Scholarly article on topic 'Seasonal to decadal prediction of the winter North Atlantic Oscillation: emerging capability and future prospects'

Seasonal to decadal prediction of the winter North Atlantic Oscillation: emerging capability and future prospects Academic research paper on "Earth and related environmental sciences"

Share paper

Academic research paper on topic "Seasonal to decadal prediction of the winter North Atlantic Oscillation: emerging capability and future prospects"

Quarterly Journal of the Royal Meteorological Society

Q. J. R. Meteorol. Soc. 142: 611-617, January 2016 B DQI:10.1002/qj.2479


Royal Meteorological Society

Seasonal to decadal prediction of the winter North Atlantic Oscillation: emerging capability and future prospects

Doug M. Smith,* Adam A. Scaife, Rosie Eade and Jeff R. Knight

Met Office Hadley Centre, Exeter, UK

Correspondence to: D. M. Smith, Met Office Hadley Centre, FitzRoy Road, Exeter, EX1 3PB, UK.

E-mail: This article is published with the permission of the Controller of HMSO and the Queen's Printer for Scotland.

European and North American winter weather is dominated by year-to-year variations in the North Atlantic Oscillation (NAO) which controls the direction and speed of the prevailing winds. An ability to forecast the time-averaged NAO months to years ahead would be of great societal benefit, but current operational seasonal forecasts show little skill. However, there are several elements of the climate system that potentially influence the NAO and may therefore provide predictability for the NAO. We review these potential sources of skill, present emerging evidence that the NAO may be usefully predictable (with correlations exceeding 0.6) on seasonal time-scales, and discuss prospects for improving skill and extending predictions to multi-year time-scales.

Key Words: seasonal prediction; decadal prediction; North Atlantic Oscillation; Arctic Oscillation

Received 28 February 2014; Revised 15 September 2014; Accepted 1 October 2014; Published online in Wiley Online Library 27 November 2014

1. Introduction

Systematic climate change from anthropogenic forcing is generally too small to account for much of the variability in regional climate on time-scales out to a decade ahead, especially in midlatitudes during winter. Instead, variability on these time-scales is mainly due to fluctuations in atmospheric circulation. Whereas El Nino/Southern Oscillation (ENSO) explains the largest proportion of tropical variability, the winter-mean (December to February) Arctic Oscillation (AO: Thompson and Wallace, 1998; also referred to as the Northern Annular Mode, NAM) or its more regional equivalent, the North Atlantic Oscillation (NAO: Walker and Bliss, 1932), are responsible for a large proportion of extratropical variability in surface climate (e.g. Hurrell, 1995). Examples of positive and negative NAO winters are shown in Figure 1. The winter of 2009/2010 is chosen as a recent negative case. This winter had the lowest recorded value of the winter NAO in over a century of sea-level pressure records (Fereday et al., 2012). The anomaly in the NAO in winter 2009/2010 exceeded 20 hPa (Figure 1(a)) and was so large that the winter mean meridional gradient of sea-level pressure was of opposite sign to the usual climatological gradient, corresponding to winter mean easterly flow across the Atlantic. This flow in turn resulted in anomalous heat transport around the Atlantic basin. Cold advection into large areas of northern Europe and southeastern parts of North America resulted in lower than average temperatures, while corresponding warm advection into eastern Canada and southern Europe increased temperatures well

above the climatological winter mean (Figure 1(b)). The winter of 1999/2000 is a prominent example of the positive phase of the NAO. In this case, a stronger than normal meridional gradient of sea-level pressure and a northward shifted, stronger westerly jet across the Atlantic delivered warm moist air to northern Europe and the eastern United States and a roughly opposite quadrupole pattern of surface temperature anomalies compared to negative NAO events.

Fluctuations in the winter-mean NAO also govern changes in the frequency of extreme daily events over large and heavily populated regions of the Northern Hemisphere (Thompson et al., 2002; Scaife et al., 2008; Kenyon and Hegerl, 2010). Many different types of extremes are related by the NAO and associated changes in the extratropical jet stream. For example, in northern Europe positive NAO conditions greatly reduce the incidence of extreme low temperatures but increase the risk of heavy rainfall and severe storms. Indeed, the positive NAO winter shown in Figure 1 contained several devastating European storms including Anatol which caused major damage in Denmark, and Lothar closely followed by Martin (Roberts et al., 2014), causing intense damage due to high winds across France, Germany, Switzerland and Italy in December 1999. A similar, albeit slightly southward shifted intensification of the Atlantic pressure gradient occurred in 2013/2014, resulting in an extremely stormy winter and record winter rainfall in the United Kingdom.

A null hypothesis for seasonal and multi-year NAO variability is that it is the residual of day-to-day internal atmospheric weather. In this case, low-frequency variations would not be predictable

© 2014 Crown copyright. Quarterly Journal of the Royal Meteorological Society published by John Wiley & Sons Ltd on behalf of the Royal Meteorological Society. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

Figure 1. Influence of the NAO on European and North American winters. Shown are maps of sea-level pressure (a,c: from HadSLP2, Allan and Ansell, 2006) and temperature (b,d: from HadCRUT4, Morice et al., 2012) anomalies (relative to 1961-1990) for the winters (December to February) 2009/2010 (a,b: negative NAO) and 1999/2000 (c,d: positive NAO). Arrows on (a) and (c) show the anomalous wind direction.

because individual weather events cannot be predicted more than a week or two ahead due to the chaotic nature of the atmosphere. However, statistical analysis (e.g. Stephenson etal., 2000) and physical modelling (discussed below) suggest that this is not the case. Given that the NAO is key for understanding many near-term climate fluctuations and identifying periods of increased risk of imminent extreme weather events, understanding other possible causes of NAO variability, beyond this null hypothesis, is of high priority. In this brief'position article' we highlight some drivers of seasonal and multi-year variability in the NAO that have already been identified and that are related to slowly varying elements of the climate system and may therefore provide predictability for the NAO. We then present emerging evidence that the NAO is predictable on seasonal time-scales, and discuss prospects for extending skilful predictions to multi-year time-scales.

