No. 72 June, 2017
CLIVAR Exchanges
CLIVAR Ocean and Climate: Variability, Predictability and Change is the World Climate Research Programme’s core project on the Ocean-Atmosphere System
Decadal Climate Variability
• Figure by Adam Phillips, picture by Christophe Cassou
Volume 25, No. 1
Past Global Changes Magazine Editorial doi: 10.22498/pages.25.1.1 Yochanan Kushnir, Christophe Cassou, Scott St George The study of Decadal Climate Variability and Predictability variability and greenhouse gas forcing, have devastated regional agriculture productivity, lead to loss of life and, characterize, understand, attribute, simulate, and predict perhaps arguably, to widespread societal instability and in the(DCVP) slow, is multi-year the interdisciplinary variations of scientific climate on enterprise global and to regional scales. Particular interest in decadal climate variations and their role in the global surface climate InSyria order to violent to anticipate conflict the and impacts war. of climate change, it is change stems from the need to detect and attribute the important for society to know how the climate response uneven rise in global mean surface temperatures (GMST) to anthropogenic forcing and the climate impact of since the beginning of the industrial period. The most natural variability will mix together to affect the near- recent expression of decadal variability in GMST has been term future. The study of DCVP aims to provide science- the slowdown in warming between roughly 1998 and based information to decision makers through research, 2012. This period, often termed as the “hiatus”, triggered observations, and decadal predictions. This goal remains intensive debate in the public domain, even if global challenging despite decades of research and of extensive temperatures had exhibited long undulations before, progress in observing and modeling the climate system. including two cooling events in the late 19th to early 20th Predicting the impact of internal decadal climate variability century and in the mid 20th century and two intervals is complicated by our incomplete understanding of the of rapid warming, one from about 1910 to 1940 and the nature of the underlying phenomena, in particular their other between the early 1970s and 1998. While these physical origins and their interaction with external forcing. departures from the expected warming due to the steady Existing obstacles in DCVP research thus test our ability to increase in greenhouse gas forcing have been attributed in attribute past variations to the combined role of internal part to natural (volcanic) and anthropogenic (industrial) variability and external forcing, as well as to reliably predict aerosols, there is ample evidence that long-term internal the near-term climate on global and regional scales. interactions between climate system components – the ocean and the atmosphere, in particular – have also been Progress in DCVP research can only be made through involved. international, cross disciplinary collaborations between
Decadal and longer variations in sea surface temperatures the Earth’s climate at timescales of a decade or longer, (SSTs) have a rich and non-uniform spatial pattern thisscientists. area ofBecause research of the is difficultieswholly dependent to observe on and emerging model related to variations in the distribution of precipitation connections between those who perform, collect and and associated atmospheric convection in the tropics, to analyze instrumental observations of the present, those alterations in position and strength of the storm tracks who develop and analyze proxies of past climate, and with at midlatitudes, to changes in sea-ice extent at polar scientists who develop models and perform dedicated latitudes in both hemispheres, etc. Changes in atmospheric modeling experiments. To review ongoing research on circulation thus contributes to changes in regional DCVP and propose the road to future progress on the climates worldwide and importantly over the continents, subject, the International WCRP CLIVAR Project and PAGES directly affecting humans and their environment. The held an international workshop with representatives of most prominent example of the terrestrial response to these various disciplines in November of 2015, under decadal climate variability is the long-lasting decline of the patronage of the International Centre for Theoretical rainfall in the North African Sahel in the second half of the Physics in Trieste Italy. This issue of Exchanges grew out 20th century, which included the devastating famines of from the presentations and discussions features in this the 1970s & 1980s. These decadal-scale shifts have been workshop. attributed to slow variations in North Atlantic SSTs, which have also affected Atlantic tropical cyclone activity over the The articles in this issue of Exchanges were selected and same time frames. Similarly, the multi-year pulses of North reviewed by the members of the CLIVAR Working Group American droughts (e.g., the Great Plains “dust bowl” in on DCVP and the PAGES 2k Network. These contributions the 1930s and the recent protracted dry period in the are meant to provide brief reviews that address the Southwest US), which impacted lives and livelihoods in the progress made in understanding and resolving different US and northern Mexico, have been attributed to the state key issues in DCVP. We greatly appreciate the voluntary efforts made by these authors to capture the exciting and Mediterranean region, in South European countries along rapidly growing literature on the subject in these brief theof the northern tropical rim Pacific and andin the tropical Maghreb Atlantic and the Oceans. Middle For East, the summaries and hope that they will stimulate further recurrent heat waves and prolonged dry spells since the research collaboration on the subject. mid 1960s, attributed to a combination of internal decadal
1 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 An overview of decadal-scale sea surface temperature variability in the observational record
doi: 10.22498/pages.25.1.2 Clara Deser, Adam Phillips National Center for Atmospheric Research, Boulder, USA
Introduction Due to their thermal and mechanical inertia, the oceans SST data coverage play a key role in decadal-scale climate variability (DCV) and provide a potential source of initial-value oceanOur focus temperature on SST is records motivated are by measured both practical near andthe Characterizing oceanic DCV is challenging, however, surfacephysical from considerations. ships-of-opportunity, On the practical starting side, with the longestbucket duepredictability to the limited for duration low-frequency of the climateobservational fluctuations. record samples in the 19th and early 20th centuries followed combined with the sparse and irregular data coverage. by engine-intake measurements (e.g., Woodruff et al., These constraints also hinder assessments of the robustness of the patterns and timescales of DCV, and communication between the atmosphere and the ocean, understanding of the governing mechanisms. In this brief and2008). thus On represent the physical a key side, quantity SSTs forare probingthe main DCV agent (for of a note, we provide an overview of the main phenomena discussion of the upper-ocean mixed layer heat budget, of DCV in the historical sea surface temperature (SST) c.f. Deser et al., 2010). data record, discuss proposed interpretations and causal mechanisms, and highlight outstanding research questions. Fig. 1 (left column) shows maps of SST data coverage duringbased three on the representative International 20-year Comprehensive periods spanning Ocean theAtmosphere late 19th and Data 20th Set centuries: (ICOADS) (Woodruff1870-1899, et 1920-1939, al., 2008) and 1970-1989. These maps show the percentage of months with at least one measurement in a 2°latitude by 2° longitude grid box in the 20-year period indicated. We note that the instrumental coverage falls off rapidly before 1870, and that satellites provide nearly global coverage starting in the 1980s (see Woodruff et al., 2008 and Deser et al., 2010). The discrete outlines of commercial shipping routes and their changes over time are readily
Broadly speaking, the North Atlantic, western South apparent, especially in the earlier time periods (Fig. 1). density of observations, with reasonable coverage back toAtlantic, approximately and northern 1870. Indian Data Ocean coverage contain in thethe highest North
Pacific is limited before about 1920, in the Tropics before changingabout 1960, spatial and in coverage the Southern of SST Ocean measurements before the advent from historicalof satellite ship-based remote sensingarchives (Fig.must 1).be taken The into uneven account and Figure 1: Distribution of sea surface temperature observations from the International Comprehensive Ocean Atmosphere Data Set. Maps show the percentage of months with at least temporal coverage of other SST data sets is available at in any analysis of DCV. Further information on the spatio- one measurement in a 2 degree latitude by 2 degree longitude climatedataguide.ucar.edu. grid box during (a) 1870-1899, (b) 1920-1939, and (c) 1970- 1989. Timeseries (1870-2015) show the percentage of grid The main phenomena of DCV boxes that have at least one observation per month within In our view, there is no unique “best” approach to the regions outlined in Fig. 1c. (d) North Pacific (20°-70°N, 110°E-100°W), (e) North Atlantic (0°-60°N, 80°W-0°W), and (f) Southern Ocean (50°-70°S, 0°W-360°E). defining the main phenomena of DCV. Here, we adopt a basin-specific perspective, which has the advantage that
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 2 any inter-basin linkages (including those lagged in time) in 1890 (mindful of the reduced coverage in NPAC before are not built-in to the analysis protocol. Similarly, we analyze monthly data (lightly smoothed with a 3-point global regression maps are all based on the period since 19501920), to and accommodate the SO record the startinglack of data in 1950. over However,the Southern the low-frequency behavior. In this regard, it is important torunning bear inmean) mind so the as tonull avoid hypothesis artificially that building any low-pass in any that time. Ocean (and to a lesser extent, the Tropical Pacific) before physically meaningful (i.e., it may not be distinguishable The three SSTA* patterns show a great deal of similarity fromfiltered a random time series process). will exhibit DCV, but it need not be in their global structures, despite that they are based
Temperature, version 3b (ERSSTv3b) dataset, which equator,on different reminiscent index regions. of the For low-frequency example, NPAC “tail” shows of We use the NOAA Extended Reconstruction Sea Surface a pan-Indo/Pacific pattern with symmetry about the employs a statistical procedure on the ICOADS data to ENSO (also termed the “Pacific Decadal Oscillation PDO studies,fill in missing we subtract grid boxes the global (Smith mean et al., SST 2008); anomaly other (SSTA) data 2016).or “Inter-Decadal It also features Pacific linkages Oscillation to the IPO) Atlantic (Zhang in et the al., fromsets yield the SSTAsimilar at resultseach grid (not box shown). for each Following month andprevious year form1997; of Power alternating et al., polarities1999; Vimont, with latitude,2005; Newman with positive et al., (hereafter, we use the nomenclature SSTA* to denote this values over the northern North Atlantic, and negative residual from the global mean) unless noted otherwise. This procedure is intended to remove any secular global trends that may be associated with changes in external betweenvalues over SSTA* the in Pacificthe North sector and ofSouth the Atlantic, Southern distinct Ocean radiative forcing such as those due to human-induced (Fig. 2a). NATL exhibits an out-of-phase relationship increases in greenhouse gas concentrations and sulfate the same SSTA* polarity over the northern NATL aerosols accompanying fossil fuel burning. We shall from that based on NPAC (Fig. 2b). However, it shares return to the issue of how well this procedure achieves magnitude, as that based on NPAC, and it shows negative its intended purpose in Section 5. and the same PDO-like structure, albeit with weaker
values throughout the Southern Ocean (Fig. 2b). principalWe define component the main phenomena(PC) time series of DCV of monthly for each SSTA* basin overseparately the domain as follows. (20°-70°N, North 110°E-100°W) Pacific (NPAC): following leading Mantua et al. (1997). North Atlantic (NATL): time series of SSTA* averaged over the domain (0°-60°N, 80°W-0°W) following Trenberth and Shea (2006). Southern Ocean (SO): time series of SSTA (not SSTA*) averaged over comparisonthe domain (50°-70°S,with NPAC 0°W-360°E) and NATL. followingThese regions Fan et are al. (2014). We have inverted the SO time series to facilitate associated with each time series, we regress SSTA* (SSTA outlined in Fig. 1c. To obtain the global-scale patterns index time series. for the case of the SO) at each grid box on the standardized Before showing the spatial patterns of DCV, we return to above,the issue represented of data coverage. as the Thepercentage right-hand of grid panels boxes of Fig. that 1 show time series of data coverage in each region defined Figure 2: Spatial and temporal characteristics of sea surface have at least one observation in each month. Consistent temperature anomaly (SSTA) variability in selected ocean with the data coverage maps, the NPAC region shows basins. (Left column) Global SSTA* regression maps (degrees >50% of grid boxes with at least one observation starting C) based on the (a) leading principal component of North Pacific SSTA*, (b) North Atlantic SSTA*, and (c) inverted region shows >50% of grid boxes present since about Southern Ocean SSTA. All indices were standardized prior to around 1920 except during the 1940s (Fig. 1d). The NATL computing the regression maps. Index regions are outlined by black boxes. (Right column) Standardized 3-month running 1885 except for the World Wars and around 1900 (Fig. mean time series (1880-2015) of the (a) leading principal 1e). Finally, coverage in the SO region is always <40%, component of North Pacific SSTA*, (b) North Atlantic SSTA*, and <10% before 1950 (Fig. 1f). A large seasonal cycle and (c) inverted Southern Ocean SSTA. Asterisk indicates that the global mean SSTA was removed prior to computing chooseis evident to show in the the SO NPAC region, and with NATL peak time coverage series starting during the time series and regression maps. summer (Fan et al., 2014). In view of these results, we
3 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Internal vs. externally-forced DCV very similar to that based on NATL, except for the sign Finally, the pattern based on the (inverted) SO index is As mentioned above, DCV is traditionally identified over the northern tropical Atlantic (Fig. 2c). as the residual from the global mean SSTA; the latter similar during their period of overlap (1950-2015), with isinterpreted not spatially as the uniform secular (Xie fingerprint et al., 2010) of human-induced and thus the pronouncedThe NPAC, NATL decadal-scale and SO index variability time series evident are remarkably even in removalclimate change. of the However,global-mean human-induced SSTA may fall climate short change of its intended purpose. The validity of this approach can be each record swings from positive to negative and back tested with large initial-condition ensembles of historical to3-month positive running during mean 1950-2015, data (Figs. with 2d-f). a suggestion In particular, that simulations with comprehensive coupled climate models, such as 40-member Community Earth System NPAC shows less prominent decadal-scale variability and Model Large Ensemble (CESM-LE) (Kay et al., 2015). aNATL weaker leads relationship NPAC and with SO by NATL 10-20 than years. after Before 1950. 1950, We The CESM-LE has been used to isolate externally-forced refrain from quantifying these statements due to the low and internally-generated components of simulated number of degrees of freedom associated with such short NATL DCV, with implications for observed NATL DCV records of DCV.
Causes of DCV since(Tandon 1920 and may Kushner, be externally-forced, 2015; Murphy etand al., that 2016). empirical These The substantial degree of commonality in the global-scale methodologiesstudies indicate used that ato significant separate theseportion components of NATL DCV in the single observational record may be inadequate. shortpatterns records associated and sparse with Pacific,sampling, Atlantic make andit challenging Southern empirical approaches for separating forced and internal toOcean identify DCV, these combined phenomena with the robustly fact that in they terms are ofbased spatial on componentsFurther work of isDCV, needed not only to in evaluate the NATL the but efficacy in other of and temporal character. These constraints also make it ocean basins as well. These approaches include linear (or other forms of) detrending, removal of the global- dynamical processes operating on decadal time scales and/ordifficult whether to assess they whether are best theyviewed arise as manifestations from distinctive of a “random walk” process or processes (see also Newman al.,mean, 2015), optimal and Empiricalfingerprinting Ensemble (Ting Mode et al., Decomposition 2009; Ting et et al., 2016). The concept of global-scale SSTA “hyper- (Wual., 2014), et al., pattern2011). scaling (Hoerling et al., 2011; Bichet et
Clement et al., 2011) has been advanced to account for Concluding remarks themodes” similarity (Dommenget of global-scale and Latif, SSTA 2008; patterns Dommenget, regardless 2010; We have presented a brief overview of the main of how they originate. This concept relies on the notion phenomena of DCV based on simple analyses of observed that the lower the frequency, the more global the pattern, due to the interplay between SSTA and the large-scale atmospheric circulation. SST over the historical record using a basin-specific approach. Our results indicate that DCV is apparent even These issues highlight the need for a combined approach characteristicsin unfiltered seasonal including data, their and global-scale that DCV patterns in the North and based on observations, paleo-climate records and chronologies,Atlantic, North especially Pacific and since Southern 1950. OceanResults share are manymore modeling to delineate robust phenomena of DCV and ambiguous before that time. Given the shortness of the to understand their causes. Indeed, modeling studies observational record relative to the time scales of interest, based on thousands of years of simulation for robust we believe that DCV is best viewed in terms of a case statistics suggest that Atlantic DCV and its global- study approach rather than as a robust and stationary scale teleconnections may originate from the mutual statistical characterization. Long (thousands of years) model control simulations provide an effective tool for assessing the robustness and global-scale linkages of interaction of the oceanic Atlantic Meridional Overturning DCV, provided the model has a credible representation Circulation (AMOC) and the large-scale atmospheric Ruprich-Robertcirculation in the et form al., 2016), of the althoughNorth Atlantic the mechanisms Oscillation outstanding research question is the extent to which DCV continue(NAO) (Delworth to be under et al., investigation. 2016; Delworth Similar and Zeng,conclusions 2016; isof externally-forced the relevant processes vs. internally-generated. governing DCV. Finally, an
References arise for Southern Ocean DCV (Latif et al, 2013; Latif et withal., 2015; a different Zhang decorrelation et al., 2016). Finally,time scale Pacific and DCVregional may 2015: Estimating the Anthropogenic Sea Surface emphasisreflect a arising combination from dynamical of stochastic and processes,thermodynamic each TemperatureBichet, A., P. Kushner,Response L. Using Mudryk, Pattern L. Terray,Scaling. and J. Climate, J. Fyfe, 28,3751–3763, doi: 10.1175/JCLI-D-14-00604.1. Newman et al., 2016). air-sea interaction (Clement et al., 2011; Okumura, 2013; Clement, A., P. DiNezio, and C. Deser, 2011: Rethinking the
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 4 Murphy, L. N., K. Bellomo, M. Cane and A. Clement, 4056-4072, doi: 10.1175/2011JCLI3973.1. Ocean's Role in the Southern Oscillation. J. Climate, 24, Res.2016: Lett., The in Role review. of Historical Forcings in Simulating the and X. Yang, 2016: The central role of ocean dynamics in Observed Atlantic Multidecadal Oscillation. Geophys. Delworth, T. L., F. Zeng, L. Zhang, R. Zhang, G. A. Vecchi1 Newman, M., M. A. Alexander, T. R. Ault, K. M. Cobb, C. Deser, E. Di Lorenzo, N. J. Mantua, A. J. Miller, S. Minobe, connecting the North Atlantic Oscillation to the Atlantic Multidecadal Oscillation, J. Climate, in press. oscillation,H. Nakamura, revisited. N. Schneider, J. Climate, D. J.in Vimont, press. doi: A. S. 10.1175/ Phillips, Delworth, T. L. and F. Zeng, 2016: The impact of the North JCLI-D-15-0508.1.J. D. Scott, and C. A. Smith, 2016: The Pacific decadal Atlantic Oscillation on climate through its impact on the Atlantic Meridional Overturning Circulation. J. Climate. Deser,DOI: 10.1175/JCLI-D-15-0396.1 C., M. A. Alexander, S. -P. Xie, and A. S. Phillips, variability: Role of stochastic atmospheric forcing from 2010: Sea surface temperature variability: patterns Okumura, Y. M., 2013: Origins of tropical Pacific decadal and mechanisms. Ann. Rev. Mar. Sci., 2010.2, 115-143, doi:10.1146/annurev-marine-120408-151453. the South Pacific. J. Climate 26 (24), 9791-9796.
Dommenget, D. and M. Latif, 2008: Generation of onPower, Australia. S., T. Casey, Clim C. Dyn, Folland, 15, A.319–324, Colman, anddoi:10.1007/ V. Mehta, hyper climate modes. Geophys. Res. Lett., 35, L02706, s003820050284.1999: Inter-decadal modulation of the impact of ENSO doi:10.1029/2007GL031087.
Dommenget, D., 2010: A slab ocean El Nino. Geophys. Res. T. Delworth, and G. Danabasoglu, 2016: Assessing the Lett., 37, L20701, doi:10.1029/2010GL044888. ClimateRuprich-Robert, impacts Y.,of R.the Msadek, observed F. Atlantic Castruccio, Mulitdecadal S. Yeager,
Global Coupled Models. J. Climate, in press. doi:Variability 10.1175/JCLI-D-16-0127.1. using the GFDL CM2.1 and NCAR CESM1 surfaceFan, T., climate C. Deser, variations and D. since P. Schneider,1950. Geophys. 2014: Res. Recent Lett., 41,Antarctic 2419-2426, sea ice doi:10.1002/2014GL059239. trends in the context of Southern Ocean Smith, T. M., R. W. Reynolds, T. C. Peterson, and J.
merged land–ocean surface temperature analysis (1880– 2006).Lawrimore, J. Climate, 2008: 21, Improvements 2283-2296. to NOAA’s historical NorthHoerling, American M., J. Hurrell, decadal A. climate Kumar, forL. Terray, 2011-20. J. Eischeid, J. Climate, P. 24,Pegion, 4519-4528. T. Zhang, X. Quan, and T. Y. Xu, 2011 (August): On doi:10.1175/2011JCLI4137.1 DOI: 10.1175/2007JCLI2100.1
Kay, J. E., C. Deser, A. Phillips, and co-authors, 2015: The Tandon, N.F., and P.J. Kushner, 2015: Does External Community Earth System Model (CESM) Large Ensemble Forcing Interfere with the AMOC’s Project: A community resource for studying climate Influence on North Atlantic Sea Surface Temperature? J. change in the presence of internal climate variability. Climate, 28, 6309-6323. DOI:10.1175/JCLI-D-14-00664.1 Bull. Amer. Met. Soc., 96, 1333–1349, doi: 10.1175/ internal twentieth-century SST in the North Atlantic. J. BAMS-D-13-00255.1. Climate,Ting, M., 22, Kushnir, 1469–1481. Y., Seager, R., Li, C., 2009: Forced and
Ting, M., Kushnir, Y., Li, C., 2014: North Atlantic sector centennial climate variability and recent decadal Latif, M., T. Martin and W. Park, 2013: Southern Ocean internal variability, Journal of Marine Systems, 133, 27- JCLI-D-12-00281.1. 38.Multidecadal SST Oscillation: External forcing versus trends. J. Climate, 26, 7767-7782. DOI: 10.1175/ Latif, M., T. Martin, W. Park, and M. Bordbar, 2015: Internal Trenberth, K. E., and D. J. Shea, 2006: Atlantic hurricanes and natural variability in 2005. Geophys. Res. Lett., 33, Impacts and Implications for Global Warming. In Climate L12704, doi:10.1029/2006GL026894. Change:Southern Multidecadal Ocean Centennial and Beyond, Variability: pp.109-124. Dynamics, Vimont, D. J. 2005: The contribution of the interannual
DOI:10.1142/9789814579933_0007. variability. J. Climate, 18, 2080-2092. ENSO cycle to the spatial pattern of decadal ENSO-like withMantua, Impacts N. J., S.on R. Salmon Hare, Y. Production. Zhang, J. M. Bull. Wallace, Amer. and Meteor. R. C. Woodruff, S. D., and co-authors, 2008: The evolving SST Soc,Francis, 78, 1069–1079.1997: A Pacific Interdecadal Climate Oscillation
record from ICOADS, in Climate Variability and Extremes
5 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 During the Past 100 Years, Adv. Global Change Res., vol. 33, edited by S. Brönnimann et al., pp. 65–83, Springer, New York.
temperature.Wu Z, Huang NE,Clim Wallace Dyn 37:759–773. JM, Smoliak BV, Chen X., 2011: On the time-varying trend in global-mean surface
Wittenberg, 2010: Global warming pattern formation: seaXie, surface S. -P., C.temperature Deser, G. A.and Vecchi, rainfall. J. J. Ma, Climate, H. Teng, 23, 966- A. T. 986, doi:101175/2009JCLI3329.1. multidecadal Atlantic meridional overturning circulation Zhang, L., T. L. Delworth, and F. Zeng, 2016: The impact of 10.1007/s00382-016-3190-8. variations on the Southern Ocean. Clim. Dyn., DOI interdecadal variability: 1900–93. J. Climate, 10, 1004– 1020.Zhang, Y., J. M. Wallace, and D. S. Battisti, 1997: ENSO-like
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 6 Global impacts of the Atlantic Multidecadal Variability during the boreal winter
doi: 10.22498/pages.25.1.7 Yohan Ruprich-Robert1, Rym Msadek2 1 Atmosphere and Ocean Sciences, Princeton University, and NOAA/GFDL, Princeton, New Jersey 2 CNRS-CERFACS, Toulouse, France
Introduction During the last century, the observed annual mean North Atlantic sea surface temperatures (SSTs) exhibited term warming trend. This multidecadal variability has multidecadal fluctuations superimposed onto a long- been referred to as the Atlantic Multidecadal Oscillation anomalous(AMO) or Variabilitypattern that (AMV). has the The same SST sign anomalies in the whole that Northdefine Atlantic, the AMV and are a maximum characterized loading by in athe basin-scale subpolar
Previousgyre (SPG) studies region have (Fig. shown 1). that the AMV is associated with, and possibly the source of, marked climate anomalies over many areas of the globe. This includes droughts over Africa and North America (Mohino et
(Mahajan et al., 2011), changes in Atlantic tropical cycloneal., 2011; activity Enfield (Vimont et al., 2001),and Kossin, decline 2007), in Arctic and recently sea ice it has been linked with the global temperature hiatus to its upstream location, the North Atlantic SST is a main actor(McGregor of the et Europeanal., 2014; Liclimate et al., 2015).variability. Additionally, Sutton anddue the existence of a causal link between the warm phase ofHodson the AMV (2005) and and warmer Sutton conditions and Dong than (2012) normal argue over for Central Europe, drier conditions over the Mediterranean basin, and wetter conditions over Northern Europe during boreal summer. A number of studies suggest also that the AMV could impact the winter North Atlantic – Europe atmospheric circulation by modulating the number of blocking events and/or by driving North Figure 1: (a) Internal (red and blue) versus external (black) components of the observed North Atlantic SST multidecadal Atlantic Oscillation-like anomalies (Hakkinen et al., variability following Ting et al. (2009) definition. (b) Regression 2011; Davini et al., 2015; Peings and Magnusdottir, 2014, map of the observed annual mean SST (ERSSTv3; Smith et al. 2015; Omrani et al., 2014; Gastineau and Frankignoul, 2008) on the internal component of the North Atlantic SST index (i.e., the AMV index); units are oC per standard deviation 2015). Furthermore, the AMV and its Pacific counterpart, linked to multidecadal changes in the frequency of of AMV index. Both SST field and AMV index time series have the Interdecadal Pacific Oscillation (IPO), have been been low pass filtered prior to computing the regression, using a Lanczos filter (21 weights with a 10-yr cutoff period). North American droughts (McCabe et al., 2004; Chylek The black latitudinal lines in b show the subpolar and tropical et al., 2014). However, whether the concomitant forcing domains used for the SPG_AMV and Trop_AMV experiments anomaliesof the Atlantic remains and uncertain. Pacific arise from a coincidence or (see section 2b). Figure from Ruprich-Robert et al. (2017). reveal a causal link between Atlantic and Pacific decadal ©American Meteorological Society. Used with permission.
