Impacts of earth system models’ systematic errors on tropical–extratropical teleconnections

Atmospheric teleconnections drive global climate variability, shaping weather extremes and their socio-economic impacts. As one of the major sources of predictability at mid-latitudes, their accurate simulation is essential for providing reliable climate predictions and supporting effective adaptation strategies. This manuscript investigates the impacts of Earth System Models' systematic errors on key simulated teleconnection patterns. Three case studies are investigated: the El Niño–Southern Oscillation (ENSO)–North Atlantic teleconnection in boreal early-winter, the Indian Ocean influence on Chile in austral early-winter, and the interference of austral early-summer Indian and Pacific Ocean teleconnections with Southern Africa. By analyzing a large ensemble of CMIP models alongside observational constraints, these teleconnections are found to be primarily driven by planetary-scale Rossby wave trains. Model deficiencies in capturing these teleconnections stem from pervasive biases in the upper-tropospheric waveguide and simulated Rossby wave trains. In the Northern Hemisphere, overly strong subtropical Pacific and Atlantic jets trap wave activity to lower latitudes, weakening the boreal early-winter ENSO signal over the North Atlantic. In the Southern Hemisphere, a weaker eddy-driven jet displaces Rossby wave rays, thereby degrading the austral early-winter Indian Ocean teleconnection over Southern America. Lastly, models failing to simulate the austral early-summer interference of Indian and Pacific Ocean teleconnections with Southern Africa systematically misrepresent Rossby wave trains over the South Atlantic, most likely owing to an underestimation of ENSO variability over the eastern Pacific. These waveguide biases are linked to background-state errors, consisting of persistent surface temperature biases. The stronger subtropical jets may result from cold mid-latitude surface conditions via thermal wind balance, while the weaker eddy-driven jet stems, through reduced baroclinicity, from an overly warm Antarctic sector and underestimated sea-ice cover. These findings underscore the relevance of mean-state fidelity in achieving reliable climate simulations, demonstrating that reliable representation of teleconnections is essential to harness their predictive potential.