ALBERT

Description: The intertropical convergence zone (ITCZ) is an important component of the tropical rain belt. Climate models continue to struggle to adequately represent the ITCZ and differ substantially in its simulated response to climate change. Here we employ complex network approaches, which extract spatio-temporal variability patterns from climate data, to better understand differences in the dynamics of the ITCZ in state-of-the-art global circulation models (GCMs). For this purpose, we 5 study simulations with 14 GCMs in an idealized slab-ocean aquaplanet setup from TRACMIP – the Tropical Rain belts with an Annual cycle and a Continent Model Intercomparison Project. We construct network representations based on the spatial correlation pattern of monthly surface temperature anomalies and study the zonal mean patterns of different topological and spatial network characteristics. Specifically, we cluster the GCMs by means of their zonal network measure distribution utilizing hierarchical clustering. We find that in the control simulation, the zonal network measure distribution is able to pick 10 up model differences in the tropical SST contrast, the ITCZ position and the strength of the Southern Hemisphere Hadley cell. Although we do not find evidence for consistent modifications in the network structure tracing the response of the ITCZ to global warming in the considered model ensemble, our analysis demonstrates that coherent variations of the global SST field are linked with ITCZ dynamics. This suggests that climate networks can provide a new perspective on ITCZ dynamics and model differences therein.