Modeling the Implications of Climate Change on Lightning, Hail and Wind in Europe

The relation between climate change and severe convective storms in Europe is not well understood. A challenge is that the resolution of most reanalyses and climate models is too coarse to explicitly resolve convective storms or their associated hazards like large hail or severe wind gusts. Therefore, we studied the relations between the occurrence of convective storms and a range of atmospheric parameters which can be derived from reanalysis or climate model data. We developed an additive regression convective hazard model (AR-CHaMo) for thunderstorm occurrence and severe weather events such as large hail and severe wind gusts, taking into account atmospheric parameters representing instability, moisture or shear.

The predicted probability of a hazard is the product of the probability that a storm occurs, and the probability of a hazard given the presence of a storm [P(hazard) = P(storm) x P(hazard|storm)]. The models were developed for the hazards hail 2 cm and 5 cm as well as for severe wind 25 m/s using EUCLID lightning data and ESWD hazard reports across central Europe for the 2008-2016 period. The method AR-CHaMo was evaluated within the time period 2008 and 2016 for central Europe using multiple statistical verification metrics.

By applying AR-CHaMo to ERA-Interim reanalysis, changes were detected in thunderstorm and hazard occurrence for the past climate (1979-2016): The frequency of thunderstorms and all related hazards generally increased due to increased latent instability. Decreases in southwest Europe result from mid-tropospheric drying. Since the frequency of hazards has increased more rapidly than that of thunderstorms, it was found that thunderstorms have become more likely to produce severe weather.

AR-CHaMo enables to study future changes in the frequency, intensity and spatial distribution of severe convective weather accounting for individual hazards. An ensemble of 14 members of the EURO-CORDEX climate simulations was used for the historical period (1971-2000) and up to 13 members for three RCP scenarios (RCP2.5, RCP4.5 and RCP8.5) for two future periods (2021-2050 and 2071-2100). Increases are found for central and eastern Europe both for thunderstorms, and for all hazards, which are largest and also most robust for the RCP8.5 scenario and at the end of the 21st century. In that scenario, the increase amplifies during the second half of the century in contrast to the other scenarios (RCP2.6 and RCP4.5). Inter-model variability is caused by both the driving global climate models and the regional models. The most impactful robust increase of up to 160% is projected for very large hail in central and eastern Europe. The changes are generally smaller for thunderstorms, and larger and more robust for the considered hazards. Therefore, thunderstorms are modeled to become more likely to produce severe weather in the future.