2. Potential sources of skill

Many previous studies have identified surface climate features that project onto the NAO and that appear to be driven by internal phenomena or external conditions that are potentially predictable on seasonal and longer time-scales (e.g. see overview in Christensen etal. (2013)). We discuss only those potential drivers of the NAO in this summary that have stood the test of time by continuing to be present in updated historical observational analyses and also show at least some reproduction in numerical models based on first principles.

Numerous studies have investigated the possible role of Atlantic sea-surface temperature (SST) changes in driving changes in the jet stream and associated surface pressure patterns resembling the NAO. Early studies (Ratcliffe and Murray, 1970) pointed to the region off Newfoundland, and some subsequent experiments produced systematic effects (Peng etal., 1997). Similarly, the studies by Rodwell etal. (1999) and Rodwell and Folland (2002) identified a potential response in the NAO associated with North Atlantic SST patterns (Figure 2(a)). However, there are issues around specified SST experiments (Bretherton and Battisti, 2000) given the large influence of atmospheric circulation on the ocean (Frankignoul and Hasselmann, 1977) and it is therefore difficult to interpret what such experiments imply for actual predictability. Nevertheless, experiments using coupled models with large SST differences in the North Atlantic show a similar influence of the ocean on the atmosphere to that inferred by Rodwell et al. (1999) and atmosphere-only simulations suggest there is a genuine influence from the ocean state onto the atmospheric jet stream and associated winter blocking frequency (Scaife et al., 2011).

While ENSO is an established cornerstone of tropical seasonal forecasting and its connection to the Pacific North American (PNA) pattern is well established, the links between ENSO and the

NAO have only recently been confirmed. Observational studies suggest a systematic tendency for negative NAO during El Nino conditions in boreal winter and the opposite during La Nina (e.g. Moron and Gouirand, 2003; Toniazzo and Scaife, 2006; Bronnimann etal., 2007). Figure 2(b) shows a similar composite result from sea-level pressure reanalyses with a negative NAO signal during El Nino. Recent modelling studies confirm this link (Manzini etal., 2006; Bell etal., 2009; Cagnazzo and Manzini, 2009; Ineson and Scaife, 2009) and suggest that interference ofthe extratropical PNA pattern with climatological stationary waves is key to providing enhanced wave driving of the stratosphere and subsequent negative-NAO-like patterns at the surface in late winter (Ineson and Scaife, 2009; Smith etal., 2010b).

A slightly more controversial area is the role of solar variability in surface NAO variations. Many statistical studies link cold northern European winters to low solar activity and mild winters to high solar activity (e.g. Lockwood etal., 2010, cf. Figure 2(c)) but the signals vary over the historical record. Lagged responses are more consistent (Gray etal., 2013) and appear to be due to ocean-atmosphere interaction (Scaife etal., 2013) which feeds back positively onto the initial atmospheric response. The fact that similar NAO responses are now being reproduced in climate model experiments (Matthes etal., 2006; Ineson etal., 2011) adds weight to the observational evidence for negative NAO during low solar activity periods.

A further external influence on the NAO on seasonal to decadal time-scales comes from explosive volcanic eruptions. There is evidence in the observational record that for two or so winters following eruptions which inject significant sulphate aerosol into the lower stratosphere, there is a positive NAO signature in extratropical surface climate (Figure 2(d)). However, only weak albeit similar signatures are reproduced in climate models (Stenchikov et al., 2004; Marshall et al., 2009), even in the latest Coupled Model Intercomparison Project phase 5 (CMIP5) climate models (Driscoll etal., 2012).

The sources of NAO variability considered above have included factors connected to the oceans and external climate forcing. It is generally the case that processes purely internal to the atmosphere have too short a time-scale to contribute significantly to seasonal to decadal prediction. However, there is one notable exception in the Quasi-Biennial Oscillation (QBO) which has a long-standing teleconnection to the winter surface climate which projects onto the NAO (Figure 2(e), Ebdon, 1975; Anstey and Shepherd, 2014). The QBO is highly predictable out to years ahead (Scaife etal., 2014b), and observed connections to the NAO are reproduced in some climate model experiments (Niwano and Takahashi, 1998; Marshall and Scaife, 2009). However, climate model forecasts in general have difficulty in reproducing the observed teleconnection (Scaife etal., 2014b).

Figure 2. Potential drivers of the NAO. Observed composites of December to February (DJF) mean-sea-level pressure associated with (a) Atlantic tripole SST pattern (using May tripole patterns updated from Rodwell and Folland (2002)), (b) ENSO, (c) the 11-year solar cycle, (d) major volcanic eruptions, (e) the quasi-biennial oscillation (QBO), (f) Atlantic multi-decadal variability (AMV), and (g) Pacific decadal variability (PDV). Units are standard deviations of annual (a-e) and decadal (f,g) time-scales. Adapted from Smith et al. (2012) where full details of indices and methodology are provided.

