What potential for improving sub‐seasonal predictions of the winter <scp>NAO</scp>?

The North Atlantic Oscillation (NAO) is the leading mode of variability across sector and a key metric extratropical forecast performance. Skilful predictions NAO are possible at medium-range (1–2 weeks) seasonal time scales. However, in dynamical prediction system, we find that sub-seasonal (1 month with lead 20–30 days) not statistically significant represent gap skill. In this study, have investigated potential for improving using large ensemble hindcasts. First, monthly only weakly related to errors medium-range. This implies performance unlikely drive improvements longer times. Second, Madden-Julian (MJO) Tropics projects onto lag 10–15 days, but its teleconnection partially represented current systems. We, therefore, assess whether improved MJO-NAO teleconnections likely predictions. We even perfect MJO forecasts small skills. work indicates timescales may predictability hence Euro-Atlantic winter climate which genuine skill difficult achieve. Potential progress area could stem from currently unknown sources initialised ensembles will be vital tool investigating these.