Summary/Abstract
With record breaking heat waves, dryness, and floods in several parts of the world in recent years the question arises whether and to what extent the hazard of hydrometeorological extreme weather events has changed, and if changes can be attributed to specific causes. The methodology of probabilistic event attribution allows to evaluate such potential changes in the occurence probabilities of particular types of extreme events. We show that such a probabilistic assessment not only provides information on changing hazards in hydrometerological events but also permits statements about multivariate combinations of hydrometeorological variables. Hence attribution studies could be targeted specifically towards relevant impacts in particular sectors, if different suitable multivariate proxies are used. We demonstrate our methodology by using combinations of temperature, precipitation and humidity in the large ensemble of climate simulations within the weather@home project in order to derive impact-relevant quantities such as a seasonal water balance and heat stress imposed on the human body. Finally, we estimate the hazard probabilities of those events in South-East Europe, a region that has recently experienced severe summer dryness (2012) in combination with multiple heat waves.