Observations suggest a link between Eurasian snow cover, particularly in October, and the following winter AO (Cohen and Entekhabi, 1999). There is also a potential physical mechanism whereby anomalously high autumn Eurasian snow cover increases the upward propagating stationary Rossby wave activity, which slows the polar stratospheric vortex. This weakening of the polar vortex then propagates downwards into the troposphere resulting in a negative-AO-like response. In general, free-running models do not reproduce these observed links (Hardiman etal., 2008), but this discrepancy could be caused by biases in their simulation of snow cover or stratospheric winds (Allen and Zender, 2011; Peings et al., 2012).

There is some evidence that ocean variability potentially influences the NAO on decadal time-scales but there is not yet strong evidence that this is reproducible in climate models. Atlantic multidecadal variability shows a promising potential link to the NAO with a warm Atlantic associated with higher pressure near Iceland which projects onto a negative NAO (Figure 2(f)). There is also some evidence of similar signals in models (e.g. Knight etal., 2006; Gastineau etal., 2013; Omrani etal., 2014), although again care is needed in the interpretation of these signals given the large atmospheric forcing influence

on the ocean. Pacific decadal variability, with a broad, low-frequency pattern of variability similar to ENSO, shows a possible link to the NAO in observations (Figure 2(g)) but this is less well established than other signals and it does not agree with the shorter time-scale variability from ENSO (Figure 2(b)), perhaps casting doubt on this as a strong driver of the NAO.

3. Model forecasts: current capability

Despite occasional signs of skill in long-range predictions of the Arctic Oscillation in the free atmosphere (Muller etal., 2005), skilful predictions of year-to-year fluctuations in the winter-mean surface NAO and associated surface climate at lead times of months have been elusive (Johansson, 2007), even in current operational systems (Doblas-Reyes etal., 2003; Arribas etal., 2011; Kim etal., 2012).

The latest Met Office seasonal forecasting system, Global Seasonal forecast system version 5 (GloSea5: MacLachlan etal., 2014), has been specifically designed to capture the potential sources of predictability of the NAO described above. Key features include:

-0.9 -0.6 -0.3 0 0.3 0.6 0.9

Figure 3. Seasonal forecast skill for DJF sea-level pressure. Anomaly correlation between observed and GloSea5 DJF sea-level pressure, computed from hindcasts starting around 1 November each year from 1993 to 2012 (Scaife etal., 2014a). Stippling shows correlations different to zero at the 5% level of significance (based on bootstrap resampling).

• A well resolved stratosphere, with a total of 85 vertical levels in the atmosphere and a lid at 80 km. This is important for capturing the downwards propagation of stratospheric wind perturbations into the troposphere (Baldwin and Dunkerton, 2001), as occurs for example in teleconnections between El Nino and the NAO (Ineson and Scaife, 2009) and the NAO response to changes in solar radiation (Ineson etal., 2011).

• A high horizontal resolution of 0.833 x 0.556° in the atmosphere and 0.25° in the ocean. The high ocean resolution in particular reduces the cold bias in North Atlantic SSTs which is typically found in coarser models which are unable to simulate the northwards deflection of the Gulf Stream. Reducing this SST bias improves the simulation of Atlantic blocking frequency (Scaife etal., 2011) which is strongly related to the NAO (e.g. Woollings etal., 2010).

• Initialisation of sea ice (K. A. Peterson, 2013; personal communication). This is necessary to capture potential influences on the NAO from anomalous Arctic sea-ice conditions, as suggested in several recent studies (e.g. Overland and Wang, 2010; Petoukhov and Semenov, 2010; Francis and Vavrus, 2012; Liu et al., 2012; Tang et al., 2013).

GloSea5 retrospective forecasts (hindcasts) covering the period 1993-2012 show correlations greater than 0.6 between the observed winter-mean surface NAO index and the ensemble mean of model retrospective forecasts starting around the beginning of November each year (Scaife etal., 2014a). There are also similarly encouraging levels of skill for predicting the AO in the latest National Centers for Environmental Prediction (NCEP) seasonal forecasting system, Coupled Forecast System version 2 (CFSv2), over the longer period 1982-2009 (Riddle etal., 2013). These studies provide emerging evidence that the NAO and AO may be significantly predictable on seasonal time-scales. This is further supported by predictions of sea-level pressure in GloSea5 (Figure 3), which show significantly high correlations for regions around the centres of action of the NAO (Iceland and the Azores) with little skill in between, as would be expected ifskill arises from predictability of the NAO. Skilful predictions of the NAO and associated sea-level pressure leads to significant skill for important surface climate impacts including extremes of temperature, wind speed and storminess over land regions in Europe and North America (Scaife et al., 2014a).

