Estimating changes in extreme snow load in Europe as a function of global warming levels

Most European regions regularly experience large snow loads due to large snow accumulations and/or intense snowfalls. In the future, the magnitude of extreme snow loads is generally expected to decrease due to global warming. However, this decrease depends strongly on the latitude and the elevation. This study aims to provide a flexible statistical framework to estimate changes in extreme snow load (50-year return levels) in Europe as a function of global warming levels, using a multi-model ensemble of snow cover projections. Different sets of nonstationary models are introduced to accommodate the nonzero probability of experiencing years without snow. In particular, parsimonious distributions (exponential, Gamma, Inverse-Gamma) are considered when zero SWE maxima are present. This approach is illustrated by an application to European regions. In a + 3 °C world, extreme snow loads are projected to decrease strongly in all of Europe compared to a warming level of +1 °C, which corresponds roughly to the current climate. • We provide a flexible framework to estimate changes in extreme snow load in Europe as a function of global warming levels. • We introduce nonstationary models dedicated to regions where snowfalls are rare (low-elevation or low-latitude regions). • In a + 3 °C world, extreme snow loads are projected to decrease strongly in all of Europe compared to the current climate.