Suchergebnisse

Abstract. The Mediterranean basin is a particularly vulnerable region to climate change, featuring a sharply contrasted climate between the North and South and governed by a semi-enclosed sea with pronounced surrounding topography covering parts of the Europe, Africa and Asia regions. The physiographic specificities contribute to produce mesoscale atmospheric features that can evolve to high-impact weather systems such as heavy precipitation, wind storms, heat waves and droughts. The evolution of these meteorological extremes in the context of global warming is still an open question, partly because of the large uncertainty associated with existing estimates produced by global climate models (GCM) with coarse horizontal resolution (~200 km). Downscaling climatic information at a local scale is, thus, needed to improve the climate extreme prediction and to provide relevant information for vulnerability and adaptation studies. In this study, we investigate wind, temperature and precipitation distributions for past recent climate and future scenarios at eight meteorological stations in the French Mediterranean region using one statistical downscaling model, referred as the "Cumulative Distribution Function transform" (CDF-t) approach. A thorough analysis of the uncertainty associated with statistical downscaling and bi-linear interpolation of large-scale wind speed, temperature and rainfall from reanalyses (ERA-40) and three GCM historical simulations, has been conducted and quantified in terms of Kolmogorov-Smirnov scores. CDF-t produces a more accurate and reliable local wind speed, temperature and rainfall. Generally, wind speed, temperature and rainfall CDF obtained with CDF-t are significantly similar with the observed CDF, even though CDF-t performance may vary from one station to another due to the sensitivity of the driving large-scale fields or local impact. CDF-t has then been applied to climate simulations of the 21st century under B1 and A2 scenarios for the three GCMs. As expected, the most striking trend is obtained for temperature (median and extremes), whereas for wind speed and rainfall, the evolution of the distributions is weaker. Mean surface wind speed and wind extremes seem to decrease in most locations, whereas the mean rainfall value decreases while the extremes seem to slightly increase. This is consistent with previous studies, but if this trend is clear with wind speed and rainfall data interpolated from GCM simulations at station locations, conversely CDF-t produces a more uncertain trend.

Abstract. ERA-40 reanalyses, and simulations from three regional climate models (RCMs) (ALADIN, LMDZ, and WRF) and from one statistical downscaling model (CDF-t) are used to evaluate the uncertainty in downscaling of wind, temperature, and rainfall cumulative distribution functions (CDFs) for eight stations in the French Mediterranean basin over 1991–2000. The uncertainty is quantified using the Cramer-von Mises score (CvM) to measure the "distance" between the simulated and observed CDFs. The ability of the three RCMs and CDF-t to simulate the "climate" variability is quantified with the explained variance, variance ratio and extreme occurrence. The study shows that despite their differences, the three RCMs display very similar performance. In terms of global distributions (i.e. CvM), all models perform better than ERA-40 for both seasons and variables. However, looking at variance criteria, RCMs are not always much better than ERA-40 reanalyses, whereas CDF-t produces accurate results when applied to ERA-40. In a second step, a combined statistical/dynamical downscaling approach has been used, consisting in applying CDF-t to the RCM outputs. It shows that CDF-t applied to the RCM outputs does not necessarily produce better results than those from CDF-t directly applied to ERA-40. It also shows that CDF-t applied to RCMs generally improves the downscaled CDFs and that the "additional" added value of CDF-t applied to the RCMs is independent of the performance of the RCMs in terms of CvM, explained variance, variance ratio and extreme occurrence.