Although sources of skill cannot be established without additional experiments, composite maps (Scaife etal., 2014a) suggest that many of the sources of predictability discussed above

1.0 0.9 0.8

■9 0.6

as . _

°0.3 0.2 0.1


0.0 0.5 1.0 1.5 2.0 2.5 3.0 Signal:noise

Figure 4. Skill versus signal-to-noise. Theoretical relationship between correlation and signal-to-noise ratio (solid curve following Kumar (2009)) and value from the ensemble mean of24 GloSea5 ensemble members (red circle with bars showing the 90% confidence interval: Scaife etal., 2014a). Uncertainties on the correlation were computed by bootstrapping both the ensemble members and retrospective forecast start dates (see Eade etal. (2014) for full details). Uncertainties in signal-to-noise were computed from the uncertainty in the standard deviation of the ensemble mean using the standard formula for the distribution of sample variance.

also appear to be operating in GloSea5, including ENSO, Arctic sea ice (especially in the Kara Sea), Atlantic SSTs and the QBO. However, the magnitude of the response in GloSea5 associated with some of these factors appears to be weak compared to observed composites for example. Interestingly, the correlation skill for the ensemble mean is higher than would be expected given the small signal-to-noise ratio present in the ensemble members (Kumar, 2009; Eade etal., 2014). This important issue will be discussed further in section 4.

Some of the potential drivers of the NAO, especially Atlantic SSTs and changes in Arctic sea ice, are also potentially predictable on interannual to decadal time-scales. Indeed, CMIP5 and other decadal predictions do show skill in predicting North Atlantic SSTs, with improvement through initialisation with observations (Doblas-Reyes etal., 2013; Hazeleger etal., 2013; Smith etal., 2010a). This is consistent with idealized model studies showing potential predictability of the Atlantic Meridional Overturning Circulation (AMOC: e.g. Griffies and Bryan, 1997; Pohlmann etal., 2004; Collins etal., 2006; Dunstone and Smith, 2010) which is expected to influence North Atlantic SST (Knight etal., 2005; Delworth etal., 2007). Assessing the skill for real-world predictions of the AMOC is hampered by a lack of observations for verification. However, Pohlmann et al. (2013) found potential predictability of the AMOC a few years ahead when assessed against a multi-model synthesis of ocean observations, potentially

providing a physical basis for improved North Atlantic SST predictions. Furthermore, some studies show that the rapid warming of the North Atlantic sub-polar gyre in the mid-1990s (Robson etal., 2012; Yeager etal., 2012) and the 1920s (Muller etal., 2014), and the cooling in the 1960s (Hermanson etal., 2014; Robson et al., 2014) could have been predicted in advance, with initialisation of anomalous AMOC playing an important role. However, despite this encouraging evidence for decadal predictability of North Atlantic SST, skilful predictions of the NAO beyond the seasonal range have not yet been achieved.

4. Future issues

The emerging NAO and AO correlation skill achieved by GloSea5 and CFSv2 is sensitive to ensemble size (Riddle et al., 2013; Scaife etal., 2014a). For single members, the correlation in GloSea5 is less than 0.2, but increases to above 0.6 for 24 members, and potentially asymptotes to greater than 0.8 for a very large ensemble. This is inconsistent with the signal-to-noise ratio (measured by the variance of the ensemble mean divided by the variance of individual ensemble members around the mean: Kumar, 2009) as shown in Figure 4. Such inconsistencies are also found in the North Atlantic in regional analysis of sea-level pressure and surface temperature on seasonal and multi-year time-scales (Eade etal., 2014). This means that the predictable component is greater in reality than in individual model members, showing that the model NAO is likely to be less constrained by the relevant driving factors compared to the observed NAO. Nevertheless, skilful predictions may be obtained by taking the mean of a large ensemble and making appropriate corrections to the variance of both the ensemble mean and individual members to ensure that the predictable component matches that in observations (Eade et al., 2014).

Further work is needed to understand why the predictable component is sometimes lower in models than observations, even though the total variance may be the same (Scaife etal., 2014a). There is mounting evidence that models respond too weakly to North Atlantic SSTs. This is shown by direct analysis of model simulations (Rodwell and Folland, 2002; Gastineau etal., 2013) and inferred from a simple theoretical explanation of the lagged response to changes in solar radiation (Scaife et al., 2013). Evidence suggests that atmospheric response to North Atlantic SSTs is stronger in higher-resolution models that resolve SST fronts in the Gulf Stream region (Minobe etal., 2008). Future high-resolution models may therefore yield improved levels of skill, or similar skill with fewer ensemble members.

Given the potential influence of North Atlantic SST on the NAO, any skill in predicting changes in the NAO over the coming decade likely requires predicting North Atlantic SSTs. Decadal variations in North Atlantic SST may occur through natural cycles of the Atlantic Multidecadal Oscillation (AMO) associated with changes in the strength of the AMOC (Knight etal., 2005; Delworth etal., 2007; Ting etal., 2009). However, there is evidence that recent variations in the AMO may have been forced by changes in anthropogenic aerosol emissions driven by socio-economic factors (Booth etal., 2012; Dunstone etal., 2013), although the magnitude of aerosol influences is debated (Zhang et al., 2013). Either way, the signal-to-noise ratio is smaller in models than in observations (Eade et al., 2014). Understanding the reasons for this and the relative importance and interplay between natural variability and external factors is therefore needed to gain confidence in predictions of future changes in North Atlantic SST and associated decadal fluctuations of the NAO.

