Predictability of the NAO Adam Scaife Crown copyright

Predictability of the NAO? Adam Scaife © Crown copyright Met Office

Winters depend on which way the wind blows: The North Atlantic Oscillation Winter 1962/63 Weak P Gradient Winter 2009/10 Cold advection into Europe Cold, calm and dry Winter 2011/12 c. f. 2013/14! Strong P Gradient Warm advection into Europe Mild, stormy and wet © Crown copyright Met Office

Glo. Sea 5 Global Seasonal Forecast System 5 Model: Had. GEM 3 H N 216 L 85 O(0. 25) Initialisation: NWP state + NEMOVAR + Sea Ice Winter Hindcasts: 24 members starting around 1 November 20 years: 1992/3 – 2012/13 Now Operational: 2 members per day -> 2 months 2 members per day -> 7 months © Crown copyright Met Office

Operational forecasts: January 2013 Observations Signals appearing in forecasts from 21 st Dec Warning provided to Dept for Transport 4 th Jan th onwards Prolonged © Crown copyrightsnow Met Officefrom ~Jan 20 Forecast

Individual winters from 1992/3 to 2011/12 Pattern of ensemble mean similar in many cases but not all © Crown copyright Met Office

Winter NAO hindcasts Correlation score = 0. 62 Extended to 20 years and 24 members for DJF © Crown copyright Met Office

Seasonal Skill for surface climate Forecast skill from NAO alone Storms Temp Wind Skilful predictions of extreme events months ahead Skill in E Atlantic and Europe is from the NAO © Crown copyright Met Office Scaife et al 2014

Sources of predictability… Strongest minus weakest cases for November predictors: ENSO, Atlantic Ocean, Kara sea-ice and Quasi Biennial Oscillation Response is weaker in model than obs © Crown copyright Met Office

Large ensembles are needed 24 member ensemble shows strong sensitivity to ensemble size Nothing special about 0. 62 and higher values are possible! © Crown copyright Met Office

Daily extremes and impacts Prediction of the NAO Þ Prediction of winter extremes Þ Prediction of Impacts Cold days (energy, transport. . ) Storms (insurance…) Windspeed (renewables…) Hydrology (river flows…) © Crown copyright Met Office Svensson et al, 2014

Summary: Skilful seasonal forecasts for the NAO, Europe & N America Low signal to noise – this is a paradox and a key research topic Risk of daily extremes governed by large scale flow Predictability of useful quantities like storminess or windspeed Further Development: Much work to do with partners on predicting impacts… © Crown copyright Met Office