Quantifying 21st-century France climate change and related uncertainties.

Abstract: We tackle here the question of past and future climate change at sub-regional or country scale with the example of France. We assess France climate evolution during the 20th and 21st century as simulated by an exhaustive range of global climate simulations. We first show that the large observed warming of the last 30years can be simulated only if anthropogenic forcings are taken into account. We also suggest that human influence could have made a substantial contribution to the observed 20th century multi-decadal temperature fluctuations. We then show that France averaged annual mean temperature at the end of the 21st century is projected to be on the order of 4.5K warmer than in the early 20th century under the radiative concentration pathways 8.5 (RCP8.5) scenario. Summer changes are greater than their winter counterpart (6K versus 3.7K). Near-future (2020–2049) changes are on the order of 2.1K (with 2.6K in summer and 1.8K in winter). Model projections also suggest a substantial summer precipitation decrease (−0.6mm/day), in particular over southern France, and a moderate winter increase, (0.3mm/day), mostly over the northernmost part of France. Uncertainties about the amplitude of these precipitation changes remain large. We then quantify the various sources of uncertainty and study how their ranking varies with time. We also propose a physically-based metric approach to reduce model uncertainty and illustrate it with the case of summer temperature changes. Finally, timing and amplitude of France climate change in case of a global average 2-K warming are investigated. Aggressive mitigation pathways (such as RCP2.6) are absolutely required to avoid crossing or barely exceeding the 2-K global threshold. However, France climate change requiring adaptation measures is still to be expected even if we achieve to remain below the 2-K global target. [Copyright &y& Elsevier]