5. Conclusions

Until recently, operational seasonal predictions typically had low skill for predicting the winter-mean surface NAO (Johansson, 2007; Arribas etal., 2011; Kim etal., 2012) and previous studies

have concluded that there could be little predictability even for extreme winters (e.g. Jung etal., 2011). However, observations and modelling studies suggest potential sources of skill, including ENSO, North Atlantic SST, Arctic sea ice, Eurasian snow cover, the QBO and solar variability. There is emerging evidence in retrospective forecasts that the NAO and AO exhibit predictability at least a month ahead (Riddle etal., 2013; Scaife etal., 2014a). Associated weather including the likelihood of damaging winter storms, near-surface wind speeds and extreme temperatures may also be predictable (Scaife etal., 2014a). Some of the drivers of the NAO would also be expected to operate on multi-year time-scales. Skilful multiannual predictions of the NAO have not yet been achieved but might be possible with improved models that better capture these potential sources of skill. A particular need for model improvement is highlighted by a mismatch between correlation skill and signal-to-noise ratio (Eade et al., 2014; Scaife etal., 2014a), suggesting that in models the atmosphere is not constrained strongly enough by the relevant driving factors.


This work was supported by the joint DECC/Defra Met Office Hadley Centre Climate Programme (GA01101), and the EU FP7 SPECS project.


Allen RJ, Ansell TJ. 2006. A new globally complete monthly historical mean sea level pressure data set (HadSLP2): 1850-2004. J. Clim. 19: 5816-5842. Allen RJ, Zender CS. 2011. Forcing of the Arctic Oscillation by Eurasian snow

cover. J. Clim. 24: 6528-6539, doi: 10.1175/2011JCLI4157.1. Anstey JA, Shepherd TG. 2014. High-latitude influence of the quasi-biennial

oscillation. Q. J. R. Meteorol. Soc. 140: 1-21, doi: 10.1002/qj.2132. Arribas A, Glover M, Maidens A, Peterson K, Gordon M, MacLachlan C, Graham R, Fereday D, Camp J, Scaife AA, Xavier P, McLean P, Colman A, Cusack S. 2011. The GloSea4 ensemble prediction system for seasonal forecasting. Mon. Weather Rev. 139: 1891-1910. Baldwin MP, Dunkerton TJ. 2001. Stratospheric harbingers of anomalous

weather regimes. Science 294: 581-584, doi: 10.1126/science.1063315. Bell CJ, Gray LJ, Charlton-Perez AJ, Joshi M, Scaife AA. 2009. Stratospheric communication of El Niño teleconnections to European winter. J. Clim. 22: 4083-4096.

Booth BBB, Dunstone NJ, Halloran PR, Andrews T, Bellouin N. 2012. Aerosols implicated as a prime driver of twentieth-century North Atlantic climate variability. Nature 484: 228-232, doi: 10.1038/nature10946. Bretherton CS, Battisti DS. 2000. An interpretation of the results from atmospheric general circulation models forced by the time history of the observed sea surface temperature distribution. Geophys. Res. Lett. 27: 767-770, doi: 10.1029/1999GL010910. Bronnimann S, Xoplaki E, Casty C, Pauling A, Luterbacher J. 2007. ENSO

influence on Europe during the last centuries. Clim. Dyn. 28: 181-197. Cagnazzo C, Manzini E. 2009. Impact of the stratosphere on the winter tropospheric teleconnections between ENSO and the North Atlantic and European region. J. Clim. 22: 1223-1238, doi: 10.1175/2008JCLI2549.1. Christensen JH, Krishna Kumar K, Aldrian E, An S-I, Cavalcanti IFA, de Castro M, Dong W, Goswami P, Hall A, Kanyanga JK, Kitoh A, Kossin J, Lau N-C, Renwick J, Stephenson DB, Xie S-P, Zhou T. 2013. Climate phenomena and their relevance for future regional climate change. In Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM. (eds.): 1217-1308. Cambridge University Press: Cambridge, UK and New York, NY. Cohen J, Entekhabi D. 1999. Eurasian snow cover variability and Northern Hemisphere climate predictability. Geophys. Res. Lett. 26: 345-348, doi: 10.1029/1998GL900321. Collins M, Botzet M, Carril AF, Drange H, Jouzeau A, Latif M, Masina S, Otteraa OH, Pohlmann H, Sorteberg A, Sutton R, Terray L. 2006. Interannual to decadal climate predictability in the North Atlantic: A multimodel-ensemble study. J. Clim. 19: 1195-1203. Delworth TL, Zhang R, Mann ME. 2007. Decadal to centennial variability of the Atlantic from observations and models. Ocean Circulation: Mechanisms and Impacts, Schmittner A, Chiang JCH, Hemming SR. (eds.) Geophysical Monograph Series 173:131-148. American Geophysical Union: Washington, DC.

Doblas-Reyes FJ, Pavan V, Stephenson DB. 2003. Multi-model seasonal hindcasts of the North Atlantic Oscillation. Clim. Dyn. 21: 501-514.

Doblas-Reyes FJ, Andreu-Burillo I, Chikamoto Y, García-Serrano J, Guemas V, Kimoto M, Mochizuki T, Rodrigues LRL, van Oldenborgh GJ. 2013. Initialized near-term regional climate change prediction. Nat. Commun. 4: 1715, doi: 10.1038/ncomms2704.

Driscoll S, Bozzo A, Gray LJ, Robock A, Stenchikov G. 2012. Coupled Model Intercomparison Project 5 (CMIP5) simulations of climate following volcanic eruptions. J. Geophys. Res. 117: D17105, doi: 10.1029/2012JD017607.

