The role of isotope-enabled GCM complexity in simulating tropical circulation changes in high-CO2 scenarios

Stable water isotopes are data-rich tracers of the hydrological cycle, and, recently, the advent of isotope-enabled climate models has allowed for investigations into the utility of water isotopes for tracking changes in the large-scale atmospheric circulation. Among the suite of published isotope-enabled climate models, those with intermediate complexity offer the benefits of efficiency, allowing for long ensemble runs. However, the ability of these models to simulate the response to global warming with the same fidelity as state-of-the-art models is questionable. Here we evaluate an intermediate complexity model, SPEEDY-IER, in a high-CO(2)scenario and compare its performance to an Intergovernmental Panel on Climate Change (IPCC)-class model, iCAM5. SPEEDY-IER can generally simulate changes in tropical circulation, the weakening of the Walker circulation, and the narrowing of the deep tropics. A deeper investigation of water isotope fields indicates SPEEDY-IER simulates qualitative trends in precipitation and vapor isotopes with fidelity, but it does not simulate amplitudes or spatial patterns of water isotope changes shown in iCAM5. This bias in SPEEDY-IER is mainly due to its coarse resolution and simplified convection scheme. We then modify the model by introducing condensation and detrainment in intermediate convection levels; this modification successfully improves SPEEDY-IER's simulation of water isotope fields, though the response of the Walker circulation to climate change is weakened. We demonstrate that evaluating water isotope fields reveals hidden biases in a climate model and guides improvements to the model's performance. Thus, the examination of water isotope fields and validation against available observations likely provides more stringent constraints for model physics.