and CNRM (stretched grid) models tend to give the least warming.

Figure 1: Histograms of (top) annual Tmax (°C) and (bottom) Tmin (°C) changes (2021-2050 minus 1961- 1990) based on CIRCE model data for the urban, rural and coastal case studies. The ensemble-mean change is shown (bars), together with changes for the five individual models. IPSL= IPSL1/IPSLreg.

2 Case study projections: temperature extremes

A selected number of climate extremes were selected as case-study specific indicators. For example, projected changes in the number of ‘very hot summer days’ (Tx95n) are shown for all case studies in Figure 2. The spatial pattern in the simulated future response is similar to Tmax being largest for the western case study of the Gulf of Valencia (ensemble mean of about 13 more ‘very hot days’ per summer) and least for the south-eastern case study of Alexandria (ensemble mean of about 5 more ‘very hot days’ per summer). There is a large model spread in the response: the IPSL1 model generally projects the greatest increase (more than 25 additional days for the Gulf of Valencia and Tuscany) while the least increase is generally for the MPI model (less than five days).

Figure 2: Histograms of changes (2021-2050 minus 1961-1990) in the number of very hot summer days (Tx95n) based on CIRCE model data for the urban, rural and coastal case studies. The ensemble-mean change is shown (bars), together with changes for the five individual models. IPSL= IPSL1/IPSLreg.

Other temperature extremes considered (not shown) include very hot nights for the urban case studies, which show the largest increase for Athens and for the summer season. Additional regional climate model sensitivity studies undertaken for the CIRCE urban case studies suggest that these results underestimate the frequency of extreme temperature events due to the exclusion of urban heat island effects.

3 Case study projections: precipitation

Total annual and extended winter season (October to March) precipitation were also selected as climate indicators (Figure 3). The ensemble mean suggests a decline in precipitation totals for all case studies except for Alexandria and the West Nile Delta and little change in the Judean Foothills. There is less spatial or inter-model consistency in precipitation projections than for temperature.

Figure 3: Histograms of changes (in mm for 2021-2050 minus 1961-1990) in (a) annual and (b) extended winter (October to March) total precipitation based on CIRCE model data for the urban, rural and coastal case studies. The ensemble-mean change is shown (bars), together with changes for the five individual models IPSL= IPSL1/IPSLreg.

Marine conditions

Sea surface temperature (SST) input through the Straits of Gibraltar. Consequently, The CIRCE models show an ensemble-mean cold projected SSS show a generally negative trend, but bias (of about 2°C) relative to the observation the model spread is large, from +0.05 to -0.35 psu. period, 1961-90. Projections for the future period 2021-2050 consistently indicate a warming of SST Sea level change (with values ranging from +0.8 °C to 1.8°C Changes in Mediterranean sea level are assessed compared to the period 1961-1990). through the steric effect [due to volume change – the largest component of sea level change (roughly Sea surface salinity (SSS) 70% of total)]. Models suggest an increasing trend SSS is also underestimated by the CIRCE ensemble in sea level, with values reaching +10 cm in 2021- (by about -0.4 psu for the period 1961-1990). 2050 and nearly +14 cm in 2050 (all relative to the Models suggest an increase in Atlantic freshwater period 1961-1990).

Acknowledgements CIRCE (Climate Change and Impact Research: the Mediterranean Environment) is funded by the Commission of the European Union (Contract No 036961 GOCE) http://www.circeproject.eu/. This information sheet forms part of the CIRCE deliverables D11.3.6, D11.4.5, and D11.5.7. The CIRCE model data were provided by INGV-CMCC (Silvio Gualdi, Enrico Scoccimarro); MF (Florence Sevault, Clotilde Dubois); IPSL-CNRS (Laurent Li); ENEA (Alessandro Dell'Aquila, Adriana Carillo); and MPIM-HH (Alberto Elizalde Arellano). For assistance in processing model data, thanks are extended to Dimitra Founda, Effie Kostopoulou, Basil Psiloglou, Kostas Varotsos (NOA), Gillian Kay (UK Meteorological Office), Dimitrios Efthymiadis and Richard Cornes (UEA).

References ► Gualdi S., Somot S., May W., Castellari S., Déqué M., Adani M., Artale V., Bellucci A., Breitgand J.S., Carillo A., Cornes R., Dell’Aquilla A., Dubois C., Efthymiadis D., Elizalde A., Gimeno L., Goodess C.M., Harzallah A., Krichak S.O., Kuglitsch F.G., Leckebusch G.C., L’Heveder B.P., Li L., Lionello P., Luterbacher J., Mariotti A., Nieto R., Nissen K.M., Oddo P., Ruti P., Sanna A., Sannino G., Scoccimarro1 E., Struglia M.V., Toreti1 A., Ulbrich U., and Xoplaki E. 2011. Future Climate Projections. In Regional Assessment of Climate Change in the Mediterranean: A. Navarra, L.Tubiana (eds.), Springer, Dordrecht, The Netherlands. ► Goodess C.M., Agnew M.D., Giannakopoulos C., Hemming D., Bindi M., Bradai M. N., Congedi L., Dibari C., El- Askary H., El-Raey M., Ferrise R., Founda D., Grünzweig J., Harzallah A., Hatzaki M., Kay G., Kostopoulou E., Lionello P., Mösso C., Oweis T., Psiloglou B., Reale M., Salem S. B., Sanchez-Arcilla A., Senouci M., Tanzarella A. and Varotsos K. 2011. Integration of the climate impact assessments with future projections. In Regional Assessment of Climate Change in the Mediterranean: A. Navarra, L. Tubiana (Eds.), Springer, Dordrecht, The Netherlands.

Authors: Clare Goodess1 ([email protected]), Christos Giannakopolous2 ([email protected]), Maria Hatzaki2 ([email protected]), Maureen Agnew1 ([email protected])

1Climatic Research Unit, School of Environmental Sciences, University of East Anglia, Norwich, UK 2Institute for Environmental Research and Sustainable Development, National Observatory of Athens, V. Pavlou and Metaxa Str., P. Pendeli, GR-15236 Athens, Greece

Date: May 2011; Updated July 2011