Ensemble prediction of subseasonal atmospheric anomalies

In the early nineties, under Eugenia Kalnay's leadership NCEP made important advances in numerical modeling to produce monthly forecasts by performing integrations of an atmospheric model with prescribed SSTs. At the same time, she founded a group to address the inherently stochastic nature of the atmosphere through ensemble forecast methods. Both lines of research and development have grown over the years resulting in the creation of sophisticated global numerical prediction systems: one, the CFSv2, which couples the atmospheric GFS to an ocean and ice model and another, the GEFS, which models the PDF forecast by representing initial and model related uncertainties. Both systems constantly improve through the use of better data assimilation schemes, higher model resolutions and more sophisticated physics. The GEFS updates much more often than the coupled system. A step towards extending the lead time of the GEFS to subseasonal time scales requires both approaches: coupling with SST and ensemble methods. A series of GEFS integrations in experimental mode have been carried out to determine a suitable configuration that could be run within the constraints of the available operational computer resources. The experimental extended runs though limited show first of all that current uncouple NWPs are able to capture the response of extended forcing in the equatorial tropics (such as during active MJO) modulating the skill of extratropical circulation to higher skill when the signal is strong and lower skill when the signal is weak. Experiments of extended GEFS with prescribed realistic SSTs indicate reduce bias in predicted tropical variables even at short lead times. In this talk, we will present a series of diagnostics of the comparison, including impact on lifespan of subseasonal atmospheric anomalies and atmospheric large-scale waves. An intuitive and practical tool introduced by Eugenia and colleagues to identify the forcing direction in locally coincident SST and atmospheric anomalies is been applied to detect geographic regions where prescribed SSTs cause the most impact to skill.