The performance of the European Consortium Earth-system model (EC-EARTH) in the boreal winter stratosphere is comprehensively assessed for the first time, in particular its version 3.3 that contributes to CMIP6. A 100-year long simulation with prescribed climatological boundary conditions and fixed radiative forcing, representative of present-day climate, is used to evaluate the simulation of the climatological stratospheric circulation and to identify model biases. Results show that EC-EARTH has a large issue with the vertical distribution of stratospheric temperature from the tropics to mid-latitudes, seemingly linked to radiative processes of ozone, leading to a biased warm middle-upper stratosphere. Associated with this model bias, EC-EARTH simulates a stronger polar vortex at upper-stratospheric levels while the Brewer-Dobson circulation at middle/lower levels is weaker than reanalysis. The amplitude of the climatological planetary waves is overall underestimated, but the magnitude of the background wave injection from the troposphere into the stratosphere is overestimated, related to a weaker polar vortex at lower-stratospheric levels and, thus, a less effective wave filtering. This bias in the westerly flow could have a contribution from parameterized waves. The overestimation of background wave driving is maximum in early-winter, and may explain the overestimated frequency of sudden stratospheric warmings at this time, as compared to reanalysis. The spatial distribution of wave injection climatology has revealed a distinctive role of the climatological planetary waves: while large-scale waves (wavenumbers 1–2) dominate the eddy heat flux over the North Pacific, small-scale waves (wavenumbers 3–4) are responsible for the doubled-lobe structure of the eddy heat flux over Eurasia. EC-EARTH properly simulates this climatological feature, although overestimates its amplitude over central Eurasia.

https://link.springer.com/article/10.1007/s00382-022-06368-0