Inter-model comparison of tropical intraseasonal variability and convectively coupled wave activity in aquaplanet experiments with an idealized warm pool

We carry out an inter-model comparison of tropical intraseasonal variability (ISV) and convectively coupled wave (CCEW) activity in the European ESMs participating in the EMBRACE project (CNRM, HadGEM, MPI, IPSLA, IPSLB, EC-Earth). The EMBRACE project (Earth system Model Bias Reduction and assessing Abrupt ClimatE change) is funded by the 7th Framework Program of the European Union; it aims at evaluating and improving the European Earth System Models (ESMs).

We focus on the question of the influence of the background state on the ability of the models to produce the different types of CCEW activity and eastward-propagating intraseasonal variability. We use the simplified framework of aquaplanet experiments as a way to better understand the behavior of each model in its full coupled/forced configuration.

The simulated CCEW activity and intraseasonal activity is analyzed in the aquaplanet ("AquaControl") simulations produced for the CMIP5 project, and compared to the coupled (CMIP) and forced (AMIP) simulations. Space-time spectral analysis of the tropical precipitation and 850 hPa zonal wind is used to diagnose the zonally propagating variability.

It appears that for each model, the CCEW spectral patterns and their main biases (compared to the observed spectra) are relatively consistent in the 3 configurations (forced, coupled, aquapla net). Larger discrepancies arise between the different models, for instance in the propagation speed of the waves. This emphasizes the strong influence of each model's parametrization scheme for tropical deep convection on the charachteristics of the simulated waves and on the degree to which convection actually couples with the dry equatorial waves.

This consistency between configurations does not apply to the eastward-propagating intraseasonal variability. For example, none of the models simulate any eastward ISV in the aquaplanet ("aquaControl") configuration, not even those models that show a strong boreal-winter MJO in their full coupled configuration (such as the CNRM model).

We carry out some additional aquaplanet experiments to investigate further the sensitivity of the CCEW and eastward ISV to the background mean state with three of the EMBRACE models: CNRM, HadGEM3 and EC-Earth. The experiments follow the protocole suggested in the earlier APE project (e.g. constant equinoxial solar radiative forcing) but building on their results, we use here a controle SST profile with a broader shape in latitude ("QOBS") than what they used. We mainly focus on experiments with an idealized warm pool added on top of the axisymmetric QOBS profile. Preliminary results show that at least one of the 3 models (CNRM) appears quite sensitive to the presence of a warm pool and seems to recover some eastward propagating intraseasonal variability in that setting. In this model, the westerly winds generated in the low levels to the west of the SST warm pool anomaly are much stronger than in the two other models. The differences between models and configuration is analyzed using Moist Static Energy (MSE) budgets of the composite intraseasonal anomalies.