Abstract
AbstractRecent studies have found evidence of multidecadal variability in Northern Hemisphere wintertime seasonal forecast skill. Here we assess the robustness of this finding by extending the analysis to a diverse set of ensemble atmospheric model simulations. These simulations differ in either the numerical model or type of initialisation and include atmospheric model experiments initialised with reanalysis data and free‐running atmospheric model ensembles. All ensembles are forced with observed sea‐surface temperatures (SSTs) and sea‐ice boundary conditions. Analysis of large‐scale Northern Hemisphere circulation indices over the Northern Hemisphere (namely the North Atlantic Oscillation, the Pacific/North American pattern and the Arctic Oscillation) reveals that in all ensembles there is larger correlation skill in late‐century periods than in mid‐century periods. Similar multidecadal variability in skill is found in a measure of total skill integrated over the whole extratropical region. Most of the differences in large‐scale circulation skill between the skilful late period (as well as the early period) and the less skilful mid‐century period seem to be due to a reduction in skill over the North Pacific and a disappearance of skill over North America and the North Atlantic. The results are robust across different models and different types of initialisation, indicating that the multidecadal variability in Northern Hemisphere winter skill is a robust feature of twentieth‐century climate variability. Multidecadal variability in skill therefore arises from the evolution of the observed SSTs, likely related to a weakened influence of the El Niño–Southern Oscillation on the predictable extratropical circulation signal during the middle of the twentieth century, and is evident in the signal‐to‐noise ratio of the different ensembles, particularly the larger ensembles.
Highlights
We begin by assessing the correlation between the observed North Atlantic Oscillation (NAO) index and the ensemble mean NAO indices in the different datasets
In this article we have examined how the multidecadal variability in seasonal forecast skill of Northern Hemisphere wintertime circulation depends on the type of initialisation and atmospheric model
We analysed a set of ensemble atmospheric model simulations, including atmospheric model simulations initialised with reanalysis data and free-running atmospheric model simulations constrained by observed ocean conditions
Summary
Four other ensemble datasets are analysed, all of which are free-running atmospheric model simulations with prescribed SSTs (i.e., “AMIP-style” simulations) In these datasets, there is no information from any initial conditions for a given season, and any decadal variability in skill is not due to variation in initialisation quality. The two ensemble datasets were taken from the Coupled Model Intercomparison Project 6 (CMIP6) database, the “AMIP-HIST” experiments from the Global Monsoons Model Inter-comparison Project (GMMIP; Zhou et al, 2016) These experiments include the natural and anthropogenic (external) historical forcings used in CMIP6 historical simulations, with prescribed SSTs and sea-ice boundary conditions from HadISST (Rayner et al, 2003). We found that this made little difference to the results because the magnitude of the trends was very small compared to the interannual variability
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call