Marquet, Pascal (Meteo-France)
The regional climate impacts of the anthropogenic greenhouse gas increase are of primary interest. One possible tool for such investigations is a General Circulation Model with high resolution over a part of the globe.
The ARPEGE-IFS climate model is well designed for such studies. A stretched-grid and tilted-pole climate version of this model has been used to study the atmospheric response over Europe to a CO2 doubling. This version possesses a variable resolution with a basic truncation of T63 and a stretching factor of 3.5.
Accordingly, the resolution varies continuously from 60 km in the Mediterranean sea (T220) to 700 km in the southern Pacific Ocean (T20). This model can be considered at the same time as a GCM far from the region of interest, including a Regional Climate Model with a fine representation of the Mediterranean basin orography and a realistic land-sea mask. The advantage of the stretched-grid model is that there is no problem of coupling with lateral boundaries since there is no independent large-scale driving GCM. It is observed that the simulated climate over Europe and North Africa is better than the climate obtained with a standard horizontal resolution of 300 km (T42 GCM).
Two 8 year simulations have been performed at CNRM, a control experiment with 1xCO2 SST and an anomaly experiment with 2xCO2 SST. These SSTs come from a 2-century simulation with the Hadley Centre coupled model. In this simulation starting at 1860, the increase in CO2 concentration follows the observed values till 1990, then goes on at a constant 1% per year rate. Two 8-year time slices, the one starting in 1985, the other in 2055, have been extracted and the Hadley Centre SSTs have been used as a forcing.
Results do not show a major disagreement with older experiments obtained with our standard climate model T42. The major impact of doubling the CO2 concentration is a warming in all seasons, with an increasing in precipitation over western Europe in winter. But it must be noticed that these predicted changes are of the order of magnitude of the errors of GCM's when they are compared with observed 1xCO2 climate. It is clear that improvements are required in the quality of large-scale circulation in the ARPEGE-IFS GCM in order to be more confident with the result of regionalizing the impacts of climate change scenario. Special attention must be paid to the synoptic waves and the blocking over east Atlantic and western Europe, where a too zonal flow is simulated.