Do stratosphere-resolving models make better seasonal climate predictions in boreal winter?

Amy H. Butler (1),(2), Alberto Arribas (3), Maria Athanassadou (3), Natalia Calvo (4), Andrew Charlton-Perez (5), Daniela Domeisen (6), Maddalen Iza (4), Alexey Karpechko (7), Craig MacLachlan (3), Alan O’Neill (5), Adam Scaife (3), and Michael Sigmond (8)

Abstract
Using an international, multi-model suite of historical forecasts from the World Climate Research Programme (WCRP) Climate-system Historical Forecast Project (CHFP), we compare the difference in seasonal prediction skill in boreal wintertime between models that resolve the stratosphere and its dynamics (“high-top”) and models that do not (“low-top”). We find that overall the high-top ensemble mean has no more skill in forecasting the wintertime North Atlantic Oscillation than the low-top ensemble mean, and skill varies widely between individual models. Increasing the ensemble size increases the skill. We then briefly examine two major processes involving stratosphere-troposphere interactions (the El Niño-Southern Oscillation/ENSO and the Quasi-biennial Oscillation/QBO) and how they relate to predictive skill on seasonal timescales, particularly over the North Atlantic and Eurasia. High-top models tend to have more realistic stratosphere-troposphere coupling related to ENSO and the QBO, which may enhance wintertime skill over high-latitudes in these models compared to low-top models during winters with ENSO or QBO forcing. However, it is not clear whether the improvement in skill in the high-top models is due entirely to better stratospheric processes.