Dunstone NJ, Smith DM. 2010. Impact of atmosphere and sub-surface ocean data on decadal climate prediction. Geophys. Res. Lett. 37: L02709, doi: 10.1029/2009GL041609.

Dunstone NJ, Smith DM, Booth BBB, Hermanson L, Eade R. 2013. Anthropogenic aerosol forcing of Atlantic tropical storms. Nat. Geosci. 6: 534-539, doi: 10.1038/ngeo1854.

Eade R, Smith DM, Scaife AA, Wallace E, Dunstone N, Hermanson L, Robinson N. 2014. Do seasonal to decadal climate predictions underestimate the predictability of the real world? Geophys. Res. Lett. 41: 5620-5628, doi: 10.1002/2014GL061146.

Ebdon RA. 1975. The quasi-biennial oscillation and its association with tropospheric circulation patterns. Meteorol. Mag. 104: 282-297.

Fereday D, Maidens A, Arribas A, Scaife AA, Knight JR. 2012. Seasonal forecasts of Northern Hemisphere winter 2009/10. Environ. Res. Lett. 7: 034031, doi: 10.1088/1748-9326/7/3/034031.

Francis JA, Vavrus SJ. 2012. Evidence linking Arctic amplification to extreme weather in mid-latitudes. Geophys. Res. Lett. 39: L06801, doi: 10.1029/2012GL051000.

Frankignoul C, Hasselmann K. 1977. Stochastic climate models. Part 2, Application to sea-surface temperature anomalies and thermocline variability. Tellus 29: 289-305.

Gastineau G, D'Andrea F, Frankignoul C. 2013. Atmospheric response to the North Atlantic Ocean variability on seasonal to decadal time scales. Clim. Dyn. 40: 2311-2330, doi: 10.1007/s00382-012-1333-0.

Gray LJ, Scaife AA, Mitchell DM, Osprey S, Ineson S, Hardiman S, Butchart N, Knight JR, Sutton R, Kodera K. 2013. A lagged response to the 11 year solar cycle in observed winter Atlantic/European weather patterns. J. Geophys. Res. Atmos. 118: 13405-13420, doi: 10.1002/2013JD020062.

Griffies SM, Bryan K. 1997. Predictability of North Atlantic multidecadal climate variability. Science 275: 181.

Hardiman SC, Kushner PJ, Cohen J. 2008. Investigating the ability of general circulation models to capture the effects of Eurasian snow cover on winter climate. J. Geophys. Res. 113: D21123, doi: 10.1029/2008JD010623.

Hazeleger W, Wouters B, van Oldenborgh GJ, Corti S, Palmer TN, Smith D, Dunstone N, Kroger J, Pohlmann H, von Storch J-S. 2013. Predicting multiyear North Atlantic Ocean variability. J. Geophys. Res. Oceans 118: 1087-1098, doi: 10.1002/jgrc.20117.

Hermanson L, Eade R, Robinson NH, Dunstone NJ, Andrews MB, Knight JR, Scaife AA, Smith DM. 2014. Forecast cooling of the Atlantic subpolar gyre and associated impacts. Geophys. Res. Lett. 41: 5167-5174, doi: 10.1002/2014GL060420.

Hurrell JW. 1995. Decadal trends in the North Atlantic Oscillation: Regional temperatures and precipitation. Science 269: 676-679.

Ineson S, Scaife AA. 2009. The role of the stratosphere in the European climate response to El Nino. Nat. Geosci. 2: 32-36.

Ineson S, Scaife AA, Knight JR, Manners JC, Dunstone NJ, Gray LJ, Haigh JD. 2011. Solar forcing ofwinter climate variability in the Northern Hemisphere. Nat. Geosci. 4: 753-757, doi: 10.1038/ngeo1282.

Johansson A. 2007. Prediction skill of the NAO and PNA from daily to seasonal time scales. J. Clim. 20: 1957-1975.

Jung T, Vitart F, Ferranti L, Morcrette J-J. 2011. Origin and predictabilityofthe extreme negative NAO winter of 2009/2010. Geophys. Res. Lett. 38: L07701, doi: 10.1029/2011GL046786.

Kenyon J, Hegerl GC. 2010. Influence of modes of climate variability on global precipitation extremes. J. Clim. 23: 6248-6262, doi: 10.1175/2010JCLI3617.1.

Kim H-M, Webster PJ, Curry JA. 2012. Seasonal prediction skill of ECMWF System 4 and NCEP CFSv2 retrospective forecast for the Northern Hemisphere winter. Clim. Dyn. 39: 2957-2973, doi: 10.1007/s00382-012-1364-6.

Knight JR, Allan RJ, Folland CK, Vellinga M, Mann ME. 2005. A signature of persistent natural thermohaline circulation cycles in observed climate. Geophys. Res. Lett. 32: L20708, doi: 10.1029/2005GL024233.

Knight JR, Folland CK, Scaife AA. 2006. Climate impacts of the Atlantic Multidecadal Oscillation. Geophys. Res. Lett. 33: L17706, doi:10.1029/2006GL026242.

Kumar A. 2009. Finite samples and uncertainty estimates for skill measures for seasonal prediction. Mon. Weather Rev. 137: 2622 -2631.

Liu J, Curry JA, Wang H, SongM, HortonRM. 2012. Impact of declining Arctic sea ice on winter snowfall. Proc. Natl. Acad. Sci. U.S.A. 109: 4074-4079.