7 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Given the many potential climate impacts of the AMV at Results decadal timescales, it is crucial to improve our knowledge a) Global impacts of the AMV of the mechanisms associated with AMV teleconnections. A better understanding of these mechanisms could help advance the prediction of AMV impacts and hence 2a-c).During The DJFM, temperature restoring pattern the three of themodels simulated to the anomalies observed decadal climate predictions. We are providing here a AMV yields, as expected, a North Atlantic warming (Fig. short description of a recent coordinated multi-models study that investigates the global impacts of the observed comparedshows some to differences the tropical with anomalies the observed is much one ofless Fig. than 1b. AMV, in which the respective role of the extratropical and theSpecifically, observed the one. relative This comes strength from of our the choice SPG anomalies to keep a time and space invariant restoring timescale for the SST. By so doing, the extratropical North Atlantic SSTs are Descriptiontropical parts of of model the AMV experiments have been identified. weakened due to the SPG deep mixed layers, which dilutes To evaluate the AMV climate impacts, we performed the imposed SST anomaly over a deeper oceanic column. idealized experiments using state-of-the art global coupled climate models, in which the North Atlantic SSTs are restored to time-invariant anomalies corresponding North Atlantic, the three models simulate remarkably to an estimate of the internally driven component of the similarRegardless global of thisteleconnections. weakness, we We find note that a slight outside warming of the observed AMV (following Ting et al. 2009’s approach; of the Indian Ocean and a negative phase of the IPO over Fig. 1). The three models used in this study are the GFDL- the Pacific. The latter has negative SST anomalies in the CM2.1 (Delworth et al., 2006; Wittenberg et al., 2006), inTropical a horseshoe-like Pacific that extendpattern, toward surrounding the higher positive latitudes SST the NCAR CESM1-CAM5 (hereafter CESM1; Kay et al., anomaliesin both Hemispheres in the West. along The three the eastern models oceanshow aboundary, warming 2015), and the GFDL-FLOR (Vecchi et al., 2014). All three of ~0.3°C over Mexico and the Eastern part of US, a models use a nominal 1˚ horizontal ocean resolution but warming over East Brazil as well as over South Asia and employ different atmospheric resolutions. Specifically, the Mediterranean area. The models also agree on the the atmospheric resolution is about 2˚ in CM2.1, 1˚ in simulated warming over Siberia and on the cooling of TwoCESM1, experiments and 0.5˚ in FLOR.were performed with the three the northwestern part of North America. In response to
the Arctic that is only found over the northeastern rim of ofmodels, the AMV namely index Full_AMV+ (i.e., plus and or Full_AMV-,minus the AMVin which pattern SST AMV+ forcing, CESM1 simulates a significant warming of anomalies corresponding to +1 or -1 standard deviation the temperature response over Northern Europe: CM2.1 region, by restoring the model SST to the observed AMV Siberia in CM2.1 and FLOR. The models also disagree on anomaliesshown in plus Fig. the 1b) model’s are imposed own SST in climatology the North Atlanticfrom 0° tend to simulate a cooling. simulates a warming there whereas CESM1 and FLOR let free, allowing a response of the full coupled climate system.to 73°N. Two Outside additional of the restoring sets of experiments region, the similarmodels to were the the atmospheric winter circulation as illustrated by precipitationWe find that and AMV geopotential leads to significantheight at changes 500 hPa in SSTs restored to the observed AMV only in the North 2 . There is a Full_AMV experiments, but with the model North Atlantic northward shift and a reinforcement of the Intertropical (Z500) anomalies (Figs. 2d-f and Fig. 3b) Atlantic subpolar gyre (SPG_AMV) or in the Tropical ensembleNorth Atlantic simulations (Trop_AMV), with were 100 performed members withfor CESM1CM2.1, Convergence Zone all over the tropical belt as well as a and CM2.1. For all experiments, we performed large issouthwestward coherent with shift a La of Niña-like the South temperature Pacific Convergence pattern order to robustly estimate the AMV climate impacts Zone. The precipitation response over the Tropical Pacific and30 members the associated for CESM1, signal-to-noise and 50 members ratio. In for order FLOR to in 3 and found that in all models the capture the potential response and adjustment of other occurrenceseen in Figs. of 2a-b.La Niña We events further roughly analyzed doubles the amplitude between oceanic basins to the AMV anomalies, all the simulations of the ENSO response conditions. In this article, we focus on the boreal winter1 the Full_AMV- and the Full_AMV+ experiments. wereclimate integrated response for to AMV10 years forcing with and fixed we externaldiscuss only forcing the east-westOver the extratropical dipole in the precipitationNorth Pacific, anomalies the AMV leadsover theto a1 weakening of the Aleutian Low (Fig. 3b) associated with an set-upensemble and mean their differences results can between be found the in AMV+Ruprich-Robert and AMV- 1Defined as the December to March seasonal mean. etsimulations. al. (2017) andFurther Castruccio details et regarding al. (in revision). the experimental 2In view of concision, only Z500 response from CESM1 is shown here, but we specify in the following when this response is different among the models. 3To do so, we defined an ENSO index based on the first EOF of the upper 200 m oceanic heat content computed over the tropical Pacific (30°S-30°N).
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 8 Figure 2: Differences between the 10-year average of the Full_AMV+ and the Full_AMV- ensemble simulations for December to March (DJFM) of (a, b, c) 2-meter air temperature and (d, e, f) precipitation. Results are shown from top to bottom for CM2.1, CESM1, and FLOR. Stippling indicates regions that are below the 95% confidence level of statistical significance according to a two- sided t-test. Note that the contours intervals of T2 in a, b, and c have been multiplied by 1.75 compared to Figure 1b.
propagation of oceanic Rossby waves from the central
North Pacific and decrease of precipitation over the west Americacoast of USpattern and Mexico (PNA) (Figs. (Barnston 2d-f). Theand Z500Livezey, anomalies 1987), thePacific SST to cooling the western is driven coast, by explainingan anomalous the advectionwarmer SST of are reminiscent of the negative phase of the Pacific North off Japan. Over the northeastern side of the North Pacific, Low and Mexico and negative anomalies over Canada and with positive Z500 anomalies centered over the Aleutian cool air from the Arctic. Furthermore, this whole North when looking at streamfunction anomalies at 200hPa Pacific response is reminiscent of that documented in the south of Hawaii. The latter center of action is more visible althoughwater hosing the experimentsimpacts are weakerof Zhang in and our Delworth experiments (2005), as expectedDong and from Sutton the weaker (2007) imposed and Okumura forcing. et al. (2009), (hereafter SF200; Fig. 3b).
DelworthThe North (2015). Pacific In SST their response study, they is also linked consistent a northward with among the three models, the North Atlantic – Europe shiftthe Aleutian of the westerlies Low weakening to a northward as discussed shift ofby the Zhang oceanic and (NAE)While theresponse North isPacific notably response weaker. is All significant models simulate and robust an gyre circulation through a Sverdrup balance and to the increase of precipitation over Southern Europe, but these
9 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Figure 3: Difference between the 10-year average of the positive and the negative phases of (a, b) Full_AMV, (c, d) Trop_AMV, (e, f) SPG_AMV for CESM1 in DJFM. (left) 2-meter air temperature (T2m) and (right) geopotential height at 500 hPa (Z500, color) and streamfunction of the wind at 200 hPa (SF200, contours at intervals of 0.8x106 m2 s-1). Stippling indicates regions that are below the 95% confidence level of statistical significance. Figure adapted from Ruprich-Robert et al. (2017). ©American Meteorological Society. Used with permission.
b) Tropical vs Extratropical SST contribution to the AMV climate impacts anomalies are only significant in CESM1. In CESM1 and We investigated the respective contribution of the tropical FLOR, these precipitation anomalies are associated with and extratropical parts of the AMV to the climate impacts geopotentiala weak anomalous anomalies north-south in CM2.1 Z500 do notdipole project that stronglyprojects described in the previous section by performing two on the NAO in its negative phase (hereafter NAO-). The additional sets of experiments in which only the subpolar
NAEonto theatmospheric NAO, even response though positive might anomaliesproject onto are a present mix of over Iceland. For CM2.1 this diagnostic suggests that the experiments(SPG_AMV) or are the shown tropical here, (Trop_AMV) but these experiments parts of the wereAMV Pattern4 alsopattern performed were imposed. with CM2.1 Only theand results we discuss from the CESM1results to-noiseboth an NAO-ratio andof the a negativeclimate responsephase of theto AMV East and,Atlantic for all models(not we shown). found that We the further NAE atmospheric quantified the response signal- PNA-like responses are primarily driven by the tropical accounts for less than 10% of the decadal variance. from both models. We find that the Pacific IPO-like and The discrepancy between the models and the weak the studies of Kucharski et al. (2015) and McGregor atmospheric response over the NAE region suggest that part of the AMV (Figs. 3c,d). This result corroborates the AMV does not strongly impact the atmosphere over cooling observed during the last decades was forced by there. We acknowledge however that our experimental et al. (2014) who suggested that the tropical Pacific protocol may lead to an underestimation of the extratropical AMV forcing and hence potentially to an the tropical Atlantic warming through a modification of underestimation of the atmospheric response over the arethe Walkermainly Circulation.explained by In the line tropical with Sutton part andof the Hodson AMV NAE region. Indeed, as discussed above, our choice to but(2005), that we they find are that reinforced the AMV by impacts the subpolar over the part Americas of the keep a time and spatially-invariant restoring timescale does not allow to strongly constrain the SST over a region 1 with deep oceanic mixed layer such as the SPG. phenomenon4This mode is defined (Figs. in observations 3a,c,e). as the second mode of variability of the atmosphere over the NAE region (e.g., Barnston and Livezey 1987).
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 10 The models show marginal impacts over North Africa of rainfall over Europe. The Walker Circulation response is and Europe in terms of T2m anomalies in response to the tropical AMV anomalies only, whereas a warming of North Africa and a cooling of Europe is simulated in associated with broad Pacific anomalies that project onto response to the SPG anomalies. This cooling is consistent likethe InterdecadalSST response Pacific is tightly Oscillation linked (IPO)to a negative in its negative phase phase. In the three models the northern part of the IPO- relativelywith the Z500 warm dipole ocean anomaly to the seenrelatively over thecool NAE European region, areof themainly Pacific driven North by American the Tropical teleconnection part of the pattern AMV. which tends to decrease the atmospheric flow from the (PNA). We find that both the IPO and PNA-like responses continent in winter. Further, the Z500 dipole response in the SPG_AMV experiment is shifted eastward compared drivingOur results decadal stress changes the importance on a global played scale, byespecially the North in iswith the the primarily NAO response driver of seen the in NAE the atmospheric Full_AMV experiment response Atlantic Ocean variability associated with the AMV in but(Fig. that 3b). both This the suggests tropical that and the the subpolar extratropical part partsof the of AMV the an important role in global climate variability observed AMV contribute to the overall NAE atmospheric response. duringthe Pacific. the Theylast alsocentury. indicate In thatthe thepresent AMV hasstudy, played we
an estimate of the internal component of the observed global atmospheric response in CESM1 than in CM2.1. AMV,specifically which focus has onbeen the shownclimate asimpacts predictable associated to some with The SPG_AMV experiment generates a strikingly larger extent on multi-year to decadal timescale (e.g., Robson thatFor theare former,weaker thebut subpolarsimilar in gyre pattern part to of those the AMV driven leads by results are therefore encouraging for the prospect of theto impacts tropical in part T2m of and the Z500 AMV. over This the is Northconsistent Pacific with region the gettinget al., 2012; skillful Yeager decadal et al., predictions 2012; Msadek over et regions al., 2014). outside Our of the North Atlantic through the impacts of AMV. The teleconnections we highlight between the Atlantic alsoweak suggests but significant that part warming of the tropical simulated signature in the of thetropical AMV isNorth forced Atlantic by the in subpolar the CESM1 part SPG_AMVof the AMV experiment. as suggested This by Chikamoto et al. (2012, 2015), who showed that phase Dunstone et al. (2011) and Smirnov and Vimont (2012) and the Pacific are also consistent with the studies of but that this mechanism is model-dependent. might be predicted few years in advance if the sign and amplitudeshifts of the of the IPO AMV as those are predicted. observed in the late 1990’s Summary and discussion We investigated the climate impacts associated with the The general impacts and mechanisms described in internal component of the observed Atlantic Multidecadal the present study are based on three climate models that show quite similar results despite their different theirVariability North Atlantic (AMV) usingSSTs to the the GFDL-CM2.1,observed anomalies. the NCAR- This robustness of our conclusions regarding AMV impacts. coupledCESM1, and approach the GFDL-FLOR allowed us coupled to determine models, the by full restoring climate atmospheric resolution. This gives confidence in the response to the imposed North Atlantic anomalies. multimodel framework, using other coupled climate However, conducting such experiments within a conclusions. This will be done as part of the CMIP6 that, despite the large-scale warming of the Northern Decadalmodels, Climate will be Prediction highly beneficial Project (DCPP), to strengthen which calls our Over the North Atlantic European (NAE) region, we show for coordinated experiments following a protocol similar to the one proposed in this study (Boer et al., 2016). NorthernHemisphere Europe continents temperature simulated response. in all modelsThey disagree during alsothe borealon the winter NAE (DJFM),atmospheric the models circulation disagree response, on the Acknowledgment which projects on the negative phase of the North The analysis and plots of this paper were performed with the NCAR Command Language (version 6.2.0, 2014), disagreement between the models and the weak signal- Boulder, Colorado (UCAR/NCAR/CISL/VETS, http:// to-noiseAtlantic Oscillationratio of the (NAO)NAE atmospheric for CESM1 response and FLOR. reveal The strong uncertainties on the role played by the AMV in dx.doi.org/10.5065/D6WD3XH5). NCAR is sponsored last century. They also suggest the need to repeat such by the National Science Foundation (NSF). The CESM coordinatedthe decadal experiments variations of with the NAOother observed models. during the is supported by the NSF and the US Department of Energy. This work is supported by the NSF under the theCollaborative Climate Variability Research and EaSM2 Predictability grant OCE-1243015 Program grant to drives a change in the Walker Circulation that drives NCAR and by the NOAA Climate Program Office under precipitationFor the three anomalies models, we over find the that whole the AMVtropical warming belt. The AMV warming leads also to reduced rainfall over the NA13OAR4310138 to NCAR and GFDL. western part of the US and Mexico and to a weak increase
11 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 References December 16, 2013). seasonality and persistence of low-frequency atmospheric circulationBarnston, A.patterns. G., and Mon. R. E. Weather Livezey, Rev., 1987: 115, Classification, 1083–1126, Enfield, D. B., A. M. Mestas‐Nuñez, and P. J. Trimble, 2001: Res.The AtlanticLett., 28, Multidecadal 2077. Oscillation and its relation to doi:10.1175/1520-0493(1987)115<1083:CSAPOL>2.0. rainfall and river flows in the continental U.S. Geophys. Boer,CO;2. G. J., and Coauthors, 2016: The Decadal Climate Prediction Project. Geosci. Model Dev. Discuss., 0, 1–32, the North Atlantic SST Variability on the Atmospheric doi:10.5194/gmd-2016-78. http://www.geosci-model- CirculationGastineau, G.,during and the C. Frankignoul,Twentieth Century. 2015: InfluenceJ. Clim., 26, of dev-discuss.net/gmd-2016-78/. 2969–2972, doi:10.1029/1999GL900613.
R. Msadek, T. Deloworth, 2017: Modulation of Arctic Sea IceCastruccio, Loss by F., Atmospheric Y. Ruprich-Robert, Teleconnections S. Yeager, G. fromDanabasoglu, Atlantic Variability.Hakkinen, S.,Science P. B. (80-. Rhines, )., 334, and 655–659, D. L. Worthen, doi:10.1126/ 2011: Multi-decadal Variability. Geophys. Res. Lett., in revision. science.1205683.Atmospheric Blocking http://www.sciencemag.org/cgi/ and Atlantic Multidecadal Ocean doi/10.1126/science.1205683. Chikamoto, Y., M. Kimoto, M. Watanabe, M. Ishii, and T.
Atlantic stepwise climate change during the 1990s. M. B. Andrews, J. R. Knight, A. A. Scaife, and D. M. Smith, Geophys.Mochizuki, Res. 2012: Lett., Relationship 39, 2–7, doi:10.1029/2012GL053901. between the Pacific and Hermanson, L., R. Eade, N. H. Robinson, N. J. Dunstone, and associated impacts. Geophys. Res. Lett., 41, 5167– Chikamoto, Y., and Coauthors, 2015: Skilful multi-year 5174,2014: Forecastdoi:10.1002/2014GL060420. cooling of the Atlantic http://doi.wiley. subpolar gyre predictions of tropical trans-basin climate variability. Nat. com/10.1002/2014GL060420. Commun., 6, 6869, doi:10.1038/ncomms7869. http:// Kay, J. E., and Coauthors, 2015: The Community Earth System Model (CESM) Large Ensemble Project: A www.nature.com/doifinder/10.1038/ncomms7869. Community Resource for Studying Climate Change 2014: Imprint of the Atlantic multi-decadal oscillation in the Presence of Internal Climate Variability. Bull. Chylek, P., M. K. Dubey, G. Lesins, J. Li, and N. Hengartner, Am. Meteorol. Soc., 96, 1333–1349, doi:10.1175/ climate: past, present, and future. Clim. Dyn., 43, 119– BAMS-D-13-00255.1. http://journals.ametsoc.org/doi/ 129,and Pacificdoi:10.1007/s00382-013-1933-3. decadal oscillation on southwestern http://link. US abs/10.1175/BAMS-D-13-00255.1. springer.com/10.1007/s00382-013-1933-3.
Tropical origin for the impacts of the Atlantic andKucharski, mean state F., F. S.in Syed,the twentieth A. Burhan, century I. Farah, in andobservations A. Gohar, MultidecadalDavini, P., J. Variability Von Hardenberg, on the Euro-Atlantic and S. Corti, climate. 2015: and2015: CMIP5. Tropical Clim. Atlantic Dyn., influence 44, 881–896, on Pacific doi:10.1007/ variability Environ. Res. Lett., 10, 094010, doi:10.1088/1748- s00382-014-2228-z. http://link.springer.com/10.1007/ 9326/10/9/094010. http://stacks.iop. s00382-014-2228-z. e13c7c1e10966cb8fd2a872e288a8724. Li, X., S.-P. Xie, S. T. Gille, and C. Yoo, 2015: Atlantic-induced org/1748-9326/10/i=9/a=094010?key=crossref. pan-tropical climate change over the past three decades. Nat. Clim. Chang., 1–13, doi:10.1038/nclimate2840. andDelworth, simulation T. L., characteristics. and Coauthors, J. Clim., 2006: 19, GFDL’s 643–674, CM2 nclimate2840. doi:10.1175/JCLI3629.1.global coupled climate models. Part I: Formulation http://www.nature.com/doifinder/10.1038/
Variability by a Weakened Atlantic Thermohaline Mahajan, S., R. Zhang, and T. L. Delworth, 2011: CirculationDong, B., and in R.a CoupledT. Sutton, GCM. 2007: J. EnhancementClim., 20, 4920–4939, of ENSO andImpact Sea ofIce theVariability. Atlantic J. MeridionalClim., 24, Overturning6573–6581, doi:10.1175/JCLI4284.1. http://journals.ametsoc.org/ doi:10.1175/2011JCLI4002.1.Circulation (AMOC) on Arctic Surfacehttp://journals.ametsoc. Air Temperature doi/abs/10.1175/JCLI4284.1. org/doi/abs/10.1175/2011JCLI4002.1.
Dunstone, N. J., D. M. Smith, and R. Eade, 2011: Multi-year predictability of the tropical Atlantic atmosphere driven frequencyMcCabe, G. inJ., M.the a Palecki,United States.and J. L. Proc. Betancourt, Natl. Acad. 2004: Sci. Pacific U. S. Lett., 38, n/a – n/a, doi:10.1029/2011GL047949. http:// A.,and 101, Atlantic 4136–4141, Ocean influencesdoi:10.1073/pnas.0306738101. on multidecadal drought doi.wiley.com/10.1029/2011GL047949(Accessedby the high latitude North Atlantic Ocean. Geophys. Res.
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 12 Science (80-. )., 309, 115–118. papers2://publication/ doi/10.1126/science.1109496. McGregor, S., A. Timmermann, M. F. Stuecker, M. H. England, M. Merrifield, F.-F. Jin, and Y. Chikamoto, 2014: Chang.,Recent 4, Walker 888–892, circulation doi:10.1038/nclimate2330. strengthening and http:// Pacific to Basin-Scale Warming and Cooling of the North Atlantic www.nature.com/nclimate/journal/v4/n10/full/cooling amplified by Atlantic warming. Nat. Clim. Sutton, R. T., and D. L. R. Hodson, 2007: Climate Response http://journals.ametsoc.org/doi/abs/10.1175/ JCLI4038.1.Ocean. J. Clim., 20, 891–907, doi:10.1175/JCLI4038.1. Mohino,nclimate2330.html?WT.ec_id=NCLIMATE-201410. E., S. Janicot, and J. Bader, 2011: Sahel rainfall and decadal to multi-decadal sea surface temperature variability. Clim. Dyn., 37, 419–440, doi:10.1007/s00382- on a shift in European climate in the 1990s. Nat. Geosci., 5, 010-0867-2. http://link.springer.com/10.1007/s00382- 788–792,Sutton, R. doi:10.1038/ngeo.T., and B. Dong, 2012: http://dx.doi.org/10.1038/ Atlantic Ocean influence 010-0867-2. ngeo 1595.
Msadek, R., and Coauthors, 2014: Predicting a Decadal Ting, M., Y. Kushnir, R. Seager, and C. Li, 2009: Shift in North Atlantic Climate Variability Using the in the North Atlantic*. J. Clim., 22, 1469–1481, doi:10.1175/JCLI-D-13-00476.1.http://journals. doi:10.1175/2008JCLI2561.1.Forced and Internal Twentieth-Century http://journals.ametsoc. SST Trends ametsoc.org/doi/abs/10.1175/JCLI-D-13-00476.1.GFDL Forecast System. J. Clim., 140707141053007, org/doi/abs/10.1175/2008JCLI2561.1(Accessed December 16, 2013).
Okumura, Y. M., C. Deser, A. Hu, A. Timmermann, and S.-P. andXie, 2009:Atmospheric North Pacific Pathways. Climate J. ResponseClim., 22, to 1424–1445, Freshwater Clim.,Vecchi, 27, G. 7994–8016, A., and Coauthors, doi:10.1175/JCLI-D-14-00158.1. 2014: On the Seasonal doi:10.1175/2008JCLI2511.1.Forcing in the Subarctic Northhttp://journals.ametsoc. Atlantic: Oceanic http://journals.ametsoc.org/doi/abs/10.1175/Forecasting of Regional Tropical Cyclone Activity. J. org/doi/abs/10.1175/2008JCLI2511.1(Accessed JCLI-D-14-00158.1. December 16, 2013). Vimont, D. J., and J. P. Kossin, 2007: The Atlantic Meridional Mode and hurricane activity. Geophys. Res. Lett., 34, 2014: Stratosphere key for wintertime atmospheric L07709, doi:10.1029/2007GL029683. http://doi.wiley. responseOmrani, N. to E.,warm N. S. Atlantic Keenlyside, decadal J. Bader, conditions. and E. Clim. Manzini, Dyn., com/10.1029/2007GL029683 (Accessed December 16, 42, 649–663, doi:10.1007/s00382-013-1860-3. 2013).
Wittenberg, A. T., A. Rosati, N. C. Lau, and J. J. Ploshay, T. Delworth, and G. Danabasoglu, 2017: Assessing the Ruprich-Robert, Y., R. Msadek, F. Castruccio, S. Yeager, 722,2006: doi:10.1175/JCLI3631.1. GFDL’s CM2 global coupled climate models. Part GlobalClimate Coupled Impacts ofModels. the Observed J. Clim., Atlantic 30, Multidecadal2785–2810, III: Tropical Pacific climate and ENSO. J. Clim., 19, 698– doi:10.1175/JCLI-D-16-0127.1.Variability Using the GFDL CM2.1 http://journals.ametsoc. and NCAR CESM1 Yeager, S., A. Karspeck, G. Danabasoglu, J. Tribbia, and org/doi/10.1175/JCLI-D-16-0127.1.
Schubert, S., and Coauthors, 2009: A US CLIVAR J.H. Clim., Teng, 25, 2012: 5173–5189, A Decadal doi:10.1175/JCLI-D-11-00595.1. Prediction Case Study: Late Project to Assess and Compare the Responses of http://journals.ametsoc.org/doi/abs/10.1175/Twentieth-Century North Atlantic Ocean Heat Content. JCLI-D-11-00595.1 (Accessed December 12, 2013).
Global Climate Models to Drought-Related SST Forcing Patterns: Overview and Results. J. Clim., 22, 5251– doi/pdf/10.1175/2009JCLI3060.1.5272, doi:10.1175/2009jcli3060.1.
JCLI-D-11-00104.1.Forcing of Tropical Atlantichttp://journals.ametsoc.org/doi/ Variability during Boreal abs/10.1175/JCLI-D-11-00104.1.Summer and Fall. J. Clim., 25, 2056–2076, doi:10.1175/ Response to a Substantial Weakening of the Atlantic ThermohalineZhang, R., and Circulation.T. L. Delworth, J. Clim., 2005: 18, Simulated 1853–1860. Tropical forcing of North American and European summer climate. Sutton, R. T., and D. L. R. Hodson, 2005: North Atlantic
13 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Arctic sea ice seasonal-to-decadal variability and long-term change
doi: 10.22498/pages.25.1.14 Dirk Notz Max Planck Institute for Meteorology, Hamburg, Germany
Introduction The large-scale loss of Arctic sea ice in recent decades is that determines the regional evolution of Arctic sea ice one of the most prominent indicators of the ongoing global (e.g., Bushuk et al., 2015). climate change. This derives from three main reasons. The underlying research is in part driven by very practical applications, such as ship routing, but will also ofFirst, changes climate in the change global-mean is amplified climate in the are Arctic more (“Arcticreadily increase our fundamental understanding of air–ice–sea observedamplification”, at high e.g. Pithanlatitudes et al. compared (2013)), soto consequences middle and interactions at high latitudes. Related activities are for lower latitudes. Second, while many observables change example coordinated by the Polar Prediction Project gradually with global mean climate, Arctic sea ice is among those observables that might eventually cross a binary threshold from “existing” to “non-existing”, which Worldwith its Weather flagship Research Year of PolarProgramm Prediction WWRP), 2017-2019 by the Polar(www.polarprediction.net, Climate Predictability under Initiative the auspices (www.climate- of WMO And third, as a consequence, changes in “Arctic sea ice cryosphere.org/wcrp/pcpi, under the auspices of the coverage”amplifies the are perception easier to ofgrasp the underlying and communicate gradual trend. to a general public than changes in more abstract metrics such as “global mean temperature”. activityWMO World (www.arcus.org/sipn). Climate Research Program WCRP), and by the Sea ice Prediction Network with its Sea ice Outlook The observed changes in Arctic sea ice are not only a clear In this latter activity, various research groups try to forecast the minimum Arctic sea ice area coverage in also has a number of sometimes far-reaching worldwide September based on the observed state of the sea ice cover consequences.local indicator ofThese large-scale include climate physical change; phenomena the ice such loss from May onwards. The groups use a variety of methods, as the possible impact on mid-latitude weather system ranging from heuristic methods to seasonal prediction
2, but systems based on coupled climate models. An analysis of also societal consequences such as the opening of new the forecast quality of the various methods has shown a orshipping the disruption routes or the of necessary the oceanic changes uptake in the of lifelihood CO of the Arctic indigenous population. better results than any other (Stroeve et al., 2014). mixed picture, with no single method giving significantly The importance of sea ice loss both as an indicator and as an active player in the ongoing climate change has studies with seasonal prediction systems usually result motivated some intense research into understanding the This finding is possibly surprising in that idealised temporal evolution of sea ice on time scales from seasons to years in advance (Blanchard-Wrigglesworth et al., 2016). decades. In this contribution, I use a combined analysis of Inin significantthis framework prediction referred skill to of as many “perfect months model”, up toa single a few the observational record and of climate model simulations model simulation is taken as the “observed truth”, while additional model simulations with slightly perturbed initial conditions are used to estimate whether this Seasonalto explain variability: and summarize The importance some of of these atmospheric findings. “truth” can be forecast. chaos Recent years have seen an increase in research activities The striking difference in the forecast quality of such aimed at forecasting the evolution of Arctic sea ice on time idealised studies compared to those trying to forecast scales of a few months. There is good reason to believe the real world might be explicable by three main factors. that such seasonal-scale forecasts should be possible, with model studies emphasizing in particular the rather knowledge of the initial state of the system. As shown long memory of the sea ice state imprinted in the sea ice byFirst, Bunzel the forecastet al. (2016), skill stronglythe incomplete depends knowledge on a proper of
2011), and the long memory of the oceanic heat content between different satellite data sets can cause differences thickness fields (e.g., Blanchard-Wrigglesworth et al., May sea ice concentration as reflected by the differences
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 14 Figure 1: (a) Evolution of modeled and observed Arctic sea ice area in May and July (1979–2016). (b) Modeled Arctic sea ice area in July as a function of observed Arctic sea ice area in May of the same year. (c) Observed Arctic sea ice area in July as a function of observed Arctic sea ice area in May of the same year. Observations for all panels are based on the Arctic sea ice index (Fetterer et al., 2002, updated 2016). Model simulations for all panels are based on the first ensemble member of the CMIP5 simulations of MPI-ESM-LR (Notz et al., 2013). in forecast mean September Arctic temperature of world application. Both are absent in a perfect-model several degrees, and in forecast September Arctic sea study, as its forecasts are usually started from a model ice area of 2 million km2. The incomplete knowledge of state that is perfectly consistent with model physics. In initial conditions of the sea ice cover arises because the contrast, the initial state for any simulation starting from microwave signature that is usually used to assess ice a model state based on observations will usually be more concentration reacts sensitively to snow coverage and or less inconsistent with model physics, possibly causing melt-pond formation on the ice, for example. Different substantial drift that can quickly compensate for any algorithms compensate these uncertainties in different added value from the assimilation of observations. ways, causing substantial differences of the observed sea ice area. Because of positive feedbacks such as A third factor that might cause the systematically better the ice-albedo feedback, these differences in initial forecast skill in idealised model studies relates to possible model errors in the simulation of the persistence of the and contribute to the much lower forecast quality of real- Arctic sea ice cover. Take, for example, the relationship worldconditions applications are substantially compared amplifiedto perfect during model summerstudies. between Arctic sea ice area in May and Arctic sea ice Incomplete knowledge of the state of the underlying area in July during the observational period 1979–2016 ocean certainly also contributes to these uncertainties. correlated because of their underlying trend, both in the Second, the forecast skill might be negatively affected by observations(Fig. 1a). The andtime in series the model of these simulations. two months are highly the initial shock and drift in the forecast runs in a real-
15 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 However, the detrended time series are only significantly acorrelated chance of in far the less model than simulations1 % that these (Fig. time 1b), series with theare uncorrelated.Pearson rank In coefficient contrast, ofchances the correlations are above 30 indicating % that observed time series of sea ice area in May and the detrendedthere is no significantobserved timecorrelation series betweenof sea ice the area detrended in July
This(Fig. 1c).suggests that at least an idealised study based on the particular model employed here (MPI-ESM-LR) will result in an unrealistically large potential forecast skill of seasonal predictions. We are currently examining in particular trying to investigate the underlying reasons forwhether this differentthis finding behaviour also holds in forthe other model models, compared and are to reality.
