Drivers of the asymmetric ITCZ bias in climate models

Abstract Realistic representation of the Intertropical Convergence Zone (ITCZ) remains a persistent challenge in coupled climate models, and the physical drivers of model biases are not fully understood. A particular issue is the relative roles of sea surface temperature (SST) and atmospheric thermodynamic structure in the formation and seasonal migration of the ITCZ. Here, we revisit this issue by analyzing the different behaviors of the ITCZ in 9 climate models. We apply an empirical orthogonal function (EOF)-based diagnostic method and characterize the difference in ITCZ among these models into different modes of hemispheric precipitation asymmetry. The leading mode explains the majority of inter-model variance and defines an asymmetric ITCZ index that captures the meridional structure of tropical precipitation biases. This index is strongly associated with the vertical structure of virtual temperature and specific humidity, whereas its response to SST influence is substantially weaker, indicating that just a more realistic representation of SST in a model does not guarantee a reduced ITCZ bias in the model. Using a single-column model framework, we further demonstrate that modifications to vertical moisture profiles produce a more coherent and robust precipitation response than SST perturbations, indicating that atmospheric moisture exerts a dominant control on ITCZ asymmetry. The two analyses yield consistent results. They suggest that reducing ITCZ biases in climate models requires realistic SSTs as well as an accurate representation of the atmospheric thermodynamic structure over tropical oceans.