Modélisation des climats futurs certitudes et incertitudes

Click to increase image sizeClick to decrease image size

Sïnce the middle of the nineteenth century the composition of the Earth atmosphere has begun to change in a manner which is unprecedented over the Quaternary era. The carbon dioxid concentration, for example, had been oscillating between extrernal values of 180 ppm (during glaciations) and 280 ppm (during interglacial eras). It had stayed almost constant over the last 10000years, but has now reached a value of 360 ppm, much of the increase taking place during the last decades. Similarly methane concentration has been multiplied by more than a factor 2, and other gases (nitrous oxides) have also experienced an exponenrial growth. There is no doubt that this situation is the result of human activities: energy consumption, industrial or agricultural activities, deforestation. As most of these gases have a long residence time in the atmosphere, where they tend to accumulate, we may expect the level of these perturbations to increase strongly throughout the 21st century. To diagnose the possible impact of these trends in terms of climate impact, the only available tools are numerical models. These numerical models constitute in a way virtual planets, where the atmospheric and oceanic flow, the continental hydrology, and their interactions, are represented through physical equations. Models are not a perfect representation of the Earth system, and they will never be. We also cannot expect the climate system to he fully predictable. But models are strongly constrained by conservation equations, they behave in many aspects like the real planet and reproduce quite realistically the mean geographical and seasonal fluctuations of the precipitation, the temperature or the winds as well as some of the most important natural fluctuations at intraseasonal or interannual (El Nino, North Atlantic Oscillation) time scales. So models are useful to describe a range of possible future climates. Present projections indicate in all cases a significant change, with a global surface warning in 2 100 between 2^oC and 6^oC: half of the uncertainry is due to the economic projections, half is due to the complex behavior of the climate system (in particular: clouds). 1t is striking that the lower bound of these estimates already represents an important perturbation. The changes in temperature would induce changes in precipitation rates or areas, in storminess or cyclone tracks - these regional consequences are however difficult to predict in details. We may expect the sea level to rise from 20 to 90 cm. Changes in a more distant future (2 centuries) might be much larger, although such projections are less reliable. An important feature of the changes to come may be their partially unpredictable character: many of the local events accompanying a global climate change will arise as surprises. In that sense the rate of change will be the primary factor of danger. And if it is probably too late to stop climate modification, there is sill time to favor a situation where adaptation to its impacts will be much easier.