Summertime Rossby waves in climate models: substantial biases in surface imprint associated with small biases in upper-level circulation

Abstract. In boreal summer, circumglobal Rossby waves can promote stagnating weather systems that favor extreme events like heat waves or droughts. Recent work showed that amplified Rossby wavenumber 5 and 7 show phase-locking behavior which can trigger simultaneous warm anomalies in different breadbasket regions in the Northern Hemisphere. These types of wave patterns thus pose a potential threat to human health and ecosystems. The representation of such persistent wave events in summer and their surface anomalies in general circulation models (GCMs) has not been systematically analyzed. Here we validate the representation of wavenumbers 1–10 in three state-of-the-art global climate models (EC-Earth, CESM, and MIROC), quantify their biases, and provide insights into the underlying physical reasons for the biases. To do so, the ExtremeX experiments output data were used, consisting of (1) historic simulations with a freely running atmosphere with prescribed ocean and experiments that additionally (2) nudge towards the observed upper-level horizontal winds, (3) prescribe soil moisture conditions, or (4) do both. The experiments are used to trace the sources of the model biases to either the large-scale atmospheric circulation or surface feedback processes. Focusing on wave 5 and wave 7, we show that while the wave's position and magnitude are generally well represented during high-amplitude (> 1.5 SD) episodes, the associated surface anomalies are substantially underestimated. Near-surface temperature, precipitation and mean sea level pressure are typically underestimated by a factor of 1.5 in terms of normalized standard deviations. The correlations and normalized standard deviations for surface anomalies do not improve if the soil moisture is prescribed. However, the surface biases are almost entirely removed when the upper-level atmospheric circulation is nudged. When both prescribing soil moisture and nudging the upper-level atmosphere, then the surface biases remain quite similar to the experiment with a nudged atmosphere only. We conclude that the near-surface biases in temperature and precipitation are in the first place related to biases in the upper-level circulation. Thus, relatively small biases in the models' representation of the upper-level waves can strongly affect associated temperature and precipitation anomalies.