Annual variability: The importance of negative feedbacks In addition to seasonal forecasts on time scales of a few months, also forecasts on time scales of a few years have made some headlines over the past decade. These headlines were usually related to claims that the Arctic would lose its remaining summer sea ice within just a few years. The underlying reasoning of such claims was often related to a discussion of a possible ’tipping point’ that is related to the ice-albedo feedback. Given the substantial loss of Arctic sea ice in the past few years, the ocean could potentially absorb enough heat to rapidly melt the remainder of the sea ice cover. system strongly suggests that this reasoning is unrealistic. However, our current understanding of the Arctic climate experiments in which all Arctic sea ice was synthetically A first indication for this finding derived from model thus maximising the possible ice-albedo feedback (Tietscheremoved fromet al., the2011). Arctic Despite Ocean such at the maximised onset of feedback, summer, the ice cover recovered in these experiments within just a few years. This is because on annual time scales, negative feedbacks dominate the evolution of the Arctic sea ice cover. Three negative feedbacks are particularly its heat to the atmosphere during winter, causing a rapid lossimportant: of much First, of the the heat open that ocean was very accumulated effectively in releases the ice- free water during summer. Second, the thin ice that forms Figure 2: (a) Evolution of modeled and observed Arctic sea during winter can grow much more rapidly than ice that ice area in September (1850–2016). (b) Evolution of ten-year survived the summer, because heat can more effectively linear trends of Arctic sea ice area, plotted at the end point be transported from the ocean to the atmosphere when of the ten year averaging period. (c) Evolution of thirty-year the ice cover is thin (Bitz and Roe, 2004). Third, as ice linear trends of Arctic sea ice area, plotted at the end point of the thirty year averaging period. (d) Evolution of modeled forms later in the season, it will carry a thinner insolating and observed Arctic sea ice area in September (1850–2100) snow cover as any snow fall occurring before ice as a function of cumulative anthropogenic CO2 emissions. formation simply falls into the open ocean (Notz, 2009). Observations for all panels are based on the Arctic sea ice index (Fetterer et al., 2002, updated 2016). Model simulations The effectiveness of these negative feedbacks on an annual for all panels are based on the 100 member ensemble of MPI- ESM-LR. time scale is not only apparent in our model simulations;
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 16 the observed time series of Arctic summer sea ice also observed sea ice area that occured in the year 2007 has carries a clear signature of such mechanisms. A year with been an extreme event compatible with internal climate a strong drop in ice coverage during September is usually variability and therefore cannot directly be compared followed by an increase in September ice coverage to the ensemble mean across several models or to the in the following year. More formally, the time series ensemble mean across multiple simulations with one model. If this characterisation of this sea ice loss as an Marotzke, 2012). If indeed the ice albedo feedback was as extreme event was correct, any realistic climate model effectiveshows significant on annual negativetime scales autocorrelation as implied by statements (Notz and should on average simulate a slower ice loss than has supporting the entire loss of Arctic summer sea ice been observed (see also Notz (2015) for a detailed within this decade, one would certainly expect that any discussion). year with a strong drop in ice coverage would be followed by a year with yet another drop. This is found neither in Regarding the future evolution of sea ice, the model the observational record, nor in model simulation. This simulations with MPI-ESM-LR suggest a similar range of underpins the dominance of negative feedbacks, which possible ten year trends than over the past few decades. stabilize the Arctic ice cover and prevent a possible “tipping point”. that the level of internal climate variability simulated is correct,Hence, inthe extreme sea ice cover cases might and providing in the future the hypothesispotentially Decadal variability: The importance of internal variability this century, or, in contrast, gain an average of 100 000 Internal climate variability not only governs a substantial kmonce2 peragain year lose for ice a asdecade fast as despite during thethe ongoingfirst decade global of part of the sea ice evolution on seasonal-to-interannual warming (see also Swart et al., 2015). time scales as discussed above, but also affects the longer exemplify this, large ensembles of simulations of coupled evolution of Arctic sea ice, the underlying reasons for Earthterm trends System of Modelssea ice (Swartare particularly et al., 2015; helpful Notz, 2015).(Swart Toet internalIn order variability to more must confidently be understood predict better. the near-termA number al., 2015). At the Max-Planck-Institute for Meteorolgy, of recent studies point in particular to the impact of oceanic heat transport into the Arctic for driving low- of simulations with our Earth System Model MPI-ESM- frequency variability of the ice cover, including a possible LRwe forhave the recently historical finished period, a 100 and member for two largepossible ensemble future contribution to the recent acceleration of sea ice loss emission scenarios, RCP2.6 and RCP4.5. In comparing (e.g., Årthun et al., 2012). These studies emphasize the simulated Area sea ice area during September with that a scenario with a much slower sea ice loss for the next decade is plausible if oceanic heat transport were the observed sea ice area is at the upper edge of the ensemblethe observational spread during record the 1979–2016,earlier years weof the find record, that weakening of the oceanic heat transport would not and approaches the mean of the ensemble in the more onlyto weaken affect the (Zhang, sea ice 2015; cover Yeager itself, etbut al., also 2015). its future Such predictability on seasonal time scales (Germe et al., 2014). This then directly links the challenge of decadal Ten-yearrecent past long (Fig. trends 2a). within the observational record forecasting of sea ice to that of its seasonal forecasting as have consistenly been negative with values ranging from described in the previous section. a mean loss of 23 000 km2 per year during the period 1990–1999 to a mean loss of 180 000 km2 during the Long-term changes: The importance of the external forcing period includes the two record minima that have been period 2003–2012 (black line in Fig. 2b). This latter a substantial contributor to the evolution of Arctic sea 2007-2016, the average ice loss has been around iceOn area.longer The time 100 scales, member internal simulations variabililty suggest also a possibleremains 50observed 000 km in2 per2007 year. and All 2012. these For numbers the most are recent well periodwithin spread in observed September sea ice area of around 2 the range simulated by individual ensemble members million km2 of MPI-ESM-LR, which show over ten year-long even long-term trends over 30 years show substantial periods a sea ice evolution ranging from a mean loss of for any given year (Fig. 2a). In terms of trends, around 200 000 km2 per year to a mean gain of around over the past 30 years, sea ice area in the Arctic could 100 000 km2 havevariability remained (Fig. roughly2c). For constantexample, or the could model have suggests decreased that particular, it is noteworthy that the mean ten-year long roughly as quickly as observed. This large spread in 30- per year (shaded range in Fig. 2b). In year long trends again suggests that using these trends the most recent past are very close to the mean trend of as metrics for the purpose of model evaluation can be thetrends 100 for member the first ensemble, 20 years ofincluding the satellite a slowdown record and of thefor misleading (Notz, 2015), in particular if the observed ice loss during the 1990s and an accelaration during the evolution of Arctic sea ice corresponds to a possible early 2000s. This agreement during substantial periods extreme event. of the record suggests that the rather sudden drop in
17 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Despite the large impact of internal variability, the dominant role of external forcing in the observed • evolution of Arctic sea ice is clear (Notz and Marotzke, the two sea months ice cover. in advance There isinherently no “tipping difficult. point” beyond 2012). The weight of the external forcing becomes whichOn annual the timeloss scales,of the negativeremaining feedbacks summer stabilizesea ice particularly apparent if one examines the average becomes unstoppable evolution of Arctic sea ice coverage in the 100 member •
2 can cause a substantial acceleration or temporary recoveryOn decadal of timethe scales,sea ice internal cover climatethat renders variability the iceensemble cover asdirectly a function follows of cumulativethe cumulative anthropogenic emissions. CO In evaluation of individual model simulations based on particular,emissions (Fig.there 2d). is no The clear long-term difference evolution in mean of thesea seaice their short-term trends impossible. coverage between RCP 2.6 and RCP 4.5 for any given •
2 emission. substantial spread in possible 30-year long trends supportingOn longer timefor scales,the production internal variability of large causes model a cumulativeWe have recently CO been able to explain this relationship, ensembles. Nevertheless, the impact of anthropognic which is largely based on the fact that the position of forcing on the long-term sea ice evolution is clear, the outer edge of the sea ice cover is determined by the with an average loss of 3 m2 of September sea ice net local energy balance (Notz and Stroeve, 2016). Any 2 emission.
2 concentration increases the incoming longwave radiation at the ice edge, causing the Referencescover per metric ton of anthropogenic CO latterrise in to atmosphericmove northward CO to a region with less incoming combined processes lead to a roughly linear relationship heatÅrthun, on BarentsM., T. Eldevik, Sea ice L. variability H. Smedsrud, and retreat,Ø. Skagseth, J. Clim., and 25, R. shortwave radiation. For geometric reasons, these B. Ingvaldsen, 2012: Quantifying the influence of Atlantic 2 4736–4743, doi:10.1175/JCLI-D-11-00466.1. emissions in the obervational record and in all CMIP5 modelbetween simulations. Arctic sea In icethe obervations, loss and anthropogenic about 3 m2 of COsea Bitz, C., and G. Roe, 2004: A mechanism for the high rate
2 emissions, while the models on 2 average only simulate an ice loss of 1.7 m 2 3623–3632. iceemissions. are lost per ton of CO of sea ice thinning in the Arctic Ocean. J. Clim., 17 (18), per ton of CO Blanchard-Wrigglesworth, E., K. C. Armour, C. M. Bitz,
2 and E. DeWeaver, 2011: Persistence and inherent emissions and Arctic sea ice area is roughly linear strongly predictability of Arctic sea ice in a GCM ensemble and The fact that the relationship between cumulative CO 2 emissions for the observations, J. Climate, 24, 231–250. suggestsare also apparent,a dominating in particular role of the in CO the mean across all Blanchard-Wrigglesworth, E., and Coauthors, 2016: simulationsevolution of ofsea our ice 100-member area. However, ensemble. other external Most striking drivers Multi-model seasonal forecast of Arctic sea ice: forecast are temporary increases in Arctic sea ice area following uncertainty at pan-Arctic and regional scales. Clim. Dyn., large volcanic eruptions during the historical period, most 1–12, 10.1007/s00382-016-3388-9. recently in 1991 after the Pinatubo eruption, in 1982 after the eruption of El Chichon and in 1963 after the eruption Bushuk, M., D. Giannakis, and A. J. Majda, 2015: Arctic sea ice reemergence: The role of large-scale oceanic and Because of the large internal variability and the relatively atmospheric variability, J. Climate, 28, 5477–5509. short-livedof Mount Agung response, (compare these also eruptions Zanchettin are et impossible al., 2014). to identify in the temporal evolution of individual simulations nor the observational record, but they apparently have contributed to a synthetic improvement affectedBunzel, by F., observational D. Notz, J. Baehr, uncertainty W. A. of Müller, Arctic andsea ice K. of CMIP5 sea ice simulations relative to CMIP3 sea concentration.Fröhlich, 2016: Geophys. Seasonal Res. climate Lett., forecasts 43 (2), significantly2015GL066 ice simulations (Rosenblum and Eisenman, 2016). 928, 10.1002/2015GL066928.
Conclusions This short overview presents some recent work on the updated 2016: Sea ice index. Digital media, National variability and long-term evolution of Arctic sea ice area. SnowFetterer, and F., Ice K. Data Knowles, Center, W. Boulder, Meier, Coloradoand M. Savoie, USA. 2002, sea ice coverage as this is the month with the strongest Germe, A., M. Chevallier, D. S. y Mélia, E. Sanchez Gomez, observedFor space trends.constraints, The discussionthe focus was can onlybe summarized on September as and C. Cassou,2014, Interannual predictability of Arctic follows: sea ice in a global climate model: Regional contrasts and • temporal evolution, Clim. Dyn., 43(9-10), 2519–2538. variability and its imprint on sea ice renders skillful predictionsOn seasonal of September time scales, sea ice atmospheric coverage more internal than Notz, D., 2009: The future of ice sheets and sea ice:
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 18 Between reversible retreat and unstoppable loss. Proc. U.S.A., 112, 4570–4575, doi:10.1073/pnas.1422296112. Nat. Ac. Sci., 106 (49), 20 590–20 595, doi:10.1073/ pnas.0902356 106.
164,Notz, 10.1098/rsta.2014.0164. D., 2015: How well must climate models agree with observations? Phil. Trans. R. Soc. A, 373 (2052), 20140
Marotzke, 2013: Arctic sea ice evolution as modeled by MPI-ESM.Notz, D., J. A. Adv. Haumann, Model. Earth H. Haak,Syst., 5, J. 173–194, Jungclaus, 10.1002/ and J. jame.20016. external driver for arctic sea ice retreat. Geophys. Res. Lett.,Notz, 39 D., (8), and L051 J. Marotzke,094, 10.1029/2012GL051094. 2012: Observations reveal
2 emission. Science, aag2345,Notz, D., and 10.1126/science.aag2345. J. Stroeve, 2016: Observed Arctic sea ice loss directly follows anthropogenic CO phase clouds cause climate model biases in Arctic wintertimePithan, F., B. temperature Medeiros, andinversions. T. Mauritsen, Clim Dyn, 2013: 43 Mixed- (1-2), 289–303, 10.1007/s00382-013-1964-9. ice Retreat in CMIP5 than in CMIP3 due to Volcanoes. J. Climate,Rosenblum, 29 (24), E., and 9179–9188, I. Eisenman, 10.1175/JCLI-D-16-0391.1. 2016: Faster Arctic Sea
Wrigglesworth, 2014: Predicting September sea ice: Stroeve, J., L. C. Hamilton, C. M. Bitz, and E. Blanchard- 2013. Geophys. Res. Lett., 41 (7), 2014GL059 388, 10.1002/2014GL059388.Ensemble skill of the SEARCH Sea ice Outlook 2008-
trends.Swart, N.Nature C., J. C. Clim. Fyfe, Change,E. Hawkins, 5 (2), J. E. 86–89,Kay, and 10.1038/ A. Jahn, nclimate2483.2015: Influence of internal variability on Arctic sea ice
Recovery mechanisms of Arctic summer sea ice. Geophys. Res.Tietsche, Lett., S., 38 D. (L02707), Notz, J. H. 10.1029/2010GL045698.Jungclaus, and J. Marotzke, 2011:
Yeager, S. G., A. R. Karspeck, and G. Danabasoglu, 2015: Predicted slowdown in the rate of Atlantic sea ice loss, Geophys. Res. Lett., 42, doi:10.1002/2015GL065364.
Inter-hemisphericZanchettin, D., O. asymmetry Bothe, C. in Timmreck, the sea ice J.response Bader, to A. Beitsch, H. F. Graf, D. Notz, and J. H. Jungclaus, 2014: Earth Syst. Dynam., 5, 223–242. volcanic forcing simulated by MPI-ESM (COSMOS-Mill). of summer Arctic sea ice extent, Proc. Natl. Acad. Sci. Zhang, R., 2015: Mechanisms for low-frequency variability
19 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Decadal climate variability and the global energy balance doi: 10.22498/pages.25.1.20 Richard P. Allan Department of Meteorology, University of Reading, UK
Introduction This trend is punctuated by episodic warm El Niño events The Earth’s energy balance represents a nexus between (e.g. 1997/98, 2015/16) and cool La Niña episodes (e.g radiative forcings which set the trajectory of climate 1999, 2011) which alter global monthly mean surface change and feedbacks which determine the nature and temperature by up to around 0.2-0.3 K. At longer timescale, magnitude of the response. Yet entwined within the increased La Niña frequency linked to multi-decadal observed decadal variability and trends are complex, unforced interactions within the climate system. The 2013) suppressed decadal surface warming rates during energy and water cycles are intimately linked and 2000-2010strengthening (Xie of the& Kosaka,Walker circulation2017). The (L’Heureux opposite et was al., observed precipitation changes contain signals from observed during 1980-1990 characterized by recurrent and strong El Nino events, which boosted the warming to radiative forcing and responses to the longer-term trend. The Ts anomalies are well represented by AMIP unforced fluctuations as well as rapid adjustments dominate the effective heat capacity of the climate system.heating Itor is cooling; essential this to monitoris mediated key byindicators the oceans of climate which including the evolving energy budget to interpret global change in the context of intrinsic multidecadal variability. surface temperature, atmospheric moisture, precipitation andFig. the 1 displaystop of atmosphere variability energy and changebalance inover global-mean the period 1979-2016. This includes a mixture of observationally- based estimates combined with the European Centre reanalysis (ERAI) (Dee et al., 2011), which continually adjustsfor Medium-range a numerical Weather model by Forecasts applying (ECMWF) data assimilation interim to an evolving and diverse set of global observations. Also shown are atmosphere-only “AMIP” experiments from phase 5 of the Coupled Model Intercomparison Project (CMIP5) (Taylor et al., 2011) which apply realistic radiative forcing and observed surface temperature and sea ice distributions over the 1979-2008 period (an ensemble mean with a one standard deviation spread across models are displayed). The surface temperature, ofwater atmosphere vapor and energy precipitation budget estimatesvariability from depicted Allan in et Fig. al. (2014b).1a-c update Allan et al. (2014a) while Fig. 1d exploits top Figure 1: Deseasonalised monthly anomalies with respect Considering deseasonalised monthly surface temperature to 1995-2000 in global mean (a) surface temperature, (b) column integrated or surface water vapor, (c) precipitation 2012), the globe has warmed at 0.16 K/decade when and (d) top of atmosphere net radiation for a combination considering(Ts) anomalies the period from HadCRUTv4.51988-2015 chosen (Morice to coincide et al., of satellite and surface observationally-based estimates, with the introduction of the special sensor microwave atmosphere-only climate models using prescribed observed imager (SSM/I) series of satellite instruments in 1987. sea surface temperature and sea ice (AMIP) and a the ERAI reanalysis over the period 1979-2016 (3 month smoothing is applied).
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 20 simulations (which prescribe observed sea surface Low altitude moisture provides the fuel for rainfall temperatures while land Ts is explicitly simulated) and events (Trenberth et al., 2003) yet global precipitation ERAI (which also prescribes ocean surface temperature is determined by atmospheric energy balance, primarily but land Ts is somewhat constrained by observations attributable to net radiative energy loss (Allen and through data assimilation). ERAI anomalies up to 0.2 K global driving factors, combined with the heterogeneous to the lack of interpolation of observed values over the distributionIngram, 2002; of precipitation O’Gorman and et al.,associated 2012). measurement Given these Arctichigher (Cowtan than HadCRUT4 and Way, in 2014). late 2016 are likely due in part limitations, it is no surprise that variability and trends in
Atmospheric column integrated water vapour closely tracks the temperature changes, as expected from the precipitationglobal precipitation from the (Fig. Global 1c) contrastPrecipitation markedly Climatology to that strong temperature dependence of saturation vapor of water vapor and temperature (Fig. 1a-b). Global mean pressure determined by the Clausius Clapeyron equation, satellite-based and land surface gauge-based estimates, and there is broad agreement between the range of appearsProject (GPCP to display v2.3; greaterAdler et month al., 2017), to month a combination variability of surface in situ observations, satellite-based datasets and compared to longer time-scale changes than temperature AMIP simulations. The satellite estimates sample the ice- free ocean (a combination of microwave measurements similarity to AMIP5 simulations appears less coherent or water vapor; co-variability with these variables and
2013)taken and from are the here F08/F11/F13/F17 combined with ERAI series over of remaining Defence (Fig. 1c) with barely significant global precipitation regions:Meteorological these indicate Satellite a moistening Program of satellites; 1.2 %/decade Wentz, for warmingtrends during 2000-2012, 1988-2015 consistent (0.3 %/decade; with understanding r=0.19) and ofno radiativesignificant forcing trend duringand precipitation the period ofresponse slower surface (Allan suppressed (by about -0.2 %/decade) during the 2000- 20121988-2015; period interestinglyof slower surface this warming. trend is only marginally surprise that SSM/I-based estimates agree with GPCP sinceet al., SSM/I 2014a; data Saltzmann, over the 2016). ice-free It ocean is reassuring is used butin the no The resultant additional water vapor continuum generation of GPCP estimates while over other regions absorption in the more transparent window regions of the infrared spectrum cause a reduction in surface loss changing observing system seriously compromises of clear-sky longwave radiation of ~1.4 Wm-2 per mm of thedata global is identical precipitation in this mergedvariability estimate. depicted However, by ERAI the precipitable water vapor (Allan, 2009) which translates to reduced clear-sky surface net longwave radiative loss 2014a). Interannual coupling of GPCP precipitation with of ~0.4 Wm-2 per decade, consistent with more detailed as previously reported (Dee et al., 2011; Allan et al., modelling estimates (Wild et al., 2008). Enhanced consistent with estimates of temperature dependent absorption of sunlight by the increasing water vapor HadCRUT Ts over this period is 3.0±0.7 %/K (r=0.37), additionally reduces net radiative energy loss by the al., 2017). Global precipitation increases with global Ts atmosphere and contributes to solar “dimming” at the primarilyprecipitation due sensitivity to the enhanced (Andrews radiative et al., 2010;loss for Myhre higher et surface and atmospheric temperatures, set by the thermodynamics of the coupled system (Roderick et al. surface (Haywood et al., 2011). increases with Ts at 7.2±0.4 %/K based on linear additional absorption of sunlight by higher water vapor regressionObserved (r global = 0.87), mean in agreement column with integrated the combination moisture 2014; Myhre et al., 2017) although this is tempered by the of Ts and moisture trends. This is consistent with simple thermodynamics which strongly determine global low 2-3%loadings in global (Allan precipitation 2009) and during modified warm by El sensible Niño events heat altitude water vapor although variability and change over (e.g.flux changes.2010 and However, 2016) coincideapparent withshort-term increases increases in Ts of land appears less constrained (Simmons et al. 2010). just 0.2-0.3 K, a much greater precipitation sensitivity Climate model AMIP simulations capture the SSM/I- than anticipated from energy budget considerations and based variability and earlier Scanning Multi-channel Microwave Radiometer (SMMR)-based microwave of circulation systems and energy in the climate system. estimates while ERAI anomalies are in close agreement indicative of a subtler influence of spatial reorganisation after the 1991-1993 period during which an unrealistic Variability in net downward top of atmosphere radiation drop in global ocean moisture affects the reanalysis. imbalance by satellites are generally well captured by There is a remarkable agreement in interannual the AMIP5 simulations (Allan et al., 2014b) indicating that radiative forcing and feedback response are well simulated when realistic ocean surface temperature andvariability the satellite-based between independent estimates of HadCRUHcolumn integrated surface is prescribed. ERAI also captures month to month moisture,specific humidity with a consistent observations increasing (Willett trend et al., over 2008) the variability in the radiation budget remarkably well given coinciding 1988-2004 period (Allan et al., 2014a).
that cloud cover, which dominates these fluctuations, is not directly assimilated. However, decadal variability
21 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 and trends are unrealistic and the reanalysis does not represent volcanic radiative forcing as evident from within the climate system: combining satellite radiation the lack of response to the 1991 Pinatubo eruption. budget(Muller measurements and O’Gorman, with 2011) reanalysis and for energy tracking transports energy Variability is dominated by cooling following the Pinatubo volcanic eruption in 1991 (up to -3 Wm-2 caused by the decadal patterns of ocean heating (Liu et al., 2017) and potentiallyto estimate constrainsurface fluxes ocean can energy be used transports to identify (Trenberth regional El Niño events in which a warmer atmosphere loses more energyreflective to volcanicspace through aerosol infrared haze in emission. the stratosphere) This reduced and to the next. These advances take observing systems and energy uptake is of order 1 Wm-2 although an increase and Fasullo, 2017) and their changes from one decade in energy uptake of about 0.2 Wm-2 can occur as El Niño builds (Johnson and Birnbaum, 2017) and substantial regionalclimate models climate to change their limits responses (Desbruyères to radiative et al., forcings 2016; reorganisation of energy in the upper 400m of the ocean itPalmer, is necessary 2017; Wild, to disentangle 2017). To further the distinct constrain energy long-term budget occurs (Roemmich et al., 2015). Recent estimates of net radiative imbalance at the top of the atmosphere of variability of the climate system. 0.6-0.8 Wm-2 for 2005-2015 (Johnson et al., 2016) are responses and feedbacks influencing internal decadal primarily determined by ocean heat content changes Acknowledgements Support was provided from the UK National Centre and assumptions are applied in anchoring the satellite recordsmeasured (Loeb by Argo et buoys;al., 2012) this andwhich additional themselves observations provide for Atmospheric Science (NCAS) and the Natural excellent representation of interannual variability and for Earth Observation (NCEO) and National Centre decadal trends. The net imbalance is remarkably stable N006054/1) and DEEP-C (NE/K005480/1) projects. The over time with trends of just 0.02±0.01 Wm-2 per decade Environment Research Council SMURPHS (NE/ over the period 1988-2015, substantially smaller than Coupled Modelling is acknowledged for developing the the expected uncertainty. This stability indicates no CMIPWorld model Climate archive, Research and Programme's we thank the Workingclimate modelling Group on hiatus in anthropogenic radiative forcing despite slower groups for producing and making available their model global surface warming at the beginning of the 21st century (Xie & Kosaka, 2017) although there is intriguing provided coordinating support and led development of evidence of distinct global energy budget response to Ts softwareoutputs; forinfrastructure CMIP, the U.S. in Department partnership of withEnergy's the PCMDIGlobal variability and long-term climate response (Brown et al., climate model data sets were extracted from the British that influences interannual fluctuations, internal decadal AtmosphericOrganization forData Earth Centre System (http://badc.nerc.ac.uk/ Science Portals. AMIP5 home) and the Program for Climate Model Diagnosis and 2014; Xie et al., 2015). Intercomparison (pcmdi3.llnl.gov/esgcet). GPCP v2.3 key climate indicators combining a range of observations, data were extracted from http://gpcp.umd.edu. Merged reanalysesOngoing monitoring and model of Earth’ssimulations energy is budgetvaluable and for other (i) radiation budget data (v3) are available from http:// detecting unrealistic behaviour in observing systems, (ii) www.met.reading.ac.uk/~sgs02rpa/research/DEEP-C/ and trends and (iii) improving understanding of physical processesidentifying and unusual feedbacks. or significant Isolating internally climate fluctuations generated researchGRL/ and centre.CERES EBAFSSM/I v2.8 and and SSMIS ERBS v7 wide products field of wereview interannual to decadal variability from longer term (WFOV) v3 data was provided by the NASA Langley climate responses is essential for interpreting changes hadcrut4/.also obtained online (ftp.ssmi.com). HadCRUT4 data is and Venugopal, 2017) and the fundamental driving available from http://www.metoffice.gov.uk/hadobs/ factorsin the globalinvolving water Earth’s cycle energy (Gu et balance al., 2016; (Palmer Sukhatme and References Adler et al. (2017) An Update (Version 2.3) of the GPCP potentially be exploited in elucidating regional feedbacks Monthly Analysis (in preparation). McNeall, 2014; Trenbeth et al., 2016). This variability can understandingon internal decadal of howvariability the spatial (Brown nature et al., 2014;of climate Zhou and A. Bodas-Salcedo (2014a) Physically consistent et al., 2016; Xie et al., 2015) as well as in advancing responsesAllan, R. P., ofC. theLiu, globalM. Zahn, atmospheric D. A. Lavers, hydrological E. Koukouvagias cycle radiative forcings (Gregory and Andrews, 2016). The in models and observations, Surv. Geophys., 35, 533-552, regionalchange influences manifestation how ofsensitive changes the in globalthe energy climate budget is to doi:10.1007/s10712-012-9213-z.
