Assessing the role of internal variability on projections of Northern Europe surface air-temperature at near-term (2020-2039) using a storyline approach

Despite ever-increasing anthropogenic forcing due to continuing GHG emissions, a slowdown in the rate of global warming has been monitored in the early 2000s and has been mostly attributed to the effect of internal climate variability. The weight of internal variability with respect to external forcing increases at regional scale and it is crucial to understand and quantify its role in future climate outcomes. In this study, we assess the near-term climate change over Europe by contrasting a suite of large ensemble of socio-economic pathways (SSP) projections conducted over 2020-2039, with historical simulations over 1995-2014 used as reference. The uncertainty associated with future anthropogenic forcing is taken into account by analysing 4 SSP projections (SSP1-2.6, SSP2-4.5, SSP3-7.0, SSP5-8.5), while the one associated with internal variability is assessed through the large ensembles of 30 members. Projections averaged over 4 SSP give a mean warming of +1.1°C over Northern Europe in Dec-Feb winter, as quantified by the difference between 20-yr averages over 2020-2039 compared to present-day (1995-2014). Since the signal (related to the forced response) to noise (related to internal variability) is small in all scenarios, ranging from 0.2 for SSP3-7.0 to 0.5 for SSP5-8.5, the 4 near-term ensembles are combined to create a super ensemble of 120 members to further evaluate the role of internal variability, whose likely range, defined by the 5th and 95th percentiles, is equal to 3.0°C for the region of interest. We apply the so-called storyline paradigm and find 4 families of future outcome, each family corresponding to a couple of given phases of the two main drivers of internal variability for Northern Europe winter temperature, namely the Atlantic meridional overturning circulation (AMOC) and the north Atlantic oscillation (NAO). More than a statistical classification, these storylines provide physically consistent future outcomes. The storyline characterised by stronger AMOC and positive NAO leads to the strongest warming, +2.3°C (+1.7 to +3.0), and has a large increase in precipitation, +10% (+5.2 to +16), compared to the storyline where AMOC is reduced and the NAO negative, leading to +0.19°C (-0.46 to +0.86) and +1.9% (-3.3 to +11) for temperature and precipitation only. Regarding temperature extremes, we show, regardless of the scenario, a clear decrease in the probability of extremely cold days in the near-term, and a statistically significant increase in the frequency of extremely warm days that is expected to double. There is, however, a larger dependence of extremes to the storyline family. In particular, the probability of occurrence of exceptionally warm days, defined by the 99th percentile level, is 4 times more likely in the combined AMOC and NAO positive phases storylines, while it is almost null for the counterpart one. We show here evidence that the storyline approach is a clear added-value to understand the role of internal variability in future climate and that it provides actionable information to users in presence of related irreducible variability uncertainties.