资源环境科技发展态势分析平台

Stability indices evaluate the atmospheric instability which is a basic and precursor ingredient needed for storms to develop. In this study, we evaluated changes of some atmospheric stability indices, namely Convective Available Potential Energy (CAPE), Deep-Layer Wind Shear (SHR06) and Severe Weather Threat (SWEAT), for the Iberian Peninsula, for a future climate scenario (2081-2100), considering the greenhouse gases emission scenario RCP8.5, relative to a historical period (1986-2005). The Weather Research and Forecasting (WRF) was forced by ERA-Interim, for validation purposes, and by the Max Planck Institute (MPI) Earth System Model. The novelty of this approach is the study of extreme weather events through the evaluation of conditions favourable to their development instead of directly studying them. The latter approach may be problematic since these phenomena are known to be poorly reproduced by models due to their relatively low resolution and parametrization processes such as clouds and precipitation. Our approach uses stability indices obtained from simulated variables, such as temperature and winds, which are, generally, much better simulated by models. The WRF-MPI model was validated against WRF-ERA. Overall, the WRF-MPI simulates well the three indices considered here, particularly CAPE, when compared to WRF-ERA. Their spatial patterns were similar, although there is a systematic positive bias in the WRF-MPI. This is minimized when we evaluate climate change by computing differences of WRF-MPI simulations between the future climate scenario and the historical period. In the future, it is estimated a significant increase of CAPE and SWEAT intensity, mainly in summer and autumn. It is also expected a decrease of SHR06 intensity in summer and autumn and an increase in the remaining seasons. Therefore, we may anticipate an increase of the probability of occurrence of environments favourable to the development of severe weather, mainly in the Mediterranean, mostly associated to higher CAPE values.