Allan, R. P., C. Liu, N. G. Loeb, M. D. Palmer, M. Roberts, and water cycle have been identified globally (Myhre et D. Smith and P.-L. Vidale (2014b) Changes in global net al., 2017; Bony et al., 2013) and at hemispheric (Hwang radiative imbalance 1985-2012, Geophysical Research 2015).et al., 2013; Applying Loeb a etregional al., 2016; energy Stephens budget et perspective al., 2017) tois Letters, 41, 5588-559710.1002/2014GL060962 informativecontinental scalesin understanding (Bollasina et these al., 2011; water Dong cycle and responses Sutton,
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 22 Allen, M. R., and W. J. Ingram (2002), Constraints on future changes in climate and the hydrological cycle, Wild, and K. P. Shine (2011), The roles of aerosol, water Nature, 419, 224–232, doi:10.1038/nature01092. vaporHaywood, and cloud J. M., in N. future Bellouin, global A. dimming/brightening, Jones, O. Boucher, M. J. Geophys. Res., 116, D20203, doi:10.1029/2011JD016000.
Jones (2010), Precipitation, radiative forcing and global sulfate aerosol and the southward shift of tropical temperatureAndrews, T., P.change, M. Forster, Geophys. O. Boucher, Res. Lett.,N. Bellouin, 37, L14701, and A. precipitationHwang, Y-T., Frierson, in the 20th D. M. century. W. & Kang, Geophys. S. M. Anthropogenic Res. Lett. 40, doi:10.1029/2010GL043991. 1–6 (2013).
Bollasina, Massimo, Yi Ming, and V Ramaswamy (2011) Johnson, G. C., and A. N. Birnbaum (2017), As El Anthropogenic aerosols and the weakening of the South Asian summer monsoon. Science, 334, 502-505 gains more heat, Geophys. Res. Lett., 44, 438–445, doi:10.1126/science.1204994. doi:10.1002/2016GL071767.Niño builds, Pacific Warm Pool expands, ocean
S. Denvil (2013) Robust direct effect of carbon dioxide on multidecadal strengthening of the Walker circulation tropicalBony, S., G.circulation Bellon, D. Klocke,and regional S. Sherwood, precipitation, S. Fermepinand Nature L’Heureux, M.L., S. Lee & B. Lyon (2013), Recent Geoscience 6, 447–451, doi:10.1038/ngeo1799 doi:10.1038/nclimate1840 across the tropical Pacific, Nature Clim. Ch., 3, 571–576, Brown, P. T., W. Li, L. Li, and Y. Ming (2014), Top-of- atmosphere radiative contribution to unforced decadal Roberts, M. Valdivieso, J. Edwards, P.-L. Vidale (2017) global temperature variability in climate models, Geophys. EvaluationLiu, C. Allan, of R.P.,satellite M. Mayer, and P. reanalysis-based Hyder, N.G. Loeb global , C.D. Res. Lett., 41, 5175–5183, doi:10.1002/2014GL060625. Geophysical Research, doi: 10.1002/2017JD026616 Cowtan, K. and Way, R. G. (2014), Coverage bias in the net surface energy flux and uncertainty estimates, J. Loeb, N. G., J. M. Lyman, G. C. Johnson, R. P. Allan, D. R. Doelling, T. Wong, B. J. Soden and G. L. Stephens (2012), 1944.HadCRUT4 doi:10.1002/qj.2297 temperature series and its impact on recent temperature trends. Q.J.R. Meteorol. Soc., 140: 1935– and upper-ocean heating consistent within uncertainty, Dee, D. P., Uppala, S. M., Simmons, A. J., Berrisford, P., Poli, NatureObserved Geoscience, changes 5, in 110-113, top-of-the-atmosphere doi:10.1038/ngeo1375 radiation P., Kobayashi, S., Andrae, U., Balmaseda, M. A., Balsamo, G., Bauer, P., Bechtold, P., Beljaars, A. C. M., van de Berg,
Loeb, N.G., H. Wang, A. Cheng, S. Kato, J.T. Fasullo, K.- L., Bidlot, J., Bormann, N., Delsol, C., Dragani, R., Fuentes, Transports:M. Xu and R.P.Revisiting Allan (2016) the ObservationalPrecipitation Asymmetry Constraints M., McNally,Geer, A. J.,A. Haimberger,P., Monge-Sanz, L., B.Healy, M., Morcrette, S. B., Hersbach, J.-J., Park, H., Problemon Atmospheric in Climate and Models, Oceanic Climate Cross-Equatorial Dynamics, 46, 3239- Heat B.-K.,Hólm, Peubey,E. V., Isaksen, C., de L.,Rosnay, Kållberg, P., Tavolato,P., Köhler, C.,M., Thépaut, Matricardi, J.- 3257, 10.1007/s00382-015-2766-z.
Morice, C. P., J. J. Kennedy, N. A. Rayner, and P. D. Jones N. and Vitart, F. (2011), The ERA-Interim reanalysis: qj.828configuration and performance of the data assimilation temperature change using an ensemble of observational system. Q.J.R. Meteorol. Soc., 137: 553–597. doi:10.1002/ (2012), Quantifying uncertainties in global and regional Desbruyères, D., McDonagh, E.L. & King, B.A. (2016) D08101, doi:10.1029/2011JD017187. estimates: The HadCRUT4 dataset, J. Geophys. Res., 117, Curr Clim Change Rep, 2, 127-134, doi:10.1007/s40641- 016-0037-7Observational Advances in Estimates of Oceanic Heating, on the regional response of precipitation to climate change.Muller CJ, Nat O’Gorman Clim Change PA (2011) 1:266–271 An energetic perspective Dong, B. and R. Sutton (2015) Dominant role of greenhouse-gas forcing in the recovery of Sahel rainfall, Nature Clim. Ch., doi: 10.1038/nclimate2664 Myhre, G., P. Forster, B. Samset, Ø. Hodnebrog, J. Gregory, J. M., and T. Andrews (2016), Variation in Sillmann, S. Aalbergsjø, T. Andrews, O. Boucher, G. climate sensitivity and feedback parameters during the Shindell,Faluvegi, K.D. Shine,Flaeschner, C. Stjern, T. Iversen, T. Takemura, M. Kasoar, A. Voulgarakis, S. Kharin, historical period, Geophys. Res. Lett., 43, 3911–3920, A. Kirkevåg, J. Lamarque, D. Olivié, T. Richardson, D. doi:10.1002/2016GL068406. Response Model Intercomparison Project, Protocol and preliminaryand F. Zwiers, results. 2017: Bull. PDRMIP: Amer. A Meteor. Precipitation Soc. doi:10.1175/ Driver and BAMS-D-16-0019.1, in press. 1091. doi:10.1007/s00382-015-2634-x Gu, G., Adler, R.F. & Huffman, G.J. Clim Dyn (2016) 46:
23 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 (2012) Energetic constraints on precipitation under report number 011012, Remote Sensing Systems, Santa climateO'Gorman, change, P. A., R.Surv. P. Allan, Geophys., M. P. Byrne 33, and585-608, M. Previdi doi: Rosa,Wentz, CA, F. J.,46pp. (2013), SSM/I Version-7 Calibration Report, 10.1007/s10712-011-9159-6 Wild, M. (2017) Towards Global Estimates of the Surface Palmer, M. D., and D. J. McNeall (2014), Internal variability Energy Budget, Curr Clim Change Rep, doi:10.1007/ s40641-017-0058-x models, Environ. Res. Lett., 9, 034016, doi:10.1088/1748- 9326/9/3/034016.of Earth's energy budget simulated by CMIP5 climate Wild, M., J. Grieser, and C. Schär, 2008: Combined surface solar brightening and increasing greenhouse effect favour
Res. Lett., doi:10.1029/2008GL034842 Rep,Palmer, doi:10.1007/s40641-016-0053-7 M.D. (2017) Reconciling Estimates of Ocean recent intensification of the hydrological cycle. Geophys. Heating and Earth’s Radiation Budget, Curr Clim Change Willett, K.W., P.D. Jones, N.P. Gillett and P. W. Thorne, 2008: Recent changes in surface humidity: development of the (2014), A general framework for understanding the responseRoderick, of M. the L., water F. Sun, cycle W. to H. global Lim, and warming G. D. over Farquhar land HadCRUH dataset. J. Clim..21, 5364:5383 doi:10.5194/hess-18-1575-2014 budgets for anthropogenic and natural changes during and ocean, Hydrol. Earth Syst. Sci., 18(5), 1575–1589, globalXie, S.-P., warming Y. Kosaka, hiatus, Y. M. Okumura Nature (2016)Geoscience Distinct 9, energy29–33 Roemmich, D., J. Church, J. Gilson, D. Monselesan, P . (2016) doi:10.1038/ngeo2581, L17706. Sutton and S. Wijffels (2015) Nature Clim. Change, 5, 240–245, doi:10.1038/nclimate2513. Xie, SP. & Kosaka, Y. (2017) What Caused the Global
Salzmann, M. (2016) Global warming without global Rep, 3, 128, doi:10.1007/s40641-017-0063-0 Surface Warming Hiatus of 1998–2013? Curr Clim Change doi:10.1126/sciadv.1501572 mean precipitation increase? Science Advances, 2, decadal cloud variations on the Earth’s energy budget, Simmons, A. J., K. M. Willett, P. D. Jones, P. W. Thorne, and Zhou, C., M. D. Zelinka and S. A. Klein (2016), Impact of D. P. Dee (2010), Low-frequency variations in surface atmospheric humidity, temperature, and precipitation: Nature Geosci., 9, 871–874, doi:10.1038/NGEO2828 Inferences from reanalyses and monthly gridded observational data sets, J. Geophys. Res., 115, D01110, doi:10.1029/2009JD012442.
A. Behrangi, J. E. Kay, P. J Webster (2016) The Curious Stephens, G.L., Hakuba, M.Z., Hawcroft, M. J. M. Haywood, and Energy Balances, Curr Clim Change Rep, 2, 135-147, doi:10.1007/s40641-016-0043-9Nature of the Hemispheric Symmetry of the Earth’s Water
Sukhatme, J. and Venugopal, V. (2017), Waxing and waning of observed extreme annual tropical rainfall.
Taylor KE, Stouffer RJ, Meehl GA (2011) An overview of CMIP5Q.J.R. Meteorol. and the experimentSoc., 142: 102–107. design. Bull doi:10.1002/qj.2633 Am Meteorol Soc 93:485–498. doi:10.1175/BAMS-D-11-00094.1
Trenberth KE, Dai A, Rasmussen RM, Parsons DB (2003) The changing character of precipitation. Bull Am Meteorol Soc 84:1205–1217
Cheng (2016) Insights into Earth’s Energy Imbalance fromTrenberth, Multiple K., Sources. J. Fasullo, J. Climate, K. von Schuckmann,29, 7495–7505, and doi: L. 10.1175/JCLI-D-16-0339.1.
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 24 Toward predicting volcanically-forced decadal climate variability doi: 10.22498/pages.25.1.25 Davide Zanchettin1, Francesco S.R. Pausata2,3, Myriam Khodri4, Claudia Timmreck5, Hans Graf6, Johann H. Jungclaus5, Alan Robock7, Angelo Rubino1, Vikki Thompson8 1 Department of Environmental Sciences, Informatics and Statistics, University Ca’Foscari of Venice, Italy 2 Department of Meteorology (MISU), Stockholm University, Sweden 3 Department of Earth and Atmospheric Sciences, University of Quebec in Montreal (UQÀM), Montreal (QC), Canada 4 IRD/IPSL/Laboratoire d'Océanographie et du Climat, Paris, France 5 Max Planck Institute for Meteorology, Hamburg, Germany 6 University of Cambridge, Cambridge, UK 7 Department of Environmental Sciences, Rutgers University, New Brunswick, USA 8 Met Office Hadley Centre, Exeter, UK Volcanic forcing and climate Strong volcanic eruptions inject into the stratosphere of tropical eruptions, for which the bulk of the volcanic massive amounts of chemically and microphysically aerosol cloud remains largely constrained in the tropical active gases that lead to the formation of aerosol stratosphere, simple theoretical arguments indicate that particles, affecting the Earth’s radiative balance and the aerosol radiative heating enhances the upper-level equator-to-pole temperature gradients that, by thermal 2016). Sulfate aerosol particles scatter solar radiation wind balance, can force a strengthened stratospheric backclimate to space,(Robock, which 2000; results Timmreck, in global 2012; surface LeGrande cooling et and al., polar vortex in both hemispheres, as diagnosed from slowdown of the global hydrological cycle. The particles also absorb radiation in the infrared and near-infrared et al., 2014). The consequent downward penetration bands, causing local warming of the lower stratosphere. ofclimate the westerly models wind(e.g., Stenchikovanomalies atet theal., edge2002; of Zanchettin the polar Both direct radiative forcing effects are temporary, their vortex and their interaction with topography provide time scale being set by the persistence of the volcanic further elements of a top-down atmospheric mechanism aerosol cloud in the lower stratosphere. This amounts to a couple of years in the case of the strongest recent tropical eruptions, such as the 1815 eruption of Mt typicallyof volcanic project forcing. on Ina positive the Northern anomaly Hemisphere, of the North its tropospheric effects during the first post-eruption winter impact of strong volcanic eruptions can last well beyond theTambora timescale in Indonesia of the direct (Fig. radiative 1a). However, perturbation the through climatic Atlantic Oscillation/Arctic Oscillation (NAO/AO) and the dynamic alterations it induces in the entire coupled componentassociated continental of a major warming recognized (Stenchikov general etpathway al., 2006; of climate system. These include “feedbacks” in their Graf et al., 2014; Zambri and Robock, 2016). This is a key related to changes in climatic variables that operate volcanically-forced decadal climate signals (Otterå et al., throughclassic definition changes ofin amplificationglobal-mean andsurface dampening temperature loops 2010; Zanchettin et al., 2012).
Specifically, the NAO-related post-eruption modifications of(Boucher positive et feedbacks al., 2013). involving For instance, snow thecover so-called and sea “polar ice – convectiveto the wind mixing field modifythrough the anomalous circulation turbulent in the upper heat providesamplification” one elementof climate of signals inter-hemispheric – mainly a consequence asymmetry North Atlantic Ocean and locally enhance oceanic to the decadal climate response to volcanic eruptions buoyancy effects of the post-eruption radiative cooling, leadingand freshwater to strengthened fluxes. These deep superpose water formation. on the extensiveThe slow Dynamical impacts further stem from the spatially propagation of so-formed water masses in the ocean heterogeneousthrough global radiativestructure cooling of volcanic (Zanchettin forcing. et In al., the 2014). case abyss is expected to protract the fast oceanic response
25 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 2013). Lacking further external excitation (e.g., by a successiveclimates (Zanchettin eruption), etnegative al., 2012, feedbacks 2013b; Sicreeventually et al., become predominant and the near-surface system relaxes back to the mean pre-eruption state as part of
thus sets the phase of internal modes of interdecadal a roughly bi-decadal fluctuation. The feedback loop can be protracted, with dampening intensity, beyond one climate variability (Otterå et al., 2010), whose effects
fluctuation (Swingedouw et al., 2015). In the latter case, interdecadaldeep ocean anomalies general oceanicmay remain response significant is also for foundmuch forlonger high-latitude (Gleckler eteruptions, al., 2006; for Gregory, which 2010).it is the A similardirect radiative surface cooling at subpolar latitudes linked
eruption’s hemisphere that typically leads to enhanced to the confinement of the volcanic aerosol cloud to the 2015). oceanic deep convection (Fig. 2, see also Pausata et al., Knowledge gaps The general framework outlined above is useful to identify the core dynamics involved in post-eruption
must be taken into consideration. Above of all, direct observationsdecadal climate of strong variability. volcanic However, eruptions several are very caveats few
allow robust statistics of their climate impact, and hence attribution.- only five in Therefore, the instrumental large partperiod of - ourwhich knowledge does not builds on climate model simulations and proxy-based climate reconstructions, both of which have large
Figure 1: Simulated global-average top-of-atmosphere net Incessant improvement in both tools brings old evidence radiative anomalies (a: 3-month smoothing) and simulated uncertainties and deficiencies (e.g., Zanchettin, 2017). Arctic sea ice cover evolution (b: 61-month smoothing) leading to a preferred enhanced stratospheric polar around the 1815 Tambora eruption in three climate simulation vortexback into in post-eruption the discussion. winters For instance, have been the questioned mechanisms by ensembles with the ECHAM5/MPIOM coupled climate model, differing in the ensemble-mean initial state and in the applied recent studies suggesting that the mechanism based on forcing. Dark green: full-forcing conditions (including the the thermal wind balance and outlined above does not Dalton Minimum of solar activity); Red: volcanic forcing-only always hold (e.g., Bittner et al., 2016), possibly as the conditions, including both the 1809 and 1815 eruptions; Blue: zonal wind response to direct aerosol radiative heating volcanic forcing-only conditions, without the 1809 eruption. may be dominated by other effects, such as the residual Lines (shading): mean (1- standard error of the mean). circulation response to anomalous wave activity (Toohey Black dashed lines: 5th–95th percentile intervals for signal et al., 2014). Accordingly, obvious implications for the occurrence in the control run.σ Vertical dotted lines indicate the 1809 and Tambora eruptions. Each ensemble consists known critical source of tropospheric wave disturbances of 10 simulations differing in the initial state. Note that the affectingpolar vortex stratospheric response dynamics stem from (e.g., the tropicalGraf et al., Pacific, 2014). a Arctic sea ice response is significantly different in the three Instrumental observations and climate proxy-based ensembles whereas the applied forcing, in terms of anomalous top-of-atmosphere net flux, is practically indistinguishable. reconstructions indicate that volcanic eruptions tend to For details see: Zanchettin et al. (2013a). be followed by an El Niño event. A newly discovered causal mechanism is initiated by cooling over Africa (the largest to decadal time scales through the tendency for tropical landmass), which reduces precipitation and reinvigoration of the oceanic meridional overturning forces an atmospheric Kelvin wave response that couples circulation that culminates several years – up to a decade or so – after major eruptions, as diagnosed from models. Implications for meridional ocean heat transports and thewith seasonal western cyclePacific of convectionconvection, to the trigger effect westerly of volcanism wind sea ice dynamics contribute to regional characterization anomalies and a Pacific El Nino. While modulated by of the signal, and hence to recognizable traces of volcanic signals especially in extratropical and polar regional responseonwind forcing involves over more the external Pacific persists forcing thanduring traditional, the year after the eruption, implying that the Pacific El Nino-like
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 26 discrepancies in the estimated post-eruption cooling.
the top-down mechanism of volcanic forcing, hence on This outlines the complexity of competing influences on or small tropical eruptions, whose uncertainty cascades the post-eruption positive NAO anomaly for moderate simulated by coupled models are another major source ofon uncertainty the decadal to be oceanic understood response. considering Ocean the dynamics different
time scales of simulated oceanic responses (Otterå et generatesal., 2010; Mignotbi-decadal et al., variability 2011; Zanchettin in the etoverturning al., 2012). circulationFurthermore, strength the fact seems that a climateto determine model spontaneouslyits excitability to the general response mechanism outlined above (Swingedouw et al., 2015).
Inherent sources of uncertainty The climatic response to a given volcanic eruption is
strongly depends on the characteristics of the forcing. Anhighly obvious specific. determinant First, the factor general is the response magnitude mechanism of the eruption, whose potential control can be, for instance,
mayestimated lead sea looking ice to at cover the sea regions ice response. of strong For oceanic a very strong deep convection,eruption, polar thereby amplification hampering of the deep global water cooling formation signal through insulation of the ocean-atmosphere boundary. Associated increased freshwater export from the Arctic also contributes to stabilize the ocean water column. These will lead ultimately to a tendency for weakening – instead of strengthening – of the thermohaline
Figure 2: Decadal oceanic response to a high-latitude Second, the way the volcanic aerosol cloud distributes in eruption resembling the multistage 1783 eruption of Laki circulation (e.g., Zhong et al., 2010; Mignot et al., 2011). (Iceland) simulate by the NorESM coupled climate model. a) Changes in the strength of the Atlantic Meridional Colosethe stratosphere et al., 2016). influences In this regard, both the the direct latitudinal radiative position and Overturning Circulation (AMOC) estimated as the maximum ofdynamic the erupting atmospheric volcano responses is an obvious (e.g., determinantToohey et al., factor,2014; of the zonally-integrated overturning stream function in the but similar uncertainty on the spatial structure of the Atlantic. b) Changes in Ocean Heat Content (OHC) averaged from the surface to selected depths for the global ocean. In volcanic aerosol cloud can be originated by the season of both panels the solid lines denote the ensemble average the eruption (Stevenson et al., 2017). changes and shadings represent the confidence intervals at approximate 95% level (twice the standard error of the mean) A milestone in our understanding of volcanically forced of the difference in all pairs of experiments that comprise the decadal climate variability was the recent recognition ensembles. The ensemble consists of 10 simulations differing that the mean climate state, the phase and amplitude of for the initial state. Anomalies are calculated with respect to a ongoing internal variability at the time of an eruption, control simulation. The figure has been adapted from Pausata such as that associated with major climatic modes et al. (2015).
ofincluding, additional e.g., forcing El Niño factors Southern crucially Oscillation determine (ENSO) how the or there are stratospheric (e.g., Scaife et al., 2009) and the Quasi-Biennial Oscillation (QBO), and the presence internally generated events (Khodri et al., 2017). However, of El Niño forcing on the atmospheric circulation over climate system responds to the volcanic forcing (Zhong tropospheric (e.g., Graf and Zanchettin, 2012) pathways et al., 2010; Zanchettin et al., 2012, 2013a; Berdahl and roleRobock, of background 2013; Swingedouw conditions et for al., the 2015; case Pausata of Arctic et sea al., Inthe addition, North Atlantic sampling that issues project in on simulation a negative ensembles NAO; this ice2016). response Fig. 1 to(after the 1815Zanchettin Tambora et al.,eruption 2013a) simulated shows the in (Lehnerwould counteract et al., 2016) the NAO+and uncertainty tendency above linked described. to the three ensembles in which the volcanic forcing is the same eruption’s season (Stevenson et al., 2017) are recently but background climate state and histories are different. proposed explanations for reconstructions-simulations
Results show that a significant increase in Arctic sea
27 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 ice cover is consistently diagnosed after the eruption in an eventuality. Therefore, a new SSiRC initiative called all ensembles but the average anomalies differ in both VolRes ("Volcano Response Plan after the next major magnitude and duration. The inter-ensemble differences eruption") has been launched aiming at developing a differences in the decadal feedback mechanisms activated tools and strategies to be readily applied for the next in the the coupled Arctic atmosphere-ocean-sea-ice sea ice response reflect system substantial after majorscientific volcanic plan eruption. to prepare Characterization observational of and the modelling potential the eruption. climatic impact of an eruption in the more distant future (such as envisaged for end-of-the-century warming This dependency on the background climate state can scenarios) bears additional uncertainties related to the partly explain the different, often contrasting, results dependence of dynamics of the eruption plume on the found for simulated and reconstructed post-eruption decadal variability from different volcanic eruptions subject to global temperature changes (Aubry et al., 2016). atmospheric stratification and tropopause height, in turn understanding how the timing between subsequent (Zanchettin et al., 2013a,b). This concept also allows climatic response to Volcanic forcing” (VolMIP) Finally, the “Model Intercomparison Project on the ifvolcanic two eruptions eruptions are can roughly deterministically paced at one influenceperiod of the Coupled Model Intercomparison Project phase 6 (CMIP6), above-mentionedresponse in the case decadal of a volcanicmechanism cluster. (roughly Specifically, two (Zanchettin et al., 2016) has been created as part of the decades), they will interfere constructively, as they will perturbation experiments to improve comparability of occur around the same phase of internal modes of oceanic resultsto define across a coordinated different protocolclimate models.for idealized Interest volcanic- is on variability. In contrast, if they are paced at half the period various aspects of volcanically-forced climate variability, of the mechanism (a decade or so), they will interfere destructively (Swingedouw et al., 2015). Intriguingly, both both the seasonal-to-interannual atmospheric response cases apply to the most recent strong volcanic eruptions: andwith the specific interannual-to-decadal sets of experiments response designed of to the investigate coupled ocean-atmosphere-sea-ice system. Through systematic roughly two decades (constructive interference), while El and consistent (across the different models) sampling of Agung in 1963 and El Chichón in 1982 are paced at marginsChichón andfor long-termPinatubo in predictability 1991 were paced of decadal at roughly climate one volcanicinternal forcingvariability or explain(e.g., ENSO, the lack QBO), thereof. VolMIP will allow impactsdecade (destructiveby strong volcanic interference). eruptions. This finding widens the identification of robust response mechanisms to VolMIP will also foster investigation of simulated Opportunities for progress A series of research initiatives are currently contributing currently an overlooked topic due to the known severe Southern Hemispheric responses to volcanic forcing, to building the scientific basis for reaching such an 2013).climate Moremodel generally,biases in theif volcanically-forcedSouthern Hemisphere decadal (e.g., byambitious means of objective climate bymodels filling with major interactive gaps of understanding. stratospheric climateSimpson variability et al., 2012; can Salleé be etunderstood al., 2013; Turnerthrough et the al., aerosols,A first goal of of the current forcing research generated is robust by acharacterization, given eruption excitation by volcanic forcing of internal modes of climate based on the estimated amount of gaseous sulphur species it injects in the stratosphere. The WCRP/SPARC robust simulation of such modes. Current climate models, Stratospheric Sulfur and its Role in Climate (SSiRC) variability, confidence must be built on the accurate and initiative (http://sparc-ssirc.org/ssirc.html, Timmreck spatial pattern, time scales and teleconnections of et al., 2016b) coordinates the international activities however, have difficulties in reproducing the observed on stratospheric aerosol research aiming at better understanding and hence modeling of the stratospheric dominant modes such as ENSO (e.g., Zou et al., 2014) or the aerosol layers and their controls. SSiRC will help study Atlantic Multidecadal Oscillation (Kavvada et al., 2013). why the characterization of the volcanic aerosol cloud and assessment of volcanic forcing impacts on decadal climate the radiative forcing generated by state-of-the-art global Only a few studies specifically focused on the quantitative aerosol models for a certain sulphur injection remain inpredictions the presence and of potential strong natural predictability climate (Collins, variability 2003; on theShiogama one hand et al., and 2010; because Timmreck of different et al., 2016a).magnitude Overall, and relatedhighly uncertainto the treatment (e.g., SPARC, of aerosol 2006; microphysics Zanchettin et and al., frequency of volcanic eruptions on the other hand, it climate2016). Focus physical will be inprocesses, particular onsuch model as inconsistenciesstratospheric circulation and stratosphere-troposphere coupling. volcanoes of regional climates. Improvements of decadal climateis difficult prediction to assess systems the potentialconcerning predictability implementationof from In addition, unpredictability of timing and magnitude volcanic forcing (e.g., LeGrande et al., 2016) and bias of volcanic eruptions is a major source of uncertainty and the climate community should be prepared for such are milestones toward robust prediction of volcanically- estimation and correction (e.g., Hawkins et al., 2014)
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 28 forced decadal climate variability. Along this long-term goal, within CMIP6, a joint decadal climate prediction experiment between VolMIP and the Decadal Climate Colose,and the Earth’sC.M., A. atmosphere, N. LeGrande, Washington, A.N., and DC: M. AGU;Vuille, 283-300 2016: Prediction Panel (Boer et al., 2016) will be conducted to address the climatic implications if a Pinatubo-like hydroclimate during the last millennium, Earth Sys. Dyn., eruption would have occurred in 2015. 7(3):Hemispherically 681-696, doi:10.5194/esd-7-681-2016 asymmetric volcanic forcing of tropical
In conclusion, there is emerging evidence that volcanic Gleckler, P. J., K. AchutaRao, J. M. Gregory, B. D. Santer, K. E. Taylor, and T.M. L. Wigley, 2006: Krakatoa lives: the effect through mechanisms that are increasingly better of volcanic eruptions on ocean heat content and thermal understood.forcing can significantly Milestones affect on thedecadal road climate toward variability robust expansion, Geophys. Res. Lett., 33(17):L17702 prediction of volcanically-forced decadal variability include improved understanding and implementation of aerosol forcing in decadal prediction systems and Nino, the subtropical bridge, and Eurasian climate, J. improved simulated representation and estimation of Geophys.Graf, H.-F., Res., and 117, D. doi:10.1029/2011JD016493 Zanchettin, 2012: Central Pacific El internal decadal climate variability.
