The Cold‐Pole Bias Severely Weakens Southern Hemisphere Springtime Stratosphere‐Troposphere Coupling

Abstract A too cold and too strong Antarctic stratospheric polar vortex (ASPV) in spring (the so‐called cold‐pole bias) is a common model problem. This study investigates the impact of the cold‐pole bias on Southern Hemisphere springtime stratosphere‐troposphere coupling and how this impact is affected by interactive ozone using a pair of Goddard Earth Observing System (GEOS) simulations with and without interactive chemistry. The cold‐pole bias in the GEOS simulations delays the poleward and downward progression of the ASPV and stratosphere‐troposphere coupling in spring by 1–2 months, causing severe underestimation of stratosphere‐troposphere coupling in October–November. Consequently, the simulations poorly capture or completely miss the observed springtime tropospheric predictability from ASPV conditions in late winter/early spring. Compared to the prescribed ozone simulation, interactive ozone exacerbates the cold‐pole bias by overpredicting Antarctic ozone loss, leading to degradation of springtime stratosphere‐troposphere coupling and loss of tropospheric predictability.