Lockwood M, Harrison RG, Woollings T, Solanki SK. 2010. Are cold winters in Europe associated with low solar activity? Environ. Res. Lett. 5: 024001, doi: 10.1088/1748-9326/5/2/024001.

MacLachlan C, Arribas A, Peterson D, Maidens A, Fereday D, Scaife AA, Gordon M, Vellinga M, Williams A, Comer RE, Camp J, Xavier P. 2014.

Global Seasonal forecast system version 5 (GloSea5): A high-resolution seasonal forecast system. Q. J. R Meteorol. Soc., doi: 10.1002/qj.2396.

Manzini E, Giorgetta MA, Esch M, Kornblueh L, Roeckner E. 2006. The influence of sea surface temperatures on the northern winter stratosphere: Ensemble simulations with theMAECHAM5 model. J. Clim. 19:3863-3881.

Marshall A, Scaife AA. 2009. Impact of the quasi-biennial oscillation on seasonal forecasts. J. Geophys. Res. 114: D18110, doi: 10.1029/2009JD011737.

Marshall A, Scaife AA, Ineson S. 2009. Enhanced seasonal prediction of European winter warming following volcanic eruptions. J. Clim. 22: 6168-6180.

MatthesK, KurodaY, KoderaK, LangematzU. 2006. Transfer of the solar signal from the stratosphere to the troposphere: Northern winter. J. Geophys. Res. 111: D06108, doi: 10.1029/2005JD006283.

Minobe S, Kuwano-Yoshida A, Komori N, Xie S-P, Small RJ. 2008. Influence of the Gulf Stream on the troposphere. Nature 452: 206-209, doi: 10.1038/nature06690.

Morice CP, Kennedy JJ, Rayner NA, Jones PD. 2012. Quantifying uncertainties in global and regional temperature change using an ensemble of observational estimates: The HadCRUT4 data set. J. Geophys. Res. 117: D08101, doi: 10.1029/2011JD017187.

Moron V, Gouirand I. 2003. Seasonal modulation of the El Nino-Southern Oscillation relationship with sea level pressure anomalies over the North Atlantic in October-March 1873-1996. Int. J. Climatol. 23: 143-155.

Müller WA, Appenzeller C, Schär C. 2005. Probabilistic seasonal prediction of the winter North Atlantic Oscillation and its impact on near surface temperature. Clim. Dyn. 24: 213-226.

Müller WA, Pohlmann H, Sienz F, Smith D. 2014.. Decadal climate predictions for the period 1901 -2010 with a coupled climate model. Geophys. Res. Lett. 41: 2100-2107, doi: 10.1002/2014GL059259.

Niwano M, Takahashi M. 1998. Notes and correspondence: The influence of the equatorial QBO on the Northern Hemisphere winter circulation of a GCM. J. Meteorol. Soc. Jpn. 76: 453 -461.

Omrani NE, Keenleyside NS, Bader J, Manzini E. 2014. Stratosphere key for wintertime atmospheric response to warm Atlantic decadal conditions. Clim. Dyn. 42: 649-663.

Overland JE, Wang M. 2010. Large-scale atmospheric circulation changes associated with the recent loss of Arctic sea ice. Tellus A 62: 1-9.

Peings Y, Saint-Martin D, Douville H. 2012. A numerical sensitivity study of the Siberian snow influence on the Northern Annular Mode. J. Clim. 25: 592-607.

Peng S, Robinson WA, Hoerling MP. 1997. The modeled atmospheric response to midlatitude SST anomalies and its dependence on background circulation states. J. Clim. 10: 971-987.

Petoukhov V, Semenov VA. 2010. A link between reduced Barents-Kara sea ice and cold winter extremes over northern continents. J. Geophys. Res. 115: D21111, doi: 10.1029/2009JD013568.

Pohlmann H, Botzet M, Latif M, Roesch A, Wild M, Tschuck P. 2004. Estimating the decadal predictability of a coupled AOGCM. J. Clim. 17: 4463 -4472.

Pohlmann H, Smith DM, Balmaseda MA, Keenlyside NS, Masina S, Matei D, Muäller WA, Rogel P. 2013. Predictability of the mid-latitude Atlantic meridional overturning circulation in a multi-model system. Clim. Dyn. 41: 775-785, doi: 10.1007/s00382-013-1663-6.

Ratcliffe RAS, Murray R. 1970. New lag association between North Atlantic sea surface temperature and European pressure applied to long range weather forecasting. Q. J. R. Meteorol. Soc. 96: 226-246.

Riddle EE, Butler AH, Furtado JC, Cohen JL, Kumar A. 2013. CFSv2 ensemble prediction of the wintertime Arctic Oscillation. Clim. Dyn. 41: 1099-1116.

Roberts JF, Champion AJ, Dawkins L, Hodges KI, Shaffrey LC, Stephenson DB, Stringer MA, Thornton HE, Youngman BD. 2014. The XWS open access catalogue of extreme European windstorms from 1979-2012. Nat. Hazards Earth Syst. Sci. Discuss. 2: 2011-2048.

Robson JI, Sutton RT, Smith DM. 2012. Initialized decadal predictions of the rapid warming of the North Atlantic ocean in the mid 1990s. Geophys Res. Lett. 39: L19713, doi: 10.1029/2012GL053370.