References Aubry, T. J., A. M. Jellinek, W. Degruyter, C. Bonadonna, andGraf, near-surface H.-F., D. Zanchettin, winter C.signature Timmreck, of andthe M.Northern Bittner, 2014: Observational constraints on the tropospheric warming on the rise of volcanic plumes and implications 3245, doi:10.1007/s00382-014-2101-0 forV. Radić, future M.volcanic Clyne, aerosol A. Quainoo, forcing, 2016: J. Geophys. Impact Res. of Atmos. global Hemisphere stratospheric polar vortex, Clim. Dyn., 43: 121, 13,326–13,351, doi:10.1002/2016JD025405 Gregory, J. M., 2010: Long-term effect of volcanic forcing on ocean heat content, Geophys. Res. Lett., 37, L22701, doi:10.1029/2010GL045507 cryosphere response to volcanic eruptions in the PaleoclimateBerdahl, M., and Modeling A. Robock, Intercomparison 2013: Northern Project Hemispheric 3 last millennium simulations, J. Geophys. Res. Atmos., 118, 2014: The interpretation and use of biases in decadal 12,359–12,370, doi:10.1002/2013JD019914 climateHawkins, predictions. E., B. Dong, J. Climate,J. Robson, 27:2931-2947, R. Sutton, and doi:http:// D. Smith, dx.doi.org/10.1175/JCLI-D-13-00473.1
Ineson, S., and A. A. Scaife, 2009: The role of the Lohmann,Boucher, O.,P. Rasch, D. Randall, S.K. Satheesh, P. Artaxo, S. Sherwood, C. Bretherton, B. Stevens G. stratosphere in the European climate response to El Feingold, P. Forster, V.-M. Kerminen, Y. Kondo, H. Liao, U. Niño, Nature Geosci., 2, 32–36 Change 2013: The Physical Science Basis. Contribution of and X.Y. Zhang, 2013: Clouds and Aerosols. In: Climate Kavvada, A., A. Ruiz-Barradas, A., and S. Nigam, 2013:
Working Group I to the Fifth Assessment Report of the IPCC AR5 climate simulations, Clim. Dyn., doi:10.1007/ Nauels,Intergovernmental Y. Xia, V. Bex Panel and P.M.on Climate Midgley Change (eds.)], [Stocker, Cambridge T.F., s00382-013-1712-12013.AMO’s structure and climate footprint in observations and UniversityD. Qin, G.-K. Press, Plattner, Cambridge, M. Tignor, United S.K. Allen, Kingdom J. Boschung, and New A. York, NY, USA Khodri, M., T. Izumo, J. Vialard, C. Cassou, M. Lengaigne, J. Mignot, E. Guilyardi, N. Lebas, Y. Ruprich-Robert, A.,
Danabasoglu, B. Kirtman, Y. Kushnir, M. Kimoto, G. A. volcanic eruptions trigger El Niño events, Nature Comm., Boer, G. J., D. M. Smith, C. Cassou, F. Doblas-Reyes, G. inRobock, press. and M. J. McPhaden, 2017: How tropical explosive M. Rixen, Y. Ruprich-Robert, and R. Eade, 2016: The DecadalMeehl, R. Climate Msadek, Prediction W. A. Mueller, Project K. (DCPP) E. Taylor, contribution F. Zwiers, LeGrande, A. N., K. Tsigaridis, and S. E. Bauer, 2016: to CMIP6, Geosci. Model Dev., 9, 3751-3777, doi:10.5194/ Role of atmospheric chemistry in the climate impacts of gmd-9-3751-2016 stratospheric volcanic injections, Nature Geosci., 9, 652– 655, doi:10.1038/ngeo2771
Using a large ensemble of simulations to assess the Bittner, M., H. Schmidt, C. Timmreck, and F. Sienz, 2016: to tropical volcanic eruptions and its uncertainty, volcanicLehner, F., eruptions A. P. Schurer, for detection G. C. Hegerl, and attribution, C. Deser, Geophys.and T. L. Geophys.Northern Res.Hemisphere Lett., 43(17), stratospheric 9324-9332. dynamical response Res.Frölicher, Lett., 43,2016: 2851–2858, The importance doi:10.1002/2016GL067935 of ENSO phase during
Collins, M., 2003: Predictions of climate following volcanic
Mignot, J., M. Khodri, C. Frankignoul, and J., Servonnat, eruptions, In: Robock A., Oppenheimer C. (eds) Volcanism 2011: Volcanic impact on the Atlantic Ocean over the last
29 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 millennium, Clim. Past, 7, 1439–1455, doi:10.5194/cp-7- ozone depletion, J. Geophys. Res., 107(D24): 4803, 1439-2011 doi:10.1029/2002JD002090
External forcing as a metronome for Atlantic multidecadal Otterå, O.H., M. Bentsen, H. Drange, and L. Suo, 2010: oscillationStenchikov, response G., K. Hamilton, to volcanic R. J.eruptions Stouffer, in A. the Robock, IPCC AR4V. Ramaswamy, climate models, B. Santer, J. Geophys. and H.-F. Res., Graf, 111: 2006: D07107, Arctic variability, Nature Geosci., doi:10.1038/NGEO995 doi:10.1029/2005JD006286 2015: Impacts of high-latitude volcanic eruptions on Pausata, F. S. R., L. Chafik, R. Caballero, R., and D. S. Battisti, 13788, doi:10.1073/pnas.1509153112 and C. Gao, 2017: Role of eruption season in reconciling ENSO and AMOC, P. Natl. Acad. Sci. USA, 112, 13784– modelStevenson, and proxyS., J. T. responsesFasullo, B. toL. tropicalOtto-Bliesner, volcanism, R. A. Tomas,P. Natl. Acad. Sci. USA, doi:10.1073/pnas.1612505114
Pausata, F. S. R., C. Karamperidou, R. Caballero, and theD. S.role Battisti, of initial 2016: conditions, ENSO response Geophys. to high-latitude Res. Lett., Masson-Delmotte, P. G. Butler, M. Khodri, and R. Seferian, doi:10.1002/2016GL069575volcanic eruptions in the Northern Hemisphere: 2015:Swingedouw, Bidecadal D., P. North Ortega, Atlantic J. Mignot, ocean E. Guilyardi,circulation V. variability controlled by timing of volcanic eruptions, Robock, A., 2000: Volcanic eruptions and climate, Rev. Nature Comm., 6, 6545, doi:10.1038/ncomms7545 Geophys., 38, 2: 191-219 Timmreck, C., 2012: Modeling the climatic effects of large Sallée, J.-B., E. Shuckburgh, N. Bruneau, A. J. S. Meijers, T. J. volcanic eruptions, WIREs Clim. Change, 3: 545–564, doi:10.1002/wcc.192
Bracegirdle, and Z. Wang, 2013: Assessment of Southern 1845–1862,Ocean mixed doi:10.1002/jgrc.20157 layer depths in CMIP5 models: Historical bias and forcing response, J. Geophys. Res.-Oceans, 118, J.Timmreck, Mills, R. C.,Neely, G. W. A. Mann, Schmidt, V. Aquila, J.-X. Sheng, C. Brühl, M. M.Toohey Chin, andS. S. D. Dhomse, Weisenstein, J. M. English,2016b: R.ISA-MIP: Hommel, A L.co-ordinated A. Lee, M. T. Yokohata, M. Ishii, T. Nozawa, M. Kimoto, 2010: intercomparison of Interactive Stratospheric Aerosol Shiogama, H., S. Emori, T. Mochizuki, S. Yasunaka, models, Geophys. Res. Abstr., 18, EGU2016-13766, EGU Potential Predictability of the Natural Variability in General Assembly 2016 Near-TermPossible Influence Climate ofPredictions, Volcanic ActivityAdv. Meteorol., on the 657318, Decadal doi:10.1155/2010/657318 2016a: The impact of stratospheric volcanic aerosol on Sicre M. A., M. Khodri, J. Mignot, J. Eiriksson, K. L. decadalscaleTimmreck, C., climate H. Pohlmann, predictions, S. Geophys. Illing, and Res. C. Lett., Kadow, 43, Knudsen, U. Ezat, I. Closset, P. Nogues, and G. Massé, 834–842, doi:10.1002/2015GL067431 2013: Sea surface temperature and sea ice variability in the subpolar North Atlantic from explosive volcanism of the late thirteenth century, Geophys. Res. Lett., 40 (20), 5526-5530. doi: 10.1002/2013GL057282. Toohey, M., K. Krüger, M. Bittner, M., C. Timmreck, and mechanismsH. Schmidt, 2014:and sensitivity The impact to forcing of volcanic structure, aerosol Atmos. on Chem.the Northern Phys., 14, Hemisphere 13063–13079, stratospheric doi:10.5194/acp-14- polar vortex: 13063-2014 Simpson, I., P. Hitchcock, T. Shepherd, and J. Scinocca, Wind2012: BiasesSouthern in a Annular Comprehensive Mode Dynamics GCM, J. Clim., in Observations 26, 3953— Turner, J., T. J. Bracegirdle, T. Phillips, G. J. Marshall, and 3967,and Models. doi:10.1175/JCLI-D-12-00348.1 Part 1: the Influence of Climatological Zonal ice extent in the CMIP5 models, J. Clim., 26: 1473-1484 SPARC, 2006: SPARC Assessment of Stratospheric Aerosol J. S. Hosking, 2013: An initial assessment of Antarctic Sea Properties (ASAP), L. Thomason and Th. Peter (Eds.), summer monsoon reduction after volcanic eruptions available at www.sparc-climate.org/publications/sparc- inZambri, Coupled B., andModel A. Robock,Intercomparison 2016: Winter Project warming 5 (CMIP5) and reports/SPARC Report No. 4, WCRP-124, WMO/TD - No. 1295, simulations, Geophys. Res. Lett., 43, 10,920-10,928, doi:10.1002/2016GL070460 Stenchikov, G., A. Robock, V. Ramaswamy, M. D.
on decadal climate variability and predictability, Curr. PinatuboSchwarzkopf, eruption: K. Hamilton, effects of andvolcanic S. Ramachandran, aerosols and Clim.Zanchettin, Ch. Rep., D., 3:2017: 150, Aerosol doi:10.1007/s40641-017-0065-y and solar irradiance effects 2002: Arctic Oscillation response to the 1991 Mount
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 30 hemisphericZanchettin, D., asymmetry O. Bothe, C. Timmreck,in the sea-ice J. Bader, response A. Beitsch, to H.-F. Graf, D. Notz, and J. H. Jungclaus, 2014: Inter- Earth Syst. Dynam., 5, 223–242, doi:10.5194/esd-5-223- 2014volcanic forcing simulated by MPI-ESM (COSMOS-Mill),
Bi-decadalZanchettin, variability D., C. Timmreck, excited H.-F.in the Graf, coupled A. Rubino, ocean– S. atmosphereLorenz, K. Lohmann, system by K. strongKrueger, tropical and J. H.volcanic Jungclaus, eruptions, 2012:
1167-1 Clim. Dyn., 39:1-2, 419-444, DOI:10.1007/s00382-011-
Zanchettin, D., O. Bothe, H.-F. Graf, S. J. Lorenz, J. responseLuterbacher, to strong C. Timmreck, volcanic and eruptions, J. H. Jungclaus, J. Geophys. 2013a: Res. Atm.,Background 118(10): conditions 4090-4106, influence doi:10.1002/jgrd.50229 the decadal climate
DelayedZanchettin winter D., C. Timmreck,warming: a O. robust Bothe, decadalS.J. Lorenz, response G. Hegerl, to H.-F. Graf, J. Luterbacher, and J.H. Jungclaus, 2013b: 40, 204–209 doi:10.1029/2012GL054403 strong tropical volcanic eruptions?, Geophys. Res. Lett.
W.Zanchettin, T. Ball, S. Bauer, D., M. S. Khodri, Bekki, S. C. S. Dhomse, Timmreck, A. N. M. LeGrande, Toohey, A.G. W.Schmidt, Mann, E.L. P.Marshall, Gerber, G. M. Hegerl, Mills, M. A. Robock,Marchand, F. S. U. R. Niemeier, Pausata, V. Poulain, E. Rozanov, A. Rubino, A. Stenke, K. Tsigaridis,
Project on the climatic response to Volcanic forcing (VolMIP):and F. Tummon, experimental 2016: design The and Model forcing Intercomparison input data for CMIP6, Geosci. Model Dev., 9, 2701-2719, doi:10.5194/ gmd-9-2701-2016
D. A. Bailey, D. P. Schneider, and A. Geirsdottir, 2010: Centennial-scaleZhong, Y., G. H. Miller, climate B. L.change Otto-Bliesner, from decadally-paced M. M. Holland, explosive volcanism: a coupled sea ice-ocean mechanism, Clim. Dyn. 23:5–7, doi:10.1007/s00382-010-0967-z model simulations of the impacts of the two types of El Niño onZou, the Y., U.S.J.-Y. Yu,winter T. Lee, temperature, M.-M. Lu, and J. Geophys. S. T. Kim, 2014:Res. -Atmos., CMIP5 119(6):3076–3092. doi:10.1002/2013JD021064
31 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Towards the prediction of multi-year to decadal climate variability in the Southern Hemisphere doi: 10.22498/pages.25.1.32 Scott Power1, Ramiro Saurral2, Christine Chung1, Rob Colman1, Viatcheslav Kharin3, George Boer3, Joelle Gergis4, Benjamin Henley4, Shayne McGregor5, Julie Arblaster5, Neil Holbrook6, Giovanni Liguori7 1 Bureau of Meteorology, Australia 2 CIMA, Ciudad Universitaria, Argentina 3 Environment and Climate Change, Canada 4 University of Melbourne, Australia 5 Monash University, Australia 6 University of Tasmania, Australia 7 Georgia Institute of Technology, USA Introduction Multi-year (2-7 years) and decadal climate variability and Vimont, 2004). Although a similar mechanism livelihoods and economies. Consequently, learning more ENSO via the North Pacific Meridional Modes (Chiang about(MDCV) the cancauses have of this a profoundvariability, influencethe extent onto which lives, tropical-extratropical interactions, including via both it can be predicted, and the greater the clarity that we thehas ocean not yet and been atmosphere, proposed are for thethought South to Pacificbe at least DCV, can provide on the climatic conditions that will unfold partly responsible for inducing decadal to multidecadal over coming years and decades is a high priority for the initiatives by WCRP, CLIVAR, and in the Decadal Climate variability in the South Pacific too (e.g., McGregor et al., Predictionresearch community. Project (Boer This et importance al., 2016) that is reflected target this in areanew 2007, 2008, 2009a and b; Farneti et al., 2014; Zhang et decadalal., 2014; and Ding multi-decadal et al., 2015). variability Inter-basin (e.g., McGregor interactions et we know, and have recently learnt, about the causes and are also thought to play a role in driving Pacific Ocean of research. Here we briefly examine some of the things and current skill in its prediction. al., 2014; Kucharski et al., 2016, Chikamoto et al., 2016). predictability of Southern Hemisphere MDCV (SH MDCV), duePalaeoclimate to spectral reconstructionsbiases in the different of MDCV proxy in records, the Pacific non- Causes of SH MDCV stationarityvary significantly of teleconnections in their spectral and characteristics, the use of various likely As with other parts of the globe, internally generated statistical reconstructions methods (Bateup et al., 2015).
(Kirtman et al., 2013). Major components of internal External forcing, both natural and anthropogenic, can climate variability is a major element of SH MDCV (e.g., Bindoff et al., 2013 and references therein). Natural variability for the Southern Hemisphere are decadal driversalso drive include decadal volcanic and longer-term eruptions variability(e.g. Church in theet al.,SH variations in ENSO, the Interdecadal Pacific Oscillation 2005), while anthropogenic external drivers include (IPO, Garreaud and Battisti, 1999; Power et al., 1999; changes in greenhouse gas concentrations, aerosols (Cai Salinger et al., 2001; Folland et al., 2002; Wu and Hsieh, et al., 2010) and stratospheric ozone (Kirtman et al., 2003; Holbrook et al., 2011; Christensen et al., 2013; MeehlKosaka et and al., Xie,2016), 2013; the EnglandSouthern et Annular al., 2014; Mode Holbrook (SAM et al., 2014; Watanabe et al., 2014; Henley et al., 2015; How2013; well Arblaster do climate et al., 2014; models Eyring simulate et al., 2013). MDCV in the SH? (Shiotani, 1990; Thompson et al., 2000; Watterson, 2009; Pohl et al., 2009; Yuan and Yonekura, 2011; Jones et al., 2016), and the Indian Ocean Dipole. etThe al., ability 2013). of They climate concluded models that to simulate models reproduce Earth's climate many described by Di Lorenzo et al. (2015) as a seasonally- importantwas assessed modes in the of IPCC variability. Fifth Assessment This includes Report modes (Flato of basedNorth Pacificred noise decadal process variability involving has the recently interaction been between extratropical atmospheric variability and relevance to the SH: ENSO, the Indian-Ocean Dipole and the Quasi-Biennial Oscillation. Models have improved in
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 32 some respects since the last generation although, in the entirely for the right reasons. inconsistencieslinked to intermittent between major reconstructions, volcanic eruptions models (Hope and case of ENSO, some of this improvement might not be instrumentalet al., 2016). indices Nevertheless, used to Hopeevaluate et changes al. (2016) over report past While models are extremely valuable tools that enable and simulating past changes of a highly variable coupled system.centuries, In reflectingthe reconstructions, the complexity variability of reconstructing arises from us to improve our understanding of SH MDCV, they are internal climate processes, different forced responses, modelsnot without tend their to underestimatelimitations. For the example, magnitude while ofCMIP5 SST or non-climatic proxy processes that are still not well models tend to capture the spatial pattern of the IPO, understood (Ault et al., 2013). variability associated with the IPO (Power et al., 2016; andHenley multidecadal et al., 2017), changes and both in the the strength magnitude of the of WalkerDCV in assisted by the availability of additional data from high trade winds (England et al., 2014; McGregor et al., 2014) latitudeImproving regions. the simulation The palaeoclimate of SH MDCV reconstruction would be greatly of the SAM by Abram et al. (2014) displays good agreement with tendCirculation to be too (Kociuba oscillatory and Power, on too 2015). short Thesea time-scale. deficiencies This CMIP5–PMIP3 last millennium simulations. Although makesappear itto hard be due, for theat least models in part, to maintain to modeled multi-year ENSOs thatand the reconstruction shows a progressive shift towards et al., 2016). the positive trend in the SAM since 1940 is reproduced longer-term anomalies (Kociuba and Power, 2015; Power bySAM's multi-model positive phaseclimate as simulations early as the forced fifteenth with century, rising A further illustration of the limitations of CMIP5 models greenhouse gas levels and ozone depletion (Abram et al..
the instrumental data from high latitudes, and associated temperaturein simulating from SH MDCV pre-industrial is seen in runs Fig. of1. CMIP5It shows models. time- 2014). These results are likely to reflect the brevity of Theseries model-to-model of the decadal range variability in the in magnitude SH surface and air et al., 2016). character of the variability is remarkable. Some models deficiencies in model representation of SH climate (Jones exhibit variability that has a range of a few tenths of a degree, while the range of some other models is three or more times larger. The character of the variability also differs markedly among models. In some models there is pronounced multi-decadal variability, whereas in other models variability occurs on much shorter time-scales. While further research is needed to ascertain which of
Clearly, given the large model to model differences these simulations more realistically captures SH MDCV. theevident current in Fig.1, generation care is ofneeded climate in assigningmodels. confidence to conclusions drawn on the basis of SH MDCV simulated in of the past 1000 years in climate model simulations and There have been other studies investigating SH MDCV reconstructions For example, a recent study by Hope et al. (2016) examined the decadal characteristics of ENSO Indexspectra calculated based on fromseven six published CMIP5–PMIP3 ENSO reconstructions,last millennium simulations.and indices of The Nino post-1850 3.4 SSTs andspectrum the Southern of each Oscillation modelled or reconstructed ENSO series captures the observed instrumentalspectrum to varyingspectral degrees. characteristics However, across no thesingle multiyear model or decadal ENSO bands. reconstruction completely reproduces the variability is observed in the reconstructions and Figure 1: Time series of the (detrended) 10-year running simulationsAppreciable of changes the pre-1850 in the levelperiod. of While decadal much ENSO of annual mean surface air temperature, averaged over the SH this represents internally generated variability (see from the CMIP5 models (K), ordered from smallest to largest. e.g. Power et al., 2006), some of the variability may be Values calculated from the first 200 years of pre-industrial experiments offsets for display purposes.
33 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Predictability of MDCV in the SH In climate science the term “predictability” has a different to Australian climate. More recently, Wittenberg et al. meaning to “predictive skill”. Predictability provides (2014)in ENSO concluded activity andthat itspotential associated predictability teleconnections was an estimate of the upper limit to predictive skill, in the absence of technical problems - apart from uncertainty years ahead, but not on decadal time-scales. Power in initial conditions. Predicability is usually estimated andevident Colman in ENSO(2006) variability also concluded in their that model the relative several from a model’s ability to predict its own evolution given imperfect initial conditions. The presumption is that the results from a well-behaved climate model can andinfluence more pronounced of decadal variabilityas depth and on latitude ocean variabilityincreased. provide information on the predictability of the actual Thisas a general whole tendedpattern tois beconsistent weak in with the the tropical more Pacificrecent climate system. Predictability arises from internal variability, external forcing (e.g. increasing greenhouse analysis of predictability in SST (Frederiksen et al. 2016). estimate of the relative importance of initial conditions andgas concentrations),external forcing and for interactions the predictability between them.of time- One additionalaveraged temperaturepotential skill in on the average Southern over Hemisphere the southern is hemispheredepicted in Fig.on all2 (dashed time-scales, lines). from Initialisation one month provides to 10 years. The relative contribution of initialisation to the potential skill tends to diminish the longer the time- scale, as the relative contribution from external forcing increases.
forcing,The estimated with internal potential variability predictability making of five-year comparatively means across the SH is dominated by the contribution of external withlittle thecontribution possible exception(Kirtman ofet theal., 2013,east Antarctic Fig. 11.1). region This Figure 2: The relative importance of initial conditions and southis consistent of Africa. with Whether the findings or not of this Meehl represents and Hu a (2010), robust external forcing for the climate prediction and predictability of time-averaged temperature in the SH. The correlation exception remains unclear. skill scores for the ensemble mean of initialized temperature forecasts, and the model-based "potential" correlation skill Additional studies have examined the regional scores, area averaged over the SH, are presented as orange lines. The same quantities, but for uninitialized climate experiments in which model conditions are perturbed at simulations, are plotted as green lines. All lines represent apredictability particular point of in MDCV time in inan integration the SH using and the idealised degree area-averaged SH values. Solid lines: hindcast skill; Dashed to which the ensuing variability is affected is assessed. lines: potential skill. Results are for temperature averaged Power and Colman (2006) used this strategy to assess over periods from a month to a decade. (Figure prepared the predictability on internally generated variability by V.V. Kharin based on the results reported in Boer et al. in their climate model. They found that off-equatorial (2013)).
stresses also drives subsequent and therefore predictable regions in the South Pacific exhibited variability that was The excitation of Rossby waves from ENSO-driven wind- a delayed, low pass-filtered version of preceding ENSO variability. The ocean had acted as a low pass filter on laggedsea-level associations variability (e.g.,with Luo,extra-tropical 2003; Qiu regionsand Chen, and 2006; the the wind-stress and heat flux forcing it received, in an Holbrook et al., 2011; 2014). Other studies have examined 1977).ENSO-modified At ocean depths Frankignoul of 300 andm, this Hasselmann gave rise to process highly changes in both the ocean and atmosphere may provide predictable(Hasselmann, multi-year 1976; Frankignoulvariability. This and is Hasselmann, consistent atropics, source with of predictability results suggesting for multi-year that South variability Pacific driven in the showed that the leading mode of SST variability in with the findings of Shakun and Shaman (2009), who tropics (McGregor et al., 2007, 2008, 2009a and b; Luo meridionalet al., 2003; modeTatebe in et idealisedal., 2013; climateZhang et simulations, al., 2014). For in the South Pacific could be reasonably well-simulated example, Zhang et al. (2014) identified a South Pacific as a response to preceding ENSO-driven heat flux sub-component of subsequent climatic variability in the forcing, in analogy with the Pacific Decadal Oscillation tropics,which wind while variability the studies in the of SouthMcGregor Pacific et al.underpins highlight a Powerin the et North al. (2006) Pacific used (Newman the same strategy et al., and 2003; concluded 2016). ocean links between the extra-tropics and the tropics and that there is limited predictability in interdecadal changes their role in driving climate variability in the tropics.
