Understanding tropical precipitation biases in kilometer-scale global climate models using the atmospheric energy balance framework

Here we investigate tropical precipitation biases in the novel kilometer-scale Earth system models (ICON and IFS) developed by the EU-funded H2020 nextGEMS project. Despite the much higher resolution, these km-scale models still feature biases that are common to CMIP models: first, tropical precipitation is systematically overestimated. Second, the double ITCZ (Intertropical Convergence Zone) bias is not ameliorated, with too little rain falling close to the Equator and too much rain in the southern branch relative to the northern branch. The double ITCZ bias is consistent with Hadley circulations that feature secondary cells close to the equator. Third, both the northern and the southern ITCZ branches are displaced poleward relative to observations. 

Focusing on the tropical precipitation distribution, we more explicitly quantify existing biases through a symmetric and an antisymmetric precipitation index. Leveraging the well-established atmospheric energy balance framework, we show how hemispherically symmetric biases are positively corellated with biases in the equatorial net energy input (NEI), once any residual in its global average is removed. In both models, equatorial NEI biases primarily arise from surface latent heat fluxes. Hemispherically antisymmetric biases are instead negatively correlated with the cross-equatorial atmospheric energy transport, which is in turn linked to biases in the NEI hemispheric asymmetry. The leading sources of asymmetric biases are top-of-atmosphere radiative fluxes in IFS and surface radiative fluxes in ICON.

Finally, although we find that notorious GCM precipitation biases are not mitigated when employing km-scale grids, we also see that the atmospheric energy balance holds great potential for improving tropical precipitation patterns. In this regard key candidates for improving the energy balance are surface flux schemes, particularly for latent heat over the oceans, and cloud-radiative effects.