Modes of tropical climate variability such as the El Niño-Southern Oscillation, Indian Ocean Dipole, tropical Atlantic variability, Madden Julian Oscillation, monsoons, and interactions among tropical basins, influence hydroclimate, extremes, ecosystems, and societies far beyond the tropics. Despite sustained development, climate models continue to exhibit biases in tropical mean states, precipitation, convection, air-sea coupling, circulation, variability, and teleconnections. These errors distort predictability and contribute to uncertainty in projections of climate modes and their impacts.
This session will connect three parts of the modelling challenge: identifying the origins and consequences of persistent biases; evaluating improvements in resolution, parameterizations, coupling, initialization, and data-driven or hybrid methods; and determining how better present-day simulations can yield more credible predictions and future projections. We welcome work using observations and reanalyses, hierarchy-of-model approaches, coupled model intercomparisons, large ensembles, initialized predictions, perturbed-physics experiments, emergent constraints, machine learning, and novel diagnostics.
Contributions may address tropical atmosphere-ocean-land interactions across weather, subseasonal, seasonal, interannual, decadal, and climate-change timescales. Studies that link process understanding to model development, compare generations of models, quantify uncertainty, or assess the global and regional impacts of tropical simulation errors are especially encouraged. By bringing together model developers, observational researchers, and prediction and projection communities, the session aims to identify advances, unresolved barriers, and practical priorities for the next generation of climate simulations.