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 34 More recently, the re-emergence of remnant mixed layers prediction skill. The advent of the satellite era increased thea difficulty amount of for available the assessment information, of but SST since variability that source and to the surface has been identified as a possible source of of data only exists since the late 1970s or early 1980s, predictability for SH MDCV (D. Dommenget, pers. comm.). most of the SST hindcasts prior to the 1970s cannot be validated against ground truth, thus decreasing the byFinally, partially predictability predictable may Atlantic arise in surfacethe Pacific, temperature including variabilitythe South Pacific, (e.g. Rashid via atmospheric et al., 2010). teleconnections driven tests applied therein. In an assessment of the decadal climatesample prediction size and significance skill over southern of most Africa, of the Reason statistical et Prediction of SH MDCV al. (2006) had already noticed how the lack of data and Unlike “predictability” assessment, a “prediction” is an the decrease in density and quality of the information in estimate or collection of estimates of the future state of several regions in Africa was a major concern. the real world. Research on decadal climate prediction The relative importance of initial conditions and external forcing for the skill of existing hindcasts of time-averaged basis(Smith for et al.,producing 2007; Meehl such et predictions al., 2009; Meehl is given et al., by 2016) the predictabilityaims to provide that forecasts exists ofin MDCV.some areasPart of and the forscientific some variables, as discussed in the previous section. MDCV temperature in the Southern Hemisphere is depicted ofin up Fig. to 2three (solid years, lines). beyond Initialisation which initialisation provides additionalin current Phase of the Coupled Model Intercomparison Project systemshindcast skillappears over tothe have Southern little Hemisphere impact. The for difference averages (CMIP5)prediction (Taylor research et al., was 2012) an important and was element a subject of assessed the Fifth between the actual skill scores (solid lines) and the potential skill scores (dashed lines) suggests that greater al., 2013). CMIP5 provided a new set of forecasts that skill may be achieved as the technical issues associated werein the initialized Fifth Assessment using the Report observed of the state IPCC of (Kirtmanthe climate et with MDCV prediction are overcome. initialization by comparison with parallel experiments Among other recent contributions to this topic, Rea wheresystem. no This information allowed anon the assessment initial state of theof the benefit climate of is important since near-term climate predictions are ofet al.,the (2016)stratospheric quantified processes prediction and qualitythe stratosphere- over high affectedsystem was by provided.the initial Theconditions quantification as well of as such by changes troposphereSH latitudes, coupling concluding is thatcrucial a proper in order representation to obtain skillful predictions. Saurral et al., (2016) analyzed the in the external forcing (Meehl et al., 2009; Kirtman et al., Initialization,2013; Meehl and which Teng is 2016). accomplished using a range of noticedinfluence differences of initialization in the andhindcast climate skill drift depending on hindcasts upon of SST in the South Pacific in a set of coupled GCMs. They producedifferent ensembles techniques of (e.g. simulations Hawkins which and Sutton,are intended 2009, initialization under strong or weak ENSO conditions. to2011; account Mochizuki for uncertainties et al. 2010; in Doblas-Reyes the initial conditions. et al., 2011), As (LienertOverall, and their Doblas-Reyes, results showed 2013). lower skill in decadal with any other forecast, a set of skill scores is usually predictions than that found for the North Pacific basin computed to quantify the quality of such predictions for The increased number of observations that are being previous years (i.e. prediction of past years or hindcasts), providing a quantitative assessment of the performance surface through e.g. the Global Temperature and Salinity of the different forecast systems. made in the SH, particularly of the ocean’s surface and sub- of the global observing network, will provide invaluable In spite of the relative youth of this research area, dataProfile for Programme, the assessment ARGO, of predictability and other critical and prediction elements there are a growing number of papers that address skill in coming years. At the same time, modelling efforts predictability and prediction skill over areas of the focused on improving the representation of processes underpinning decadal variability and initial conditions
Northern Hemisphere (e.g. Griffies and Bryan, 1997; Boer, 2000; Collins, 2002; Hawkins and Sutton, 2009; Smith et seaare alsoice concentrationof great importance and thickness for the field. reanalysis The prediction and the al., 2010; Zanna, 2012; García-Serrano and Doblas-Reyes, of SH MDCV could benefit from the improvements in 2012; Boer et al., 2013; Guémas et al., 2013; among many upcoming Year of Polar Prediction could prove a very others). However, considerably less attention has been interestingingestion of opportunity those data intoto gatherthe GCMs. new For information instance, theon comparedpaid to the to SH. the Partnorthern of the hemisphere. explanation Most is likely importantly, related sea ice in order to improve understanding of high-to-mid thereto the issmaller a relative scientific lack of community in situ climate that recordsexists in in the most SH latitude linkages and to see how these new observations impact the skill of MDCV predictions. of the SH, with the exception of some ship routes in the Indian and Atlantic Oceans. Limited in situ data poses
35 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Making further progress Climate Change, 4, 564–569. predictOur understanding its behavior ofcould the causesbe increased and predictability by improving, of e.g.,MDCV : in the Southern Hemisphere and our ability to Allan, R., Endfield, G., Damodaran, V., Adamson, G., • the documentation, quality controlling and analysis Hannaford, M., Carroll, F., Macdonald, N., Groom, N., Jones, of MDCV evident in instrumental records, other integratedJ., Williamson, historical F., Hendy, climate E., Holper, research: P., Arroyo-Mora, the example J. P., historical records (see, e.g. Callaghan and Power, ofHughes, Atmospheric L., Bickers, Circulation R. and Bliuc,Reconstructions A.-M., 2016: over Toward the Earth. WIREs Climate Change, 7, 164-174, doi: 10.1002/ • the quality and increasing the range of relevant wcc.379. instrumental,2011; 2014), and historical paleo recordsand paleoclimate records. • our understanding of the impact/sensitivity of Arblaster, J.M., and N.P Gillett (Lead Authors), N. Calvo, methods used for multi-proxy paleoclimate records in order to develop methods that are not sensitive to Young, 2014: Stratospheric ozone changes and climate, the non-stationarity of teleconnections. P.M. Forster, L.M. Polvani, S.-W. Son, D.W. Waugh, and P.J. • MDCV, including the underlying dynamical processes Chapter 4. Scientific Assessment of Ozone Depletion: climate models and their simulation of observed SH Geneva,2014, Global Switzerland. Ozone Research and Monitoring Project events (e.g., strengthening the of Walker circulation – Report No. 55, World Meteorological Organization, overand statisticsthe past fewof SH decades MDCV and and the particular rate of warming important in Ault, T. R., C. Deser, M. Newman, J. Emile‐Geay, 2013: Characterizing decadal to centennial variability in the • processthe Pacific studies, over the by continuingpast half-century) to develop theoretical Research Letters, 40, 3450-3456. understanding ofof the the causes relevant of SH processes MDCV through involving e.g. equatorial Pacific during the last millennium. Geophysical both observations and model simulations, using a Bateup, R., S. McGregor and A. J. E. Gallant, 2015: hierarchy of modelling approaches • understanding of MDC predictability and its underlying dynamics in models, through further aThe pseudoproxy influence framework, of non-stationary Clim. Past, teleconnections 11, 1733-1749, on experimentation and further analysis of experiments doi:10.5194/cp-11-1733-2015.palaeoclimate reconstructions of ENSO variance using already conducted • initialisation methods, and predictive systems more Bindoff, N.L., P.A. Stott, K.M. AchutaRao, M.R. Allen, N. broadly • our ability to reliably determine hindcast skill. 2013:Gillett, Detection D. Gutzler, and K. Hansingo,Attribution G. of Hegerl, Climate Y. Hu,Change: S. Jain, from I.I. These objectives can be assisted by, e.g., : GlobalMokhov, to J.Regional. Overland, In: J. Climate Perlwitz, Change R. Sebbari 2013: and The X. Physical Zhang, • participation in CMIP6, the Decadal Climate Science Basis. Contribution of Working Group I to the Prediction Project and other relevant international initiatives • maintaining and expanding observing networks to Tignor,Fifth Assessment S.K. Allen, Report J. Boschung, of the IntergovernmentalA. Nauels, Y. Xia, V. Panel Bex andon Climate P.M. Midgley Change (eds.)].[Stocker, Cambridge T.F., D. Qin, University G.-K. Plattner, Press, M. • continuing to foster links between paleoclimate, Cambridge, United Kingdom and New York, NY, USA. climatebetter monitor modelling SH andMDCV model analysis communities • digitising and quality controlling early instrumental Boer, G. J., 2000: A study of atmosphere-ocean predictability on long time scales. Climate Dyn., 16, 469– et al., 2016), and by 477. • collectingrecords from critical SH locations palaeoclimate (Allan et data al., 2016;in key Freeman regions (e.g. Abram et al., 2015). predictability and forecast skill. Clim. Dyn., 41, 1814- References 1833,Boer, G. doi:10.1007/s00382-013-1705-0. J., V. V. Kharin, and W. J. Merryfield, 2013: Decadal Abram, N. J., Dixon, B. C., Rosevear, M. G., Plunkett, B., Gagan, Boer, G., and Coauthors, 2016: The Decadal Climate Prediction Project (DCPP) contribution assessmentM. K., Hantoro, of W.location S. and Phipps,and length S. J., 2015:considerations. Optimized to CMIP6. Geosci. Model Dev., 9, 3751-3777. Paleoceanography,coral reconstructions 30 (10), of the 2015PA002810. Indian Ocean Dipole: An Cai, W., T. Cowan, J. M. Arblaster, and S. Wijffels, 2010:
heat content trend in CMIP3 models, Geophys. Res. Lett., AnnularAbram, N. Mode J., Mulvaney, during R., the Vimeux, past F.,millennium. Phipps, S. J., Turner,Nature 37,On potentialL17709, doi:10.1029/2010GL044399. causes for an under-estimated global ocean J. and England, M. H., 2014: Evolution of the Southern
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 36 in southeastern Australia 1860–2012, associated deaths and the ongoing warming hiatus. Nature Climate Callaghan, J., and S. B. Power, 2014: Major coastal flooding Change,intensification 4, 222-227, of wind-driven doi: circulation10.1038/nclimate2106 in the Pacific 213. and weather systems. Aust. Meteor. Oceanogr. J., 64, 183– P. J. Young, S. Bekki, D. Bergmann, P. Cameron-Smith, W. J. Eyring, V., J. M. Arblaster, I. Cionni, J. Sedláček, J. Perlwitz, J.K.Christensen, Kanyanga, J.H., A. K.Kitoh, Krishna J. Kossin, Kumar, N.-C. E. Aldrian,Lau, J. Renwick, S.-I. An, I.F.A. Cavalcanti, M. de Castro, W. Dong, P. Goswami, A. Hall, D.Collins, T. Shindell, G. Faluvegi, K. Sudo, K.-D. S. Gottschaldt,Szopa and S. L. Watanabe, W. Horowitz, 2013: D. Long-termE. Kinnison, ozone J.-F. Lamarque, changes and D. associatedR. Marsh, D.climate Saint-Martin, impacts ClimateD.B. Stephenson, Change. In: S.-P. Climate Xie and Change T. Zhou, 2013: 2013: The Physical Climate in CMIP5 simulations, J. Geophys. Res. Atmos., 118, 5029– SciencePhenomena Basis. and Contribution their Relevance of Working for Future Group RegionalI to the 5060, doi:10.1002/jgrd.50316.
Tignor,Fifth Assessment S.K. Allen, Report J. Boschung, of the IntergovernmentalA. Nauels, Y. Xia, V. Panel Bex interdecadal variability driven by tropical-extratropical andon Climate P.M. Midgley Change (eds.)].[Stocker, Cambridge T.F., D. Qin, University G.-K. Plattner, Press, M. interactions.Farneti, R., F. Climate Molteni, Dynamics, and F. Kucharski,42(11–12), 2014:3337–3355. Pacific Cambridge, United Kingdom and New York, NY, USA. decadal-scale impact of volcanic eruptions on sea level Folland, C. K., J. A. Renwick, M. J. Salinger, and A. B. Mullan, andChurch, ocean J., heatN.J. White content. and Nature, J.M. Arblaster, 438, 74-77, 2005: doi:10.1038/ Significant 2002: Relative influences of the Interdecadal Pacific nature04237. Oscillation and ENSO on the South Pacific Convergence Zone. Geophys. Res. Lett., 29, 1643-1647. Collins, M., 2002: Climate predictability on interannual to climate models, Part II Application to sea-surface decadal time scales: The initial value problem. Clim. Dyn., temperatureFrankignoul, C.,anomalies and K. Hasselmann,and thermocline 1977: variability. Stochastic 19, 671–692. Tellus, 29, 10.1111/j.2153-3490.1977.tb00740.x.
Simulated modes of inter-decadal predictability in sea CSIRO, Australian Bureau of Meteorology and SPREP, surfaceFrederiksen, temperature. C.S., X. Climate Zheng, Dynamics, and S. Grainger, 46, 2231-2245. 2016: Climate2015: Climate Change in Science the Pacific: and a regionalAdaptation summary Planning of new science and management tools, Pacific-Australia DOI 10.1007/s00382-015-2699-6. Program Summary Report. Commonwealth Scientific and Kent, E. C., Angel, W. E., Berry, D. I., Brohan, P., Eastman, Industrial Research Organisation, Melbourne, Australia. Freeman, E., Woodruff, S. D., Worley, S. J., Lubker, S. J.,
Di Lorenzo, E., G. Liguori, N. Schneider, J. C. Furtado, ReleaseR., Gates, 3.0: L., a Gloeden, major update W., Ji, Z.,to Lawrimore,the historical J., Rayner,marine Variability.B. T. Anderson, Geophys. and M. Res. A. Alexander,Lett., 42, 2015:9440-9448, ENSO doi: and climateN. A., Rosenhagen, record. International G. and Smith, Journal S. R., of 2016: Climatology: ICOADS 10.1002/2015GL066281.meridional modes A null hypothesis for Pacific Climate doi:10.1002/joc.4775.
Ding, R., J. Li, and Y. Tseng, 2015: The impact of South
Garreaud, R.D. and D.S. Battisti, 1999: Interannual ENSO 10.1007/s00382-014-2303-5.Pacific extratropical forcing on ENO abnd comparisons 2113-2123and Interdecadal ENSO-like Variability in the Southern with the North Pacific. Climate Dynamics, 44, DOI: Hemisphere Tropospheric Circulation. J. Climate, 12,
T. N. Palmer, 2011: Decadal climate prediction with 2012: Understanding Atlantic multi-decadal variability Doblas-Reyes, F. J., M. A. Balmaseda, A. Weisheimer, and predictionGarcia-Serrano, skill. J., F.Geophys. J. Doblas-Reyes, Res. Lett., and C.39, A. S.L18708, Coelho, observations. J. Geophys. Res., 116, D19111. doi:10.1029/2012GL053283. the ECMWF coupled forecast system: Impact of ocean
of simulated North Atlantic multidecadal variability. Dong L., T. Zhou, A. Dai, F. Song, B. Wu, X. Chen, 2016: The Clim.Griffies, Dyn., S. M.,13, and459–487. K. Bryan, 1997: A predictability study 21251,Footprint doi:10.1038/srep21251. of the Inter-decadal Pacific Oscillation in Indian Ocean Sea Surface Temperatures. Scientific Reports, 6, M. Asif, 2013: Retrospective prediction of the global Meehl, A. Timmermann, W. Cai, A. Sen Gupta, M. J. warmingGuémas, V.,slowdown F. J. Doblas-Reyes, in the last I.decade. Andreu-Burillo, Nature Clim. and McPhaden,England, M.A. H.,Purich S. and McGregor, A. Santoso, P. Spence, 2014: Recent G. A. Change, 3, 649-653, doi:10.1038/nclimate1863.
37 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Theory. Tellus, 6, 473-485, 10.1111/j.2153-3490.1976. tb00696.x.Hasselmann, K., 1976: Stochastic climate models. Part I. Nauels,Intergovernmental Y. Xia, V. Bex Panel and P.M.on Climate Midgley Change (eds.)]. [Stocker, Cambridge T.F., UniversityD. Qin, G.-K. Press, Plattner, Cambridge, M. Tignor, United S.K. Allen, Kingdom J. Boschung, and New A. York, NY, USA. Kociuba G. and S. B. Power, 2015: Inability of CMIP5 AmericanHan, W., J. Meteorol. Vialard, M.Soc., J. 95, McPhaden, 1679-1703. W. P. M. de Ruitjer, Models to Simulate Recent Strengthening of the Walker 2014: Indian Ocean Decadal Variability: A Review. Bull. Circulation: Implications for Projections. J. Clim., 28, 20- 35. narrow uncertainty in regional climate predictions. Bull.Hawkins, Am. E., Meteorol. and R. Sutton,Soc., 2009:90, 1095-1107, The potential doi: to Kosaka, Y. and S.-P. Xie, 2013: Recent global-warming 10.1175/2009BAMS2607.1. 501(7467), 403–7. hiatus tied to equatorial Pacific surface cooling. Nature,
45(11–12),Henley, B.J. 3077–3090. and Coauthors, 2015: A Tripole Index for the Interdecadal Pacific Oscillation. Climate Dynamics, doi:10.1007/s00382-015-2705-zKucharski, F., Ikram, F., Molteni, F. et al., 2016: Atlantic forcing of Pacific decadal variability, Clim Dyn, 46: 2337. Agreement in Climate Model Simulations of the Henley, B.J. and Coauthors, 2017. Spatial and Temporal Letters, in press, 1–33. temperature.Lienert F., F. J.J. Doblas-Reyes,Geophys. Res. 2013:Atmospheres, Decadal 118, prediction 5913- Interdecadal Pacific Oscillation. Environmental Research 5922.of interannual tropical and North Pacific sea surface
Luo, J.-J., S. Masson, S. Behera, P. Delecluse, S. Gualdi, Holbrook N. J., I. D. Goodwin, S. McGregor, E. Molina, S. B. Power, 2011: ENSO to multi-decadal time scale changes mechanism.in East Australian Deep CurrentSea Research, transports Part andII,, 58,Fort 547-558, Denison aA. coupled Navarra, GCM. T. Yamagata, Geophys. 2003:Res. Lett., South 30 Pacific (24), origin2250, doi:10.1016/j.dsr2.2010.06.007.sea level: Oceanic Rossby waves as the connecting doi:10.1029/2003GL018649.of the decadal ENSO-like variation as simulated by
DecadalHolbrook climate NJ, J Li,variability M Collins, and E cross-scale Di Lorenzo, interactions: F-F Jin, T ExtratropicalMcGregor, S., A.Rossby Sen Gupta, Waves. N. J. J. Climate, Holbrook 22, and 5839-5383. S. B. Power, ICCLKnutson, 2013 M Expert Latif, C Assessment Li, SB Power, Workshop. R Huang and Bulletin G Wu, of2014: the 2009: The Modulation of ENSO Variability in CCSM3 by American Meteorological Society, 95, ES155-ES158. doi: 10.1175/BAMS-D-13-00201.1. observedMcGregor, off-equatorialS., N. J. Holbrook wind-stress and S. B. forcing. Power, 2009:J. Climate. The 2512-2525.response of a stochastically forced ENSO model to
Hope, P., Henley, B. J., Gergis, J., Brown, J. and Ye, H., 2016: s00382-016-3393-z.Time-varying spectral characteristics of ENSO over Interdecadal sea surface temperature variability in the the Last Millennium. Climate Dynamics: DOI: 10.1007/ McGregor, S., N. J. Holbrook, and S. B. Power, 2007: wind stresses and oceanic Rossby waves. J. Clim., 20 (11), D. J. Charman, K. R. Clem, X. Crosta, C. de Lavergne, I. 2643-2658equatorial Pacific Ocean. Part I: The role of off-equatorial Jones, J. M., S. T. Gille, H. Goosse, N. J. Abram, P. O. Canziani, G. J. Marshall, V. Masson-Delmotte, A. K. Morrison, A. J. Eisenman, M. H. England, R. L. Fogt, L. M. Frankcombe, Interdecadal sea surface temperature variability in the Simpkins, E. J. Steig, B. Stenni, D. Swingedouw, T. R. McGregor, S., N. J. Holbrook, and S. B. Power, 2008: Vance,Orsi, 2016:M. N. Raphael,Assessing J. A.recent Renwick, trends D. P. in Schneider, high-latitude G. R. off-equatorial wind stresses in a hybrid coupled model. J. Clim.,equatorial 21 (17), Pacific 4242-4256. Ocean. Part II: the role of equatorial/ Change, 6, 917–926. Southern Hemisphere surface climate. Nature Climate Kirtman, B., S.B. Power, J.A. Adedoyin, G.J. Boer, R. Bojariu, McGregor, S., A. Timmermann, M. F. Stuecker, M. H. Meehl, M. Prather, A. Sarr, C. Schär, R. Sutton, G.J. van England, M. Merrifield, F.-F. Jin, Y. Chikamoto, 2014: I. Camilloni, F.J. Doblas-Reyes, A.M. Fiore, M. Kimoto, G.A. Change,Recent Walker4, 888-892, Circulation doi:10.1038/NCLIMATE2330. strengthening and Pacific Climate Change: Projections and Predictability. In: Climate cooling amplified by Atlantic warming, Nature Climate ChangeOldenborgh, 2013: G. The Vecchi Physical and Science H.J. Wang, Basis. 2013: Contribution Near-term of
Working Group I to the Fifth Assessment Report of the
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 38 Meehl, G. A., J. M. Arblaster, C. M. Bitz, C. T. Y. Chung, and
Rashid, H., S.B. Power, and J. Knight, 2010: Impact of variability.H. Teng, 2016: Nature Antarctic Geoscience, sea-ice doi:10.1038/ngeo2751. expansion between GCM.multi-decadal J. Climate, fluctuations 23, 4038-4044. in the Atlantic thermohaline 2000 and 2014 driven by tropical Pacific decadal climate circulation on Indo-Pacific climate variability in a coupled prediction for transition to positive phase of the 2016: Multi-decadal changes and predictions over the Meehl, G.A., A Hu, H Teng, 2016: Initialized decadal Rea G, C. Cagnazzo, A. Riccio, F. Fierli, and F. Cairo, 7, doi:10.1038/ncomms11718. representation in CMIP5 models. EGU General Assembly Interdecadal Pacific Oscillation, Nature Communications, ConferenceSouthern Hemisphere Abstracts, region: Vienna, role ofAustria, the stratospheric European Geosciences Union. hindcasts relevant to near-term climate prediction. Proc, Natl.Mochizuki, Acad. Sci., T., et107, al., 1833–1837. 2010: Pacific decadal oscillation Reason CJ, W Landman, and W Tennant, 2006: Southern African climate and its links with variability of the Newman, M., G. P. Compo, and M. A. Alexander, 2003:
J. Climate, 16, 3853–3857, doi:10.1175/1520-0442. Salinger,Atlantic Ocean. M.J, J.A. Bull. Renwick, Amer. Meteor. and Soc.,A. B. 87, Mullan, 941-955. 2001: ENSO-forced variability of the Pacific Decadal Oscillation, Int. J. Climatol., 21, 1705-1721. Interdecadal Pacific Oscillation and South Pacific Climate. org/10.1175/JCLI-D-15-0508.1Newman, M., et al. 2016: The Pacific Decadal Oscillation, revisited. J. Climate, 29, 4399-4427, DOI: http://dx.doi. 2016: Initial shock, drift and trends in decadal climate Saurral R. I., F. J. Doblas-Reyes, and J. García-Serrano,
P.,Palmer, 2015: J.,Drought Cook, variability E., Turney, in C.,the Allen, eastern K., Australia Fenwick, and P., initialpredictions shock, in drift the and South bias Pacific adjustment. region 9-13 upon May different 2016, Cook, B., O’Donnell, A., Lough, J., Grierson, P. and Baker, Barcelona,phases of ENSO.Spain. SPECS/PREFACE/WCRP Workshop on
EnvironmentalNew Zealand summer Research drought Letters, atlas 10 (12), (ANZDA, 124002. CE 1500– Shakun, J.D., and J. Shaman, 2009: Tropical origins of 2012) modulated by the Interdecadal Pacific Oscillation. Lett., 36, L19711, doi:10.1029/2009GL040313. North and South Pacific decadal variability. Geophys. Res. Australia.Power, S. B.,Climate T. Casey, Dynamics, C. Folland, 15, A.319-324. Colman and V. Mehta, Shiotani, M., 1990: Low-frequency variations of the zonal 1999: Inter-decadal modulation of the impact of ENSO on Power, S. B. and R. Colman, 2006: Multi-year predictability Meteorol. Soc. Jpn., 68, 461–471. in a coupled general circulation mean state of the Southern Hemisphere troposphere, J. model. Climate Dynamics, 2, 247-272.
temperatureSmith, D. M., prediction S. Cusack, for A. W.the Colman, coming C. decade K. Folland, from G.a globalR. Harris, climate and model. J. M. Science, Murphy, 317, 2007: 796–799. Improved surface teleconnections.Power, S., M. Haylock, J. Climate, R. Colman, 8, 2161-2180. and X. Wang, 2006: The predictability of interdecadal changes in ENSO and ENSO
Inability of climate models to simulate recent multi- multi-yearSmith, D. M.,predictions R. Eade, N.of J.Atlantic Dunstone, hurricane D. Fereday, frequency. J. M. Power, S., F. Delage, G. Wang, and G. Kocuiba, 2016: NatureMurphy, Geosci., H. Pohlmann, 3, 846–849. and A. A. Scaife, 2010: Skilful implications for global temperature projections. Climate Dynamics,decadal changedoi:10.1007/s00382-016-3326-x. in Pacific surface temperature: Taylor, K. E., R. J. Stouffer, and G. A. Meehl, 2012: An overview of CMIP5 and the experiment design. Bull. Am. Power, S. B. and J. Callaghan, 2016: Variability in Severe Meteorol. Soc., 93, 485–498. in Southeastern Australia since the Mid–Nineteenth Tierney, J. E., Abram, N. J., Anchukaitis, K. J., Evans, Century.Coastal Flooding,Journal of Applied Associated Meteorology Storms, andand Climatology, Death Tolls 55, 1139-1149. temperaturesM. N., Giry, C., for Halimeda the past Kilbourne,four centuries K., Saenger,reconstructed C. P., fromWu, H.coral C. andarchives. Zinke, Paleoceanography, J., 2015: Tropical 30, sea-surface 226-252, doi:10.1002/2014PA002717. Qiu, B., and S. Chen, 2006: Decadal Variability in the 1751-1762.Large-Scale Sea Surface Height Field of the South Pacific Ocean: Observations and Causes. J. Phys. Oceanogr., 36, Annular modes in the extratropical circulation, part 2, Thompson, D. W. J., J. M. Wallace, and G. C. Hegerl, 2000:
39 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Trends, J. Clim., 13, 1018–1036.
