Ensembles of leading European global coupled climate models show impressive reliability for seasonal climate prediction-including useful output for probabilistic prediction of malaria incidence and crop yield.

Advances in simulating atmospheric variability with the ECMWF model are presented that stem from revisions of the convection and diffusion parametrizations. The revisions concern in particular the introduction of a variable convective adjustment time-scale, a convective entrainment rate proportional to the environmental relative humidity, as well as free tropospheric diffusion coefficients for heat and momentum based on Monin-Obukhov functional dependencies.The forecasting system is evaluated against analyses and observations using high-resolution medium-range deterministic and ensemble forecasts, monthly and seasonal integrations, and decadal integrations with coupled atmosphere-ocean models. The results show a significantly higher and more realistic level of model activity in terms of the amplitude of tropical and extratropical mesoscale, synoptic and planetary perturbations. Importantly, with the higher variability and reduced bias not only the probabilistic scores are improved, but also the midlatitude deterministic scores in the short and medium ranges. Furthermore, for the first time the model is able to represent a realistic spectrum of convectively coupled equatorial Kelvin and Rossby waves, and maintains a realistic amplitude of the Madden-Julian oscillation (MJO) during monthly forecasts. However, the propagation speed of the MJO is slower than observed. The higher tropical tropospheric wave activity also results in better stratospheric temperatures and winds through the deposition of momentum.The partitioning between convective and resolved precipitation is unaffected by the model changes with roughly 62% of the total global precipitation being of the convective type. Finally, the changes in convection and diffusion parametrizations resulted in a larger spread of the ensemble forecasts, which allowed the amplitude of the initial perturbations in the ensemble prediction system to decrease by 30%.

International audienc

[1] The scientific basis for two-tier climate prediction lies in the predictability determined by the ocean and land surface conditions. Here we show that the state-of-the-art atmospheric general circulation models (AGCMs), when forced by observed sea surface temperature (SST), are unable to simulate properly Asian-Pacific summer monsoon rainfall. All models yield positive SST-rainfall correlations in the summer monsoon that are at odds with observations. The observed lag correlations between SST and rainfall suggest that treating monsoon as a slave possibly results in the models' failure. We demonstrate that an AGCM, coupled with an ocean model, simulates realistic SSTrainfall relationships; however, the same AGCM fails when forced by the same SSTs that are generated in its coupled run, suggesting that the coupled ocean-atmosphere processes are crucial in the monsoon regions where atmospheric feedback on SST is critical. The present finding calls for reshaping of current strategies for monsoon seasonal prediction. The notion that climate can be modeled and predicted by prescribing the lower boundary conditions is inadequate for validating models and predicting summer monsoon rainfall.

The DEMETER multi‐model ensemble system is used to investigate the rationale behind the multi‐model concept. A comprehensive documentation of the differences in the single and multi‐model performance in the DEMETER hindcast data set is given. Both deterministic and probabilistic diagnostics are used and a variety of analyses demonstrate the improvements achieved by using multi‐model instead of single‐model ensembles. In order to understand the reason behind the multi‐model superiority, basic scenarios describing how the multi‐model approach can improve over single‐model skill are discussed. It is demonstrated that multi‐model superiority is caused not only by error compensation but in particular by its greater consistency and reliability.