Abstract ID: 087
Initial Conditions for Stratospheric Error Growth
Lead Author: Robert W. Lee
University of Reading, United Kingdom
Keywords: Stratosphere-troposphere coupling, Bayesian, Minimal model, Sudden Stratospheric Warmings, S2S
Abstract: The stratosphere has been shown to be a significant source of sub-seasonal tropospheric predictability for some winters, however it is not yet well understood what mechanisms define when this will come into play. We investigate the predictability associated with stratosphere-troposphere coupling in the sub-seasonal and seasonal hindcast datasets by fitting a simple, minimal model. We diagnose the contribution of the stratosphere to tropospheric forecast skill by using Bayesian methods to fit the minimal model to hindcasts in the S2S database in order to compare and contrast different prediction systems and also compare them with synthetic forecast datasets. We also contrast correlation skill in sub-sets of forecasts with weak, neutral and strong lower stratospheric polar vortex states. We also utilise windows-of-opportunity and compare with a broader range of skill metrics to test their performance in detecting model skill derived from stratosphere-troposphere coupling and the different impression they impart. Finally, we apply the minimal model to the SSW case study events of boreal winters 2018 versus 2019 to investigate further the contrasting tropospheric responses.
Andrew J. Charlton-Perez (University of Reading)