Tozer, C. R., T. R. Vance, J. L. Roberts, A. S. Kiem, M. A. J. Curran, and A. D. Moy, 2016: An ice core derived 1013- year catchment-scale annual rainfall reconstruction
System Science, 20 (5), 1703-1717. in subtropical eastern Australia. Hydrology and Earth Vera, C., 2003: Interannual and interdecadal variability of atmospheric synoptic-scale activity in the Southern doi:10.1029/2000JC000406. Hemisphere. J. Geophys. Res: Oceans, 108 (C4), 8077,
Ishii, and M. Kimoto, 2014: Nature Climate Change, 4, 893–897.Watanabe, M., H. Shiogama, H. Tatabe, M. Hayashi, M.
Watterson, I. G., 2009: Components of precipitation and temperature anomalies and change associated with
809–826. doi:10.1002/joc.1772. modes of the Southern Hemisphere. Int. J. Climatol., 29: changes in the ventilation of the southern oceans, Science, 339Waugh, (6119), D.W., F.568-570, Primeau, T.doi: DeVries, 10.1126/science.1225411, and M. Holzer, Recent 2013
Wittenberg, A.T., A. Rosati, T.L. Delworth, G.A. Vecchi, org/10.1175/JCLI-D-13-00577.1.F. Zeng, 2014: ENSO Modulation: Is it decadally predictable? J. Climate, 27, 2667-2681, doi: http://dx.doi.
Dynamics,Wu, A., and 21, W. 719-730. W. Hsieh, 2003: Nonlinear interdecadal changes of the El Nino-Southern Oscillation. Climate Yuan X. and E. Yonekura, 2011: Decadal variability in the doi:10.1029/2011JD015673. Southern Hemisphere, J. Geophys. Res., 116, D19115, observed Atlantic sea surface temperatures. J. Climate., 25,Zanna, 5047–5056. L., 2012: Forecast skill and predictability of
Variability.Zhang, H., A.J. Climate,Clement 27, and 769-783. P. Di Nezio, 2014: The South Pacific Meridional Mode: A Mechanism for ENSO-like
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 40 Initialization Shock in CCSM4 Decadal Prediction Experiments doi: 10.22498/pages.25.1.41 Haiyan Teng, Gerald A. Meehl, Grant Branstator, Stephen Yeager, Alicia Karspeck National Center for Atmospheric Research, Boulder, USA Introduction Initialized predictions tend to drift away from the (Kalnay et al., 1996). This forced ocean/ice simulation initial states towards the model’s imperfect climatology. represents the NCAR contribution to the If all predictions drift in a coherent manner that is independent of an initial state, the drift to a large extent can be removed during the posterior bias correction. forCoordinated atmosphere/land Ocean-ice are Reference taken Experimentsfrom CCSM4 phase CMIP5 II poses a serious challenge to the quality of ocean (CORE II) (Danabasoglu et al., 2014). The initial states However, lack of continuous global ocean observations 1st during 1955-2014, we ran a 10-member ensemble incoherent initial errors especially in the Tropics can uninitializedof initialized historical/RCP4.5hindcast experiments runs. with For the each ensemble January initial states that span several decades. Furthermore, spread created by perturbing the atmosphere (or both and can impact the entire globe through atmospheric atmosphere and land in the earlier set as documented by teleconnections.be amplified by air-seaConsequently coupling the as effecta prediction of such evolves initial Yeager et al., 2012) initial conditions. bias correction methods and can overwhelm the There exists a large variety of bias correction methods relativelyshocks on small predictions decadal are signals difficult that to we remove seek toby predict. simple and they are designed to reduce errors and add skills to the forecasts. To avoid complexities that can arise decadal prediction experiments with the Community in assessing the source of improvements in calibrated ClimateHere, we System describe Model the version initialization 4 (CCSM4) shock and in adiscuss set of initialized hindcasts compared to the traditional climate the challenges they cause to near-term hindcasts, in change projection experiments without initialization (also referred to as the uninitialized simulations), here we examine interannual anomalies with respect to a CCSM4particular decadal of the predictionInterdecadal experiments Pacific Oscillation (IPO). hindcast climatology that is only a function of prediction CCSM4 is a fully coupled general circulation model consisting of atmosphere, ocean, land, and sea ice that calculation of anomalies (as in Doblas-Reyes et al., 2013). Thatrange; is, noat lead observations t (t=Month1, are 2, taken …, 120) into from account a start in year our j employed (Gent et al., 2011). The atmosphere model uses (j=year 1955, ..., 2014), the interannual anomalies are linked via a flux coupler and no flux corrections are resolution of 1°and 26 layers in the vertical. The ocean a finite volume dynamical core with a nominal horizontal where is the raw hindcast, and N is the total number nominal latitude-longitude resolution of 1° (tapering of start years. By using this methodology, the reference downis a version to 1/4° of inthe latitude Parallel in Ocean the equatorial Program (POP)tropics) with and a model hindcast climatology already 60 levels in the vertical. The land and sea ice components includes most model systematic errors (including share the same horizontal grids as the atmosphere and systemic errors in the forced response, assuming the ocean models, respectively. forced response is independent to the initial states). Presuming initial errors are the same for different start The CCSM4 decadal prediction experiments (also referred years, computing differences on the right hand side of (1) to as initialized hindcasts, or hindcasts) analyzed here will leave mostly the signal in the hindcasts. A caveat is are an expansion of a previously documented set (Yeager that the initial errors often vary with start years, which is et al., 2012) that was submitted to the Coupled Model even more true for decadal predictions than for seasonal Intercomparison Project phase 5 (CMIP5) (Taylor et al., predictions, as the former covers decades with limited 2012). The ocean/sea ice initial conditions are obtained ocean observations. from a CCSM4 ocean/sea ice stand-alone simulation forced with atmospheric variables, such as surface winds, Warm shocks in the Nino3.4 SST pressure, humidity etc. from the NCEP/NCAR reanalysis air temperature, precipitation, surface fluxes, sea level theFirst, time we evolution focus on ofsea raw surface Nino3.4 temperature SST during (SST) the 10-year in the Nino3.4 region for the hindcast behavior. Fig. 1a shows
41 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Year 3-7 is the forecast range that some current decadal a raw 10-member ensemble mean hindcast from a start prediction experiments aim to target (Meehl et al., yearhindcast indicated period; by with the eachlabel colored bar, and thin the line thick representing black line representing the observed climatology (http://www. global2014a, climate 2014b, (Kosaka 2016; and Meehl Xie, and2013), Teng, has limited 2012, 2014a,initial- long.data). Strikingly, the Nino3.4 SSTs from hindcasts value2014b). predictability The tropical for Pacific, all initial which states is a pace-makerat this range of initializedesrl.noaa.gov/psd/gcos_wgsp/Timeseries/Data/nino34. during the early decades (1955-1970, dark of 3-7 years based on a “perfect model” assessment blue) tend to produce a 2°C warm spike relative to the (Branstator and Teng, 2010). When we calculate the Nino3.4 SST interannual anomalies at the Year 3-7 range contrast there are no obvious warm spikes in the later startobserved years, climatology and a number in the of hindcastsfirst two orinitialized three years. around By with a downward trend until about 1980, and then stable or(Fig.1d), somewhat we stillupward find after secular that. multidecadal In some ways variability, this is three years. year 2000 produce cold spikes of roughly -2°C in the first We further examine the hindcasts of Nino3.4 SST at three thereminiscent amplitude of of the the Year Year 2 3-7drifts anomalies (Fig. 1c). is Note much there weaker. is a different lead times by comparing predicted interannual change in the range of the y-axis from Fig.1bc to Fig.1d; What caused the shock? by subtracting the observed climatological mean. In anomalies defined by (1) to observed anomalies derived andFirst, the we correspondingexamine how closely January the observations. ocean initial conditions We focus consistencyMonth1 (Fig.1b), of the the initial hindcast states anomalies with the (red) observations. generally onreflect SST observations and subsurface by comparing temperature the Month1measured hindcasts by the Becausematch the we observationscannot effectively (black). remove This the initial reflects shock the that is most pronounced in Year 2 with the calculation in (1), the Year 2 annual mean Nino3.4 SST hindcast thermocline depth (20°C isotherm, Z20) in the equatorial anomalies show a pronounced multidecadal shift, with Pacific, with the observed SST obtained from HadISST more than 1°C anomalies in the early two decades and (Rayner et al., 2003) and the Z20 observations calculated butfrom a possible the 2009 change World in either Ocean the Database strength (Levitus or distribution et al., The observed anomalies do not have these multidecadal of2009). the initialOur main errors focus occurring is not climatologicalat about 1980 meanbiases that might variations-1°C anomalies suggesting in late 1990s that the and variations the early 2000s are examples (Fig. 1c). explain why the positive initial shock only occurs for of impacts from initialization shocks that cannot be start years between roughly 1955-1970. removed by simple bias correction techniques.
Figure 1: (a) Time evolution of raw Nino3.4 SST during the 10- year hindcast period, with each colored thin line representing a 10-member ensemble mean hindcast from a start year Figure 2: Start year vs. longitude distribution of Month1 indicated by the label bar, and the thick black line representing (top) and Year1 (bottom) errors in equatorial SST (5°S-5°N) the observed climatology. The hindcast SSTs are smoothed and 20°C isotherm depth (Z20, 2°S-2°N). Boundaries for by an unweighted 6-month running average. b-d) interannual the Nino3.4 region are outlined by the two vertical dashed Nino3.4 SST anomalies from the CCSM4 hindcasts (red) and lines. The observed SST and Z20 are obtained from HadISST observations (black) at three different forecast ranges: (b) (Rayner et al. 2003) and 2009 World Ocean Database. Month 1, (c) Year 2 and (d) Year 3-7.
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 42 The warm spikes in the Nino3.4 region in the early hindcast period do not seem to be directly caused in any obvious way by the SST initial errors, for those initial errors are generally negative (~ -1°C) during 1950-1970 and are actually cooler than the errors in start years deeper in the pre-1980 initial conditions compared with after 1980 (Fig. 2a). However, the Z20 errors are much years have a much warmer subsurface temperature bias the later period (Fig. 2b), indicating that the early start
(hence deeper Z20). This subsurface warm bias is further amplified during Year 1 (Fig. 2d) and appears to have Wepropagated further to diagnosethe surface the (Fig. interannual 2c). temperature tendency budget in the upper 100m in the Nino3.4 region:
where T is the depth averaged temperature anomaly,
net H 0, and Cp are three constants that denote the layer thickness (=100m, and Q is the net air-sea surface flux. , ρ Figure 3: Budget analysis for the mixed layer temperature and heat capacity, respectively. UETW-E and VNTS-N denote horizontalconsistent resultsand meridional are found forconvergence H=60m), water (differential density, tendency anomalies in the Nino3.4 region averaged in a) Month1-3 and b) Month 6-12 hindcasts. All terms are shown as at the west and east boundary, or the south and north interannual anomalies, including temperature tendency (red), surface flux (orange), horizontal (UET, blue) and meridional respectively, and WTT100m (VNT, green) convergence of temperature flux, temperature boundaryat the 100m of depth. the Nino3.4 Prime region) represents of temperature the interannual flux flux at the 100m depth (WTT, purple), and sum of SHF, UET, anomalies (1). The residual is the term vertical includes temperature tendencies flux VNT and WTT (black dashed). All curves are smoothed by the 10-year running averages. subgrid scale. from the diffusive temperature fluxes and fluxes from the Meanwhile, the corresponding 100m depth temperature
During the first three months, the positive interannual during the two periods is 25.0°C and 24.2°C, respectively; temperature tendency (Fig. 3a, red), which is contributing the former period is significantly warmer at the 98% to the warm spikes seen in Fig.1a, can be well explained confidence level than the later period. isby represented the first four by the terms black on dashed the right line). hand Both sidethe vertical of the advected into the mixed layer and the surface warming is tendency equation. (In Fig. 3a, the sum of the four terms fullyOnce established the warm subsurfacefor all start temperatureyears during anomalies1955-1975 are by Month 6 (not shown), the mixed layer warming is further temperature flux at the bottom of the layer (purple, Fig. positive3a) and thetemperature meridional tendency temperature anomalies flux convergenceduring the indicated by the Month 6-12 anomalous temperature 1955-1975(green, Fig. start 3a) years, make which positive are partially contributions compensated to the tendencyamplified during through 1955-1975 atmosphere-ocean being mainly coupling. produced This by is
the UET'w-E term, which is partially compensated by the other four terms on the right hand side of the temperature by a large negative zonal temperature flux convergence UET'w-E is expected (blue, Fig. 3a). from westerly wind anomalies induced by the warm velocity and temperature at the 100m depth contribute tendency equation (Fig. 3b). Positive More specifically, interannual anomalies in both vertical to the anomalously large WTT'100m during 1955-1975 in climatological zonal temperature gradient advected by theSST anomalous anomalies, zonal and incurrent a figure can not explain shown a large we findportion the don’t have reliable ocean observations to quantify the of the positive temperature tendency anomalies during each of the first three months (not shown). Although we observed decadal change in the equatorial upwelling with westerly velocities, the domain averaged vertical velocity at 100m surface wind anomalies are equatorial downwelling depth in the Nino3.4 region in Month 1 of the hindcasts 1955-1975anomalies and (red, meridional Fig. 3b). temperature Associated divergence which
make WTT'100m and VNT'S-N negative. tois the significantly last two decades higher (at(1995-2014), the 99% confidencewith the 20-year level) While the budget analysis can explain how the warm meanduring equal the to first 0.51 two m/day decades and 0.38 (1955-1974) m/day, respectively. compared spikes in the pre-1975 hindcasts are triggered and
43 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 forcing from the NCEP/NCAR reanalysis has played a spatial pattern closely resembles the El Nino composite keyamplified, role in it producing does not these single ocean out which initial atmosphericconditions. Pacific SSTs at 40°S-60°N, 120°E-70°W (Fig. 4a). While the Among more than a dozen models that participated in frequency. It shows a negative-to-positive transition in the CMIP5 decadal prediction experiments (Kirtman the(Zhang mid et1970s, al., 1997), and a positive-to-negativethe IPO time series hastransition much lowerin the et al., 2013), the Max Planck Institute Earth System Model (MPI-ESM) has a similar initial shock problem latter has been regarded as the main cause for the recent to that found in CCSM4. Their ocean initial conditions slow-downlate 1990s and in observed early 2000s global (black warming line in (EasterlingFigs. 4de). Theand were also produced by forcing the ocean-only model Wehner, 2009, Meehl et al., 2011, 2013). Motivated by the with the NCEP/NCAR reanalysis. Pohlmann et al. (2016) further pin down the origins of the shock in this model to a spurious change in the NCEP/NCAR surface wind IPO’s strong impact on global climate, and by the fact that stress over 150°W-120°W, 10°S-10°N, which is strongly predictableit contains a than midlatitude the tropical component SSTs (Teng (the andPacific Branstator, Decadal westward before the 1970s and eastward after 1980 Oscillation, PDO) (Mantua et al., 1997) that is more compared to other reanalyses. The anomalous trade wind trend in the NCEP/NCAR reanalysis may help decadal2011; Branstator prediction andefforts. Teng, 2010, 2012; Branstator et to explain the CCSM4 initialization shock because: a) al., 2011), predicting the IPO has become a goal for the through Ekman transport, the anomalously strong equatorial trade wind stress before the 1970s in the NCEP/NCAR reanalysis can directly induce anomalous investigation at NCAR parallel to that done at MPI, a new setequatorial of decadal upwelling prediction in the experimentsCCSM4 initial were states; produced b) in an from an ocean state reconstruction generated using a th Century Reanalysis, Compo differentno spurious wind pre-1975 field (20 subsurface warming in the new oceanet al. 2006) initial thanconditions. the prior Therefore, CORE-II it simulation. is likely that We those find especially warm subsurface initial conditions pre-1975 in CCSM4 are also caused by the NCEP/NCAR spurious wind stress. More importantly, our preliminary analysis of new hindcast experiments initialized from the ocean states produced by the 20th Century Reanalysis winds indicates no big initialization shock.
Impacts on the IPO hindcast Figure 4: (a) The observed IPO: SST anomalies regressed upon the IPO index (second EOF of 10-year low-pass filtered theOne near-term might hope forecast the model range can(e.g. adjust 3-5 years) quickly it is to still the HadISST during 1870-2016, time series shown as the black possibleinitialization to harvest shock some in the skill equatorial from modes Pacific or patterns so that that for line in (d) and (e). (b) and (c) regressed EOF1 SST anomalies have long predictability. Indeed, the CCSM4 experiments from CCSM4 Year 2 (b) and Year3-7 (c) hindcasts and the show considerable skill in predicting North Atlantic corresponding time series are represented by red thick line upper-ocean heat content and surface temperature up (ensemble mean) and orange shading (spread) in d) and e). et al., 2015), which is consistent with other models’ The initialization shock has an impact on Year 2 CMIP5to a decade near-term in advance climate (Yeager predictions et al., 2012;(Doblas-Reyes Karspeck
hindcasts of the IPO in CCSM4. This is shown by an modelet al., 2013;anomalies Kirtman computed et al., from2013). observations There also (Meehlis skill inet EOF analysis of the Year 2 annual mean SST anomalies hindcast simulations of the IPO SST pattern in terms of averageover the and Pacific spread Ocean is shown from by all the start red yearsthick line and and all ensemble members. The time series of EOF1 (ensemble al., 2014; Meehl et al., 2016). However, also consistent for 54% of the total variance) shows a downward raisingwith other the issue models, of how Fig. hindcast 1d indicates skill is prediction evaluated skill for trend,orange crossing shading from in Fig. positive 4d respectively, to negative andnear ityear accounts 1980. is low in the year 3-5 equatorial Pacific SST in CCSM4, While both the timing of the cross-over from positive to 1999) on the near-term time scales. the Interdecadal Pacific Oscillation (IPO) (Power et al., timenegative series (Fig. of 4d)the andYear the 2 Nino3.4 amplitude interannual of the regressed anomalies SST anomalies (Fig. 4b) in the Nino3.4 region resemble the The observed IPO pattern is often defined as the second empirical orthogonal function (EOF) of low-pass filtered shown in Fig. 1c, the regressed SST anomalies (Fig. 4b)
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 44 trend, crossing from positive to negative near year 1980. While both the timing of the cross-over from positive to in phase with the observed IPO such that those negative understatingtrend values of could the artificiallytime-dependent contribute characterization some skill inof timenegative series (Fig. of 4d)the andYear the 2 Nino3.4 amplitude interannual of the regressed anomalies SST initializationhindcasts for shock the negative in CCSM4 phase is thus of necessary the IPO. Ain betterorder anomalies (Fig. 4b) in the Nino3.4 region resemble the to correctly interpret the model skill. further demonstrate that the equatorial SST shock is associatedshown in Fig. mainly 1c, thewith regressed a trend in SST the anomaliesPC time series (Fig. with 4b) Acknowledgment
Portions of this study were supported by the regional mixeda basin-wide layer temperature IPO-like (or along ENSO-like) the Kuroshio pattern. extension Since we andWe thankGlobal Dr. Climate Holger Modeling Pohlmann Program for valuable (RGCM) comments. of the regiondon’t find in thesuch Month1 large cold hindcasts temperature from 1955-1970, anomalies inthese the cold anomalies are likely induced by the initialization U. S. Department of Energy’s, Office of Science (BER), teleconnections. Cooperative Agreement DE-FC02-97ER62402, and shock in the equatorial Pacific through atmospheric the National Science Foundation. SY acknowledges Climatesupport Variability from the Nationaland Predictability Oceanic and Program Atmospheric grants Administration (NOAA) Climate Program Office under representsWe repeat a the downward EOF analysis trend and for a the somewhat Year 3-7 different annual SY and AK are partly funded by the National Science timingmean SSTof anomaliesthe positive (Figs. to 4cnegative and e) change and the compared PC1 also NA09OAR4310163 and NA13OAR4310138, and both to Year 2, with the change occurring somewhat later Foundation (NSF) Collaborative Research EaSM2 grant year smoothing and the longer hindcast range (since the ComputationalOCE-1243015. Computingand Information resources Systems were providedLaboratory by around 1990-95 (Fig. 4e). This could be due to multi- National Center for Atmospheric Research (NCAR)'s year average range rather than initialization year, e.g. a Computing Center (NERSC) under Contract No. DE- hindcastx-axis represents for 1986-1990 the central is represented prediction on theyear plot of asthe 1988 five (CISL), and by the National Energy Research Scientific and was initialized in 1984). At this longer range, the SST AC02-05CH11231and Oak Ridge Leadership Computing in the initialized hindcasts, have damped greatly in the Facility located in the National Center for Computational anomalies associated with EOF1, representing the trend BERSciences of the (NCCS) U.S. Department at Oak Ridge of NationalEnergy. NCAR Laboratory is sponsored under Contract DE-AC05-00OR22725, which are supported by persistenceequatorial Pacific, resulting but fromalmost a nodeeper damping mixed is foundlayer. inThus, the SST anomalies in the North Pacific, possibly due to long Referencesby the National Science Foundation. initialization shock represented by a trend that in some despite the fast adjustment in the tropical Pacific, the value decadal predictability in a CGCM. J. Climate, 23, higher latitude oceans. Thus, while the high persistence 6292-6311.Branstator, G. and H. Teng, 2010: Two limits of initial- ofways midlatitude resembles anomalies an IPO/PDO has thepattern potential may topersist give longerin the predictability to anomalies present in the midlatitude Branstator, G. and coauthors, 2011: Systematic estimates initial conditions, it may also degrade predictive skill as it enhances the persistence of anomalies produced by Climate, 25, 1827-1846. initialization shock. of initial-value decadal predictability for six AOGCMs. J.
initialization on decadal predictions as assessed for CMIP5 models.Branstator, Geophys. G. and Res. H. Lett., Teng, doi:10.1029/2012GL051974 2012: Potential impact of initializationOur results and shock those can indegrade other forecastsinvestigations and sometimes (Sanchez- Gomez et al., 2016; Pohlmann et al., 2016) indicate that Compo, G. P., J. S. Whitaker, and P. D. Sardeshmukh, 2006: even produce negative skill. For example, when trying pressure data. Bull. Am. Meteorol. Soc., 87, 175-190. to predict the negative IPO prior to the mid-1970s, the Feasibility of a 100-year reanalysis using only surface suggestsinitialization the shockinitial shownshock aswould a trend contribute in EOF1 ofto bothpositive the CLIVAR, 2011: Data and bias correction for decadal Year 2 and Year 3-7 interannual anomalies (Figs. 4d,e) climate prediction. CLIVAR Publication Series 150, forvalues this oftime the period IPO for relative that time to observations period. Therefore, (Meehl et skill al., in the positive IPO SST pattern in the CCSM4 hindcasts Danabasoglu,International CLIVARG. and Projectcoauthors, Office, 2014: 4pp. North Atlantic values of the initialization shock trend were taken into simulations in coordinated ocean-ice reference account.2014; 2016) There could are conceivably also occasions have been such higher as in the if positive 1990s when the trends in the Year 3-7 anomalies, which to a large extent may be caused by the initialization shock, are experiments phase II (CORE-II). Part I: mean states. Ocean Modelling, 73, 76-107.
45 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 term regional climate change prediction. Nature Comms, hindcasts for the mid-1970s shift and early 2000s hiatus doi:10.1038/ncomms2704 Doblas-Reyes F. J. and coauthors, 2013: Initialized near- andMeehl predictions G. A. and for H. 2016-2035. Teng, 2014a: Geophys. CMIP5 Res. multi-model Lett., doi: 10.1002/2014GL059256. doi:10.1029/2009GL037810.Easterling, D. R., and M. F. Wehner, 2009: Is the simulations for the mid-1970s shift and early-2000s climate warming or cooling? Geophys. Res. Lett., 36, hiatus.Meehl G.Geophys. A. and Res. H. Teng, Lett., 2014b:doi: 10.1102/2014GL061778. regional precipitation Gent, P. and coauthors, 2011: The Community Climate System Model version 4. J. Climate, 24, 4973-4991. decadal prediction for transition to positive phase of Kalnay, E and coauthors, 1996: The NCEP/NCAR 40-year Meehl, G. A., A. Hu and H. Teng, 2016: Initialized reanalysis project. Bull. Am. Meteorol. Soc., 77, 437-471. 10.1038/ncomms11718. the Interdecadal Pacific Oscillation. Nature Comm, doi: Karspeck, A. and coauthors, 2015: An evaluation of experimental decadal predictions using CCSM4, Climate in decadal hindcasts due to errors in wind stress over Dyn, 44, 907-923. Pohlmann, H. and coauthors, 2016: Initialization shock 3486-8. Kirtman, B. and coauthors, 2013: Near-term climate the tropical Pacific. Clim Dyn, doi: 10.1007/s00382-016- Power, S. and coauthors, 1999: Interdecadal modulation (eds) Climate Change 2013: the physical science basis, change: projections and predictions. In: Stocker T.F. et al. 319-324. report of the intergovernmental panel on climate change, of the impact of ENSO on Australia. Climate. Dyn., 15, Cambridgecontribution University of working Press, group Cambridge, 1 to the fifthpp 953-1028. assessment Rayner, N. A. and coauthors, 2003: Global analysis of sea surface temperature, sea ice, and night marine air Kosaka, Y. and S.-P. Xie, 2013: Recent global-warming temperature since the late nineteenth century. J. Geophys. Res., 108:4407, doi: 10.1029/2002JD002670. 501, 403-407. hiatus tied to equatorial Pacific surface cooling. Nature, Sanchez-Gomez, E. and coauthors, 2016: Drift dynamics Levitus, S. and coauthors, 2009: Global ocean heat in a coupled model initialized for decadal forecasts. content 1955-2008 in light of recently revealed Climate Dyn., 46, 1819-1840. instrumentation problem. Geophys. Res. Lett., doi: 10.1029/2008GL037155
subsurfaceTeng H. temperature and G. Branstator, in a CGCM, 2011:J. Climate, Initial-value 36, 1813- 1834.predictability of prominent modes of North Pacific Bull.Mantua, Amer. N. Meteor.J and coauthors, Soc., 78, 1069–1079.1997: A Pacific Interdecadal Climate Oscillation with impacts on salmon production. Taylor, K.E., R.J. Stouffer and G.A. Meehl, 2012: An Meehl, G. A. and coauthors, 2011: Model-based evidence overview of CMIP5 and the experiment design. Bull Am of deep-ocean heat uptake during surface-temperature Meteorol Soc, 93, 485-498. hiatus periods. Nat. Climate Change, 1, 360-364. Yeager, S, and coauthors, 2012: A decadal prediction case Meehl G. A. and coauthors, 2013: Externally forced study: late twentieth-century North Atlantic ocean heat and internally generated decadal climate variability content. J. Climate, 25, 5173-5189.
