*Nathaniel C Johnson1, Kai-Chih Tseng1,3, Thomas L Delworth1, Liwei Jia1,2, Feiyu Lu1,3, William Cooke1, Colleen McHugh1,4, Andrew T Wittenberg1
(1.NOAA Geophysical Fluid Dynamics Laboratory, 2.University Corporation for Atmospheric Research, 3.Department of Geosciences, Program in Atmospheric and Oceanic Science, Princeton University, 4.SAIC, Science Applications International Corporation)
Keywords:extreme weather prediction, seasonal prediction, heat waves, atmospheric rivers, severe thunderstorm prediction
In this presentation, I describe the latest research advancing the seasonal prediction of extreme weather rooted in the development of the new seasonal-to-decadal prediction system at the Geophysical Fluid Dynamics Laboratory (GFDL), the Seamless System for EArth System Research (SPEAR). The SPEAR models have a coarse ocean resolution of approximately 1.0° that allows computational efficiency for the running of large ensembles and atmospheric resolution ranging from 1.0°to 0.25° that allows the simulation of various extreme weather phenomena. SPEAR provides both centennial-timescale climate change projections and routine real-time seasonal forecasts for NOAA as a contributor to the North American Multi-Model Ensemble (NMME). The SPEAR system has demonstrated the potential to produce skillful seasonal forecasts of the statistics of various extreme weather phenomena, including atmospheric river activity over western North America, boreal spring U.S. tornado occurrence, and summer heat waves over the U.S. The combination of the initialized SPEAR forecast simulations and large ensembles with historical and projected radiative forcing provides an opportunity for rapid attribution of extreme weather statistics in seasonal forecasts. I conclude with a brief discussion of some ongoing SPEAR model developments, including the development of a flux-adjusted version and a version with higher ocean resolution (approximately 25 km).