Robson JI, Sutton RT, Smith DM. 2014. Decadal predictions of the cooling and freshening of the North Atlantic in the 1960s and the role of the ocean circulation. Clim. Dyn. 42: 2353-2365, doi: 10.1007/s00382-014-2115-7.

Rodwell MJ, Folland CK. 2002. Atlantic air-sea interaction and seasonal predictability. Q. J. R. Meteorol. Soc. 128: 1413-1443.

Rodwell MJ, Rowell DP, Folland CK. 1999. Oceanic forcing of the wintertime North Atlantic Oscillation and European climate. Nature 398: 320-323.

Scaife AA, Folland CK, Alexander L, Moberg A, Knight JR. 2008. European climate extremes and the North Atlantic Oscillation. J. Clim. 21: 72-83.

Scaife AA, Copsey D, Gordon C, Harris C, Hinton T, Keeley SJ, O'Neill A, Roberts M, Williams K. 2011. Improved Atlantic blocking in a climate model. Geophys. Res. Lett. 38: L23703, doi: 10.1029/2011GL049573.

Scaife AA, Ineson S, Knight JR, Gray L, Kodera K, Smith DM. 2013. A mechanism for lagged North Atlantic climate response to solar variability. Geophys. Res. Lett. 40: 434-439, doi: 10.1002/grl.50099.

Scaife AA, Arribas A, Blockley E, Brookshaw A, Clark RT, Dunstone N, Eade R, Fereday D, Folland CK, Gordon M, Hermanson L, Knight JR, Lea DJ, MacLachlan C, Maidens A, Martin M, Peterson AK, Smith D, Vellinga M, Wallace E, Waters J, Williams A. 2014a. Skillful long-range prediction of

European and North American winters. Geophys. Res. Lett. 41: 2514-2519, doi: 10.1002/2014GL059637.

Scaife AA, Athanassiadou M, Andrews M, Arribas A, Baldwin M, Dunstone N, Knight J, MacLachlan C, Manzini E, MUller WA, Pohlmann H, Smith D, Stockdale T, Williams A. 2014b. Predictability of the quasi-biennial oscillation and its northern winter teleconnection on seasonal to decadal timescales. Geophys. Res. Lett. 41: 1752-1758, doi: 10.1002/2013GL059160.

Smith DM, Eade R, Dunstone NJ, Fereday D, Murphy JM, Pohlmann H, Scaife AA. 2010a. Skilful multi-year predictions of Atlantic hurricane frequency. Nat. Geosci. 3: 846-849, doi: 10.1038/ngeo1004.

Smith KL, Fletcher CG, Kushner PJ. 2010b. The role of linear interference in the annular mode response to extratropical surface forcing. J. Clim. 23: 6036-6050.

Smith DM, Scaife AA, Kirtman B. 2012. What is the current state of scientific knowledge with regard to seasonal and decadal forecasting? Environ. Res. Lett. 7: 015602, doi: 10.1088/1748-9326/7/1/015602.

Stenchikov G, Hamilton K, Robock A, Ramaswamy V, Schwarzkopf MD. 2004. Arctic Oscillation response to the 1991 Pinatubo eruption in the SKYHI general circulation model with a realistic quasi-biennial oscillation. J. Geophys. Res. 109: D03112, doi: 10.1029/2003JD003699.

Stephenson DB, Pavan V, Bojariu R. 2000. Is the North Atlantic Oscillation a random walk? Int. J. Climatol. 20: 1 -18.

Tang Q, Zhang X, Yang X, Francis JA. 2013. Cold winter extremes in northern

continents linked to Arctic sea ice loss. Environ. Res. Lett. 8: 014036, doi: 10.1088/1748-9326/8/1/014036. Thompson DWJ, Wallace JM. 1998. The Arctic Oscillation signature in the wintertime geopotential height and temperature fields. Geophys. Res. Lett. 25: 1297-1300, doi: 10.1029/98GL00950. Thompson DWJ, Baldwin MP, Wallace JM. 2002. Stratospheric connection to Northern Hemisphere wintertime weather: Implications for prediction. J. Clim. 15: 1421-1428. Ting M, Kushnir Y, Seager R, Li C. 2009. Forced and internal twentieth-century

SST trends in the North Atlantic. J. Clim. 22: 1469-1481. Toniazzo T, Scaife AA. 2006. The influence of ENSO on winter North Atlantic

climate. Geophys. Res. Lett. 33: L24704, doi: 10.1029/2006GL027881. Walker GT, Bliss EW. 1932. World weather V. Mem. R. Meteorol. Soc. 4:53-84. Woollings T, Hannachi A, Hoskins B. 2010. Variability of the North Atlantic eddy-driven jet stream. Q. J. R. Meteorol. Soc. 136: 856-868, doi: 10.1002/qj.625. Yeager S, Karspeck A, Danabasoglu G, Tribbia J, Teng H. 2012. A decadal prediction case study: Late 20th century North Atlantic ocean heat content. J. Clim. 25: 5173-5189, doi: 10.1175/JCLI-D-11-00595.1. Zhang R, Delworth TL, Sutton R, Hodson DLR, Dixon KW, Held IM, Kushnir Y, Marshall J, Ming Y, Msadek R, Robson J, Rosati AJ, Ting MF, Vecchi GA. 2013. Have aerosols caused the observed Atlantic multidecadal variability? J.Atmos. Sci. 70: 1135-1144, doi: 10.1175/JAS-D-12-0331.1.