Examining the changes of extratropical cyclone predictability with high resolution CMIP6-based ensemble forecasts

Sándor Maho

Little is known about how forecast uncertainties related to extratropical cyclones (ETC) change in a warmer climate. By assuming that predictability is inversively proportional to the ensemble spread, we can study how the predictability of various ETC metrics changes in ensemble simulations initialised with CMIP6 data and run with OpenIFS, an open source version of the Integrated Forecasting System developed by ECMWF. We select cases of high-impact ETCs from historical and ssp5 EC-Earth3 simulations, which we simulate with OpenIFS using the SPPT scheme to produce the 51 ensemble members. The downscaling of the original CMIP6 data and the initialisation of OpenIFS is done by a recently developed software, CLIO. Test experiments show that CLIO-OpenIFS simulations are able to represent the CMIP6 meteorological environments with reasonable accuracy. Preliminary results show that the ensemble spread decreases for ssp5 cyclones for various dry ETC metrics (i.e. central pressure, maximum wind, maximum lower-level vorticity) indicating an improved predictability. In contrast, the spread of 6-hourly accumulated ETC precipitation increases, implying decreased predictability for ssp5 cyclones.