Climate, 26, 7298-7310. associated with the Interdecadal Pacific Oscillation. J. interdecadal variability: 1900-1993. J. Climate, 10, 1004- 1020.Zhang, Y., J. M. Wallace, and D. S. Battisti, 1997: ENSO-like
Geophys.Meehl G. A.Res. and Lett., H. Teng, doi: 10.1029/2012GL053423.2012: Case studies for initialized decadal hindcasts and predictions for the Pacific region. Meehl, G. A. and coauthors, 2014a: Decadal climate prediction: an update from the trenches. Bull. Am. Meteorol. Soc., 95, 243-267. model simulations of the observed early-2000s hiatus of globalMeehl, warming.G. A., H. Teng, Nat. Climateand J. M. Change, Arblaster, 4, 898-902. 2014b: Climate
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 46 Internal and forced decadal variability: lessons from the past millennium doi: 10.22498/pages.25.1.47 Hugues Goosse1, François Klein1, Didier Swingedouw2, Pablo Ortega3 1 ELIC/TECLIM Université catholique de Louvain, Belgium 2 Environnements et Paléoenvironnements Océaniques et Continentaux (EPOC), UMR CNRS 5805 EPOC—OASU—Université de Bordeaux, Allée Geoffroy Saint-Hilaire, Pessac 33615, France 3 NCAS-Climate, Meteorology Department, University of Reading, UK
Introduction variability in the estimated temperature changes over the The climatic observations over the instrumental era as circulation in the variability at decadal to centennial time cover a period too short to document the full range of scalelast millennium. and the impact Specific of the points forcing about on climaticthe role modesof oceanic are climatecommonly variability. defined The (from study roughly of paleoclimates C.E. 1850 to provides present) then presented. We conclude with some perspectives for a longer perspective allowing to explore the behavior of future developments. the climate system in a wider range of conditions and forcings. The most recent millennium is characterized by The warming during the 20th century compared to a climate very close to the present one and offers a large the pre-industrial climate amount of long paleo-climatic time series. This allows multiple applications, such as to characterize more millennium is that at the global scale climate was relatively precisely the decadal to centennial climate variations, stableA first before clear the conclusion industrial from era (especially the analysis as compared of the last to to test the robustness of potential mechanisms, or to abrupt climate variations during glacial period), although estimate the exact probability and return period of some reconstructions for the last millennium display some climate on which the anthropogenic forcing has imposed specific events. It can also be used to define a baseline warmsignificant (the regionalso-called fluctuations.medieval climate In many anomaly regions, around the for which we have a relatively large amount of paleo 950-1250),first centuries followed of the bysecond a colder millennium period (the were little relatively ice age, data,its imprint. with generally Fortunately, a low the uncertainty last millennium on the is datinga period of the records, and reasonable estimates of the changes consortium, 2013). Those warm conditions during the in external forcing. Consequently, it has received a lot medievalaround roughly climate 1450-1850) anomaly were (Mann less et al.,homogenous 2009; PAGES than 2k of attention over the last 20 years, allowing to obtain during the 20th century, which appears as the only one some crucial results for our understanding of the climate over the past centuries characterized by a clear warming system. over all the continents, except Antarctica (Mann et al.,
Paleoclimatology requires collaboration between with this conclusion and reproduce relatively well the communities as the records are coming from diverse main2009; characteristics PAGES 2k consortium, of reconstructed 2013). Model changes results (Masson- agree natural archives such as corals, tree rings, ocean and ice cores, speleothems, and others. Deciphering the climate signal in those archives requires a deep understanding of theDelmotte dominant et al.,processes 2013; PAGESresponsible 2k-PMIP3, for the 2015).continental This the physical, chemical and biological processes resulting reinforces our confidence in their ability to reproduce in the formation of those archives. This knowledge is to centennial timescales. essential for paleoclimate reconstructions but, in turn, and global-scale temperature fluctuations at interannual gives precious evidence of the impact of climate on Role of internal variability An emerging point from the analysis of last millennium always desirable so that inherent biases to each of the climate is the dominant contribution of the internal differentnatural systems. archives can Furthermore, be partly canceled multi-proxy out. Studying studies the are last millennium is thus a good opportunity to strengthen even in the absence of natural or anthropogenic forcing) the link between the groups focusing on climate changes variability (defined here simply as the one occurring
ofin the the natural recorded external fluctuations. forcing (solar Indeed, variations, internal volcanic climate and the ones working on systems influenced by climate. eruptions)variability at often the local overwhelms to regional completely scales in climate the influence model anthropogenicIn this short note, forcing, we briefly natural present forcing a few and illustrations internal focusing first on the relative contributions of simulations (Goosse et al., 2005; Jungclaus et al., 2010).
47 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Due to this important contribution of internal climate variability, ensembles of simulations, driven by the same external forcing but using different initial conditions, are required for meaningful comparisons between the results of one model and reconstructions and to disentangle the forced and unforced components of the simulated response (Goosse et al., 2005; Jungclaus et al., A2010; commonly Otto-Bliesner used etmetric al., 2016). to quantify the relative contribution of forced and internal variability in an ensemble of simulations is the ratio of the standard deviation of the ensemble mean over the mean standard deviation of the individual departures around this ensemble mean. The forced signal dominates in regions where this ratio is larger than 1, and the internally-driven signal when it is lower. These relative contributions can change depending on the timescale considered, as set of ten last millennium simulations from the climate illustrated in Fig. 1 for the 2-meter air temperature in a Figure 1: Ratio of the standard deviation of the ensemble while,model mainlyCESM1 in(Otto-Bliesner the tropics, aet larger al., 2016). contribution At interannual of the mean of an ensemble of 10 simulations performed with CESM1 responsetimescales, to internal the forcings fluctuations can be are found dominant at multidecadal everywhere (Otto-Bliesner et al., 2016) over the mean standard deviation of the ensemble members around this ensemble mean, for timescales (i.e. 50-year means). The surface average the 2-meter air temperature. The standard deviations are computed over the period C.E. 850-1850 for (top panel) means, decadal means and 50-year means are 0.45, 0.77 annual mean values (middle panel) decadal-means and andof the 1.25, ratios respectively. over the NorthernThe corresponding Hemisphere values for annualfor the (bottom panel) 50-year means. The ensemble mean is representative of the forced response while the range of the hemispheric mean temperatures (i.e., when the surface ensemble around this ensemble mean provides an estimate Southern Hemisphere are 0.44, 0.72 and 1.27. For the of the internal variability as simulated by the model. deviations), the relative contribution of the forced responseaverage is is computed much larger first with before ratios estimating of 1.13, 1.72 the standardand 2.53 volcanic forcing (Schurer et al., 2013). Major volcanic eruptions induce a global-scale cooling in the years for the Northern Hemisphere mean temperature and of the eruptions and their cumulative effects are also Many0.75, 1.58 studies and have2.66 foralso the emphasized Southern Hemisphere the role of internalmean. following an event. Furthermore, changes in the frequency difference between the medieval climate anomaly and responsible for a significant part of the temperature thoughvariability the inexternal recent forcing and futurewill likely changes be much (Hawkins larger 2015). inand the Sutton, future 2009; (compared Boer, 2011;to the 850-1850Deser et al., period) 2014), due even to the little ice age (Schurer et al., 2013; McGregor et al., large anthropogenic greenhouse gases emissions. The It has been mentioned above that model results last millennium therefore provides an ideal baseline overall agree with reconstructions but a more detailed to estimate the magnitude of natural and of internal analysis indicates that many of them tend to simulate variability at various spatial and temporal scales. a response to volcanic eruptions that is larger than in the reconstructions, with the largest differences in the beforeFor instance, the dominant this period impact can of begreenhouse used to evaluategas forcing, the 2k-PMIP3, 2015). The correlation of temperature changes andprobability the processes of extreme responsible events suchfor these as droughts events, asor well floods as amongSouthern the Hemisphere continents (Neukomand between et al.,the 2014;hemispheres PAGES to test the stationarity of the teleconnections within the is also higher in the model simulations than in the
2015). This high spatial coherence may be due either to climate system (e.g., PAGES 2k Consortium, 2013; Ortega uncertaintiesreconstructions in (Neukomthe external et al.,forcing 2014; estimates, PAGES 2k-PMIP3, to a too Theet al., influence 2015; Coats of theet al.. natural 2016). forcing strong and homogenous response to external forcings Although natural forcing may be relatively weak at the regional scale, it is possible to detect and attribute LeGrande et al., 2016), or to an underestimation of the magnitudein the climate of internal model simulationsvariability in (Stoffel the models et al., (which 2015; during the last millennium, in particular the imprint of induces changes much less coherent among continents statistically its influence on North Hemisphere temperature
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 48 than the external forcings). Alternatively, uncertainty in modes and trigger a phase shift. The instrumental era the proxy-based reconstructions, due to the non-climatic is clearly too short to draw robust conclusions on such noise, could lead to an underestimation of the coherency of the changes between regions. Determining the origin of those discrepancies will require to investigate all those potential responses, as it mainly includes five large elements simultaneously. A more objective model-data volcanic eruptions (Krakatau in 1883, Santa María in comparison is in particular required, including forward some1902, Agungindications in 1963, exist El Chichónthat volcanic in 1982 eruptions and Pinatubo might in modules that simulate explicitly the proxy variables 1991). On the basis of analyses over the last millennium, measured such as tree ring width or isotope composition large eruption (e.g., Emile-Geay et al., 2008). Similarly, itpromote has been a positiverecently phaseshown of that ENSO the thelargest year eruptions following of a mismatch may simply come from the differences in the the last millennium are almost systematically followed variables(Evans et that al., are 2013). compared Indeed, between a significant proxies and part models. of the
Changes in ocean circulation by a positive phase of the NAO, particularly clear for the The changes in ocean circulation could potentially thatsecond volcanic winter eruptions after the eruptions may act as (Ortega a pacemaker et al., 2015, of their Fig. provide large contributions to the decadal variability 2). For the AMV and AMOC, a few recent studies suggest over the last millennium. In the North Atlantic, some model results and reconstructions have suggested that cleardecadal and variability robust conclusions over the last concerning millennium the (Otteråexact impact et al., decadal to centennial variations in the intensity of both of2011; volcanic Swingedouw eruptions et al.,on 2015).these variabilityNevertheless, modes no entirely can be the subpolar gyres and the meridional overturning circulation induced important changes in the oceanic reproduce such impacts and the intricacy to separate the heat transport, and thus had large-scale impacts on contributiondrawn, notably of the due forcing to difficulties from the of stochastic climate models internal to
Moreno-Chamarro et al., 2016). climate (e.g., Lund et al., 2006; McCarthy et al., 2015; fluctuations (Swingedouw et al., 2017). circulation is a challenge because the available proxy dataConfirming is much those more hypotheses abundant over about the the continents role of ocean than over the ocean and globally only a few marine records have high enough resolution to correctly represent
1850,decadal compatible fluctuations. with Compilations the one reconstructed of observations for have the Figure 2: Composite (black line) and individual (colored bars) continentsconfirmed a(McGregor global oceanic et al., cooling 2015). over At theregional period scales, 850- NAO responses to 8 of the strongest eruptions in the last new high-resolution oceanic observations (e.g. Reynolds millennium, as described by a multi-proxy NAO reconstruction et al., 2016) and syntheses (as in McGregor et al., covering this period (Ortega et al., 2015). Note that compared to Ortega et al. (2015), we have refined the selection of events understanding of past oceanic changes are expected in to only consider those for which the date of the eruption is the2015) coming are under years. way and significant progresses on our well constrained (see Swingedouw et al, 2017). Significance is assessed following a Monte Carlo approach with 1,000 random selections of 8 years from the NAO reconstruction. The influence of the external forcing on the modes of Significant values at the 90% and 95% confidence levels are climatic variability represented by crosses and stars, respectively. The climate system variability is organized in large- scale modes of variability that are mainly governed by the dynamics of the ocean and the atmosphere. Well- Further developments and perspectives known examples of these variability modes are the El- A promising way forward to reduce the uncertainties is the combination of paleo-data and model results to reconstruct as accurately as possible the state of the system Niño Southern Oscillation (ENSO), the North Atlantic over the past millennium. Although many challenges Oscillation (NAO) or the Atlantic Multidecadal Variability remain, such reanalyses for the past millennium are (AMV). While ENSO impacts the variability of the tropical AtlanticPacific, sector, with manyand the teleconnections AMV on sea-surface world-wide, temperature the variabilityNAO is focussed in the onAtlantic, local windpotentially variations related in thewith North the bettercurrently understanding under development of the dynamics (Goosse of et the al., system, 2012; Hakim those large-scale Atlantic meridional overturning circulation reconstructionset al., 2016; PAGES, could 2017). provide In addition a test bed, to contributing complementary to a to the last century, for decadal prediction systems (Meehl et al., 2014), in order to evaluate their skill in a wider Volcanic(AMOC). eruptions, by cooling the climate at the global scale, might strongly impact the fates of these variability the initial climatic conditions for such retrospective range of conditions. However, the ability to generate
49 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 predictions (hindcasts) is a dominant issue and the small amount of available data, notably in the ocean realm, 2005: Internal and forced climate variability during strongly limits for the moment the breadth of the tests theGoosse, last H., millennium: H. Renssen, aA. model-dataTimmermann, comparison and R.S. Bradley, using that can be performed over these past periods.
Many of the conclusions and the questions raised above ensemble simulations. Quat. Sci. Rev., 24, 1345-1360. for the last millennium correspond to research priorities for the more recent past (i.e., the instrumental period) TheGoosse, role of H., forcing E Crespin, and internal S. Dubinkina, dynamics M.F. in explaining Loutre, M. the E. as well. The advantage of the investigations covering the “MedievalMann, H. Renssen, Climate Y.Anomaly”. Sallaz-Damaz, Clim. andDyn., D. 39, Shindell, 2847–2866. 2012: last millennium is the possibility to analyze longer time series, and thus increase the signal-to-noise ratio for the detection and attribution, for example, of forced signals. Anderson, R. Tardif, N. Steiger, and W.A. Perkins, 2016: The disadvantage compared to the instrumental period is TheHakim, Last G.J., Millennium J. Emile-Geay, Climate E.J. Reanalysis Steig, D. Project: Noone, frame D.M. the scarcity and larger uncertainties of the proxy records and the larger resources needed to perform model 6745–6764. simulations spanning several centuries. Despite those work and first results. J. Geophys.l Res.: Atmos., 121, challenges, the information brought by the instrumental and pre-instrumental periods is very complementary, uncertainty in regional climate predictions. Bull. Am. justifying strong interactions and collaborations. Met.Hawkins, Soc., E.,90, and 1095-1107. R. Sutton, 2009: The potential to narrow
Acknowledgements variability over the last millennium. Clim. Past, 6, 723– 737. Jungclaus, J. H. et al., 2010 : Climate and carbon-cycle Belgium).Hugues Goosse Didier is ResearchSwingedouw Director is supported within the by Fonds the LeGrande, A., K. Tsigaridi,s and S.E. Bauer , 2016: Role National de la Recherche Scientifique (F.R.S.-FNRS- of atmospheric chemistry in the climate impacts of stratospheric volcanic injections. Nature Geoscience, 9, French Centre National de la Recherche Scientifique 652–655. note(CNRS). is a The contribution work of Pablo to the Ortega PAGES2k is funded working by the group NERC of PAGES.research Support Project for DYNAMOC PAGES activities (NE/ M005127/1).is provided by This the Lund, D.C., J. Lynch-Stieglitz, and W.B. Curry, 2006: Gulf Stream density structure and transport during the past millennium. Nature, 444, 601–604. US and Swiss National Science Foundations, US National Oceanographic and Atmospheric Administration and by Referencesthe Future Earth program. Boer, G.J., 2011: Decadal potential predictability of signaturesMann, M., Z. and Zhang, dynamical S. Rutherford, origins R. of Bradley, the Little M. Ice Hughes, Age and D. MedievalShindell, C.,Climate Ammann, Anomaly. G., Faluvegi Science, and 326, F. Ni,1256–1260. 2009: Global 1119-1133. twenty-first century climate. Climate Dynamics, 36, Coats, S., J.E. Smerdon, Cook B.I., R. Seager, E.R.Cook, and K.J. Anchukaitis, 2016: Internal ocean-atmosphere climateMcCarthy, variability G.D. I. D. Haigh,revealed J.l J.-M.by sea-level Hirschi, J.observations. P. Grist, and variability drives megadroughts in Western North Nature,D. A. Smeed, 521, 508-510. 2015: Ocean impact on decadal Atlantic America. Geophys. Res. Lett., 43, 9886-9894.
Deser C., A. Phillips, M.A. Alexander, and B. Smoliak, 2014: trend for the past two millennia. Nature Geoscience, 8, Projecting North American Climate over the next 50 671-677.McGregor, H.V., et al., 2015: Robust global ocean cooling Years: uncertainty due to internal variability. J. Climate, 27, 2271-2296. Masson-Delmotte, V. et al. , 2013. Information from Paleoclimate Archives, in: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I
Clim.,Emile-Geay, 21, 3134-3148. J., R. Seager, M. A. Cane, E. R. Cook, G.H. Haug, 2008: Volcanoes and ENSO over the past millennium. J. Cambridgeto the Fifth UniversityAssessment Press, Report Cambridge, of the Intergovernmental United Kingdom Evans, M.N., S. E.Tolwinski-Ward, D.M. Thompson, and andPanel New on York, Climate NY, Change,USA, 2013. edited by: T.F. Stocker et al., K.J. Anchukaitis, 2013: Applications of proxy system Meehl, G.A., et al., 2009: Decadal Prediction Can It Be Rev., 76, 16–28. modeling in high resolution paleoclimatology. Quat. Sci., Skillful? Bull. Amer. Met. Soc., 90, 1467–1485.
Moreno‑Chamarro E., D. Zanchettin, K. Lohmann, and J.H.
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 50 Jungclaus, 2016: An abrupt weakening of the subpolar of explosive volcanic eruptions on the main climate gyre as trigger of Little Ice Age‑type episodes. Clim. Dyn. variability modes. Glob. Planet. Changes, 150, 24-45 .
Neukom,DOI 10.1007/s00382-016-3106-7 R., et al., 2014: Inter-hemispheric temperature variability over the last millennium. Nature Clim. Change, 4, 362–367.
C.C. Raible, M. Casado, and P. Yiou, 2015: A model-tested Ortega, P., F. Lehner, D. Swingedouw, V. Masson-Delmotte, millennium. Nature, 523, 71-74. North Atlantic Oscillation reconstruction for the past
External forcing as a metronome for Atlantic multidecadal variability.Otterå, O.H., Nature M. Bentsen, Geoscience, H. Drange 3, 688-694. and L. Suo, 2010:
Landrum, S. Stevenson, N. Rosenbloom, A. Mai, and G. Strand,Otto-Bliesner, 2016: Climate B.L., E.C. Variability Brady, J.and Fasullo, Change A. since Jahn, 850 L. C.E.: An Ensemble Approach with the Community Earth System Model (CESM). Bull. Amer. Met. Soc., 97, 735-754.
PAGES, 2017: DAPS [online]. Available from http://www. pastglobalchanges.org/165-initiatives/working-group/ daps. Accessed on 10 May 2017.
PAGES 2k Consortium, 2013: Continental-scale temperature variability during the last two millennia. Nature Geosciences 6, 339–346.
PAGES 2k–PMIP3 group, 2015: Continental-scale temperature variability in PMIP3 simulations and PAGES 2k regional temperature reconstructions over the past millennium. Clim. Past, 11, 1673–1699.
Reynolds, D.J., et al., 2016: Annually resolved North Atlantic marine climate over the last millennium. Nature Communication, in press.
J. Phipps, 2013: Separating forced from chaotic climate variabilitySchurer, A. overP., G.C. the Hegerl, Past Millennium. M. E. Mann, J. ClimateS. F. B. Tett 26, 6954–and S. 6973, 2013.
Stoffel, M., et al., 2015: Estimates of volcanic-induced years. Nat. Geosci., 8, 784–788. cooling in the Northern Hemisphere over the past 1,500
Delmotte, P. G. Butler and M. Khodri, 2015: Bidecadal NorthSwingedouw, Atlantic D., ocean P. Ortega,J. circulation Mignot,E. variability Guilyardi,V. controlled Masson- by timing of volcanic eruptions. Nature Communications, 6, 6545
Swingedouw, D., J. Mignot, P. Ortega, M. Khodri, M. Menegoz, C. Cassou, and V. Hanquiez, 2017: Impact
51 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Abrupt Northward Shift of SPCZ position in the late-1920s Indicates Coordinated Atlantic and Pacific ITCZ Change doi: 10.22498/pages.25.1.52 Braddock K. Linsley1, Robert B. Dunbar2, Donna Lee1, Neil Tangri2, Emilie Dassié1,3 1 Lamont-Doherty Earth Observatory, 61 Route 9W, Palisades, NY 10964, USA 2 Department of Environmental Earth Systems Science, Stanford University, Stanford, CA, 94305, USA 3 Laboratoire d'Océanographie et du Climat: LOCEAN - IPSL, UMR 7159 CNRS/UPMC/IRD, Université P. et M. Curie, 4 place Jussieu, 75252 Paris cedex 05, France Introduction Patterns of climate variability are often studied by evaluating instrumental or paleo-data from regions that have the highest correlations to the target climate Pacific (the Pacific Decadal Oscillation (PDO)). Since the PDO appears to have both subtropical and tropical primarily evaluated by compiling or reconstructing origins (Newman et al., 2016), the congruence of ENSO mode. For example, past variations in ENSO have been and PDO nodal lines in some regions is not unexpected. trendsThis region northwest in the Southto southeast Pacific isfrom also the the Equatorcentral rainfall in the sea surface temperature (SST) in the Nino3.4 (5°N-5°S; axis of the South Pacific Convergence Zone (SPCZ) which 120°W-170°W) and Nino3 (5°N-5°S; 90°W-150°W) informationregions in the on the equatorial spatial extent Pacific of (e.g.: a target Trenberth climate 1997;mode western Pacific through Samoa and American Samoa. canUrban also et beal., gained 2000; byCobb studying et al., 2013).conditions However, on the critical nodal line perimeter of the climate pattern where on average inThe the SPCZ tropics is yet the its largestdynamics spur and even of the current Intertropical position there is no correlation (R=~0) between the climate Convergence Zone (ITCZ) and a key hydrologic feature parameter being reconstructed and the target mode. Atmosphericare poorly represented data indicate in climate a close models relationship (Vincent between 1994; Vincent et al., 2009; Cai et al., 2012; Evans et al., 2015). largest SST anomalies are focused in an elongated E-W last 30 years, instrumental precipitation data indicate patternDuring Elor Niñofootprint and that La Niñais generally events symmetric in the Pacific, around the SPCZ movements and ENSO in this region. Over the events display differences in both the amplitude and the that during most El Niño events the SPCZ moves a few longitude0° latitude of the in largest the central SST anomalies Pacific. and Individual cluster analysis ENSO occurdegrees during northward La Niña (Gouriou events (Gouriou and Delcroix and Delcroix2002; Vincent 2002, et al., 2009; Salinger et al., 1995). Southward SPCZ shifts are three primary El Niño patterns and one primary La El Niño events such as 1982/83 and 1997/98, and during of the last 50 years of Pacific SST data indicate that there someVincent moderate et al., 2009; strength Cai et Elal., Niño’s 2012). such During as 1991-1992,very strong relatedNiña pattern SST anomalies (Chen et al., surrounded 2015). The by mean a perimeter of these where ENSO SSTevent is patternson average results not positively in the classic or negatively footprint correlated in ENSO- the SPCZ can collapse onto the equator (so-called zonal slightlySPCZ events; cooler Vincent conditions et al., on 2009; average Linsley in the et area al., 2017).of the line where on average SST variability is not correlated to Both SPCZ responses during El Niño result in saltier and with ENSO (see Fig. 1). This perimeter reflects a nodal nodal line for all types of El Niño and La Niño events has advectsSPCZ central relatively rainfall salty axis water as the into SPCZ the shifts region. northeast and beenENSO. relatively Instrumental stable data over suggest the last that ~50 the years. average ENSO the westward flowing South Equatorial Current (SEC) to southeast though Samoa (14°S, 172°W) and American In an effort18 to trackPorites past changeslutea coral in thecore SPCZ from response the island to In the South Pacific, the ENSO nodal line runs northwest ENSOof Ta’u events in the and Manua the PDOIsland we group have analyzedon the eastern sub-seasonal side of skeletalAmerican δ Samoa.O in a Ta’u Island is located in the center of nodalSamoa line (14°S, for the 169.5°W) decadal mode to French of SST Polynesiavariability (17°S,in the 150°W)(Fig. 1). This location is nearly identical to the the SPCZ and on the nodal line region for both ENSO and Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 52 in American Samoa are closely related to SEC dynamics 3 4 at ~90°C. With the PDO. Variability of surface oceanographic conditions18 samples). With the Isoprime we dissolved ~80-120 μg coral powder aliquots in ~100% H PO reconstruction from Samoa and American Samoa. 3 4 at ~70°C. NBS-19 standards were and SPCZ movements. This is the first 50+ year coral δ O theanalyzed Delta 5V-Kiel to 6 times IV we per dissolved day. To ~50-80 assess external μg coral precision powders Methods inand ~100% sample Hhomogeneity,PO 209 replicate samples were In November 2011 we cored a large colony of Porites analyzed (8.2% replication). The standard deviation of lutea on the western side of the island of Ta’u located at 18 18 169 30.027) on an exposed outer reef in 7.5m (25 feet) of 0.082‰.NBS-19 standards All results analyzedare reported was relative 0.06‰ to VPDB for δ (inO. ‰). The waterS 14° 15'(water 33.74": depth W to169° top 30' of coral).01.61" (orCore S sections14 15.566, from W average difference of the replicate δ O analyses was the Ta’u-1 core were sawed longitudinally in half and The samples from the core’s live-collected top, serve to 5mm thick slabs cut at Stanford University and shipped 18 18 maximumanchor the werechronology attributed for the to δ seasonalO series tomaxima November and deionizedto the Lamont-Doherty water bath with Earth a probe Observatory sonicator (LDEO)(500W, 2011. Below this section, annual δ O minima and for isotopic analysis. At LDEO, slabs were cleaned in a comes from pseudo-coral forward modeling where we usedminima instrumental in SST, respectively. SST and Verificationsea surface ofsalinity this approach (SSS) to 3520kHz). Kv). The Slabs X-ray were positives then oven were dried used at to50°C. identify Once growth dried, 18 bandthe slabs orientation were X-rayed and thein an maximum HP cabinet growth X-ray axissystem down (at 18 each slab section as a guide to the subsampling path. generateat Ta’u is duea modeled to SST andcoral SSS. δ OTa’u-1 series annual from 2008average to 1981coral Subannual samples were hand-drilled from the slabs (Fig.18 2). This model assumes all variability18 in coral δ O at 1 mm intervals by excavating a 3 mm wide by 2 mm δ O and annual average pseudo-coral18 δ O significantly spherical carbide bit. correlate (R=0.77; p < 0.001) with annual average deep trough with a variable speed Dremel drill fit with a correlations between coral δ O and SST and SSS of -0.58
Figure 1: Average correlation pattern of SST to Nino3 SST. Red-purple colors indicate positive correlation and green-blue colors indicate negative correlation. White box indicates the Nino3 region for monitoring ENSO sign and strength and black dashed box the Nino3.4 region. Yellow circle highlights the location of our Ta’u coral site in American Samoa. Note that the zero correlation line (nodal line) runs SE-NW right through Samoa. Also shown are the locations of Fanning, Palmyra and Maiana (see Fig. 3C) and Fiji, Tonga, Tokelau and Tavalu (see Fig. 4).
Sample powders were analyzed on either an Elementar and 0.64 respectively. This pseudo-coral comparison Isoprime mass spectrometer equipped with a dual- 18 spectrometer with dual-inlet and Kiel IV carbonate interannualverifies our changeschronology in SST and and indicates SSS have that additive Porites effects coral reactioninlet and device. Multiprep The or instruments a Themo-Fisher are Deltain the V+ same mass δ O at Ta’u is a function18 of both SST and SSS and where core extends from November 2011 to ~ January 1800. on interannual coral δ O variability. This 2.6m section of 2 reference gas and dewatered phosphoriclaboratory atacid LDEO is made and using have the been same cross-calibrated. protocols for Results and Discussion They use the same CO 18 Although on average there is no correlation between SST the analysis of this upper 2.6m of core (n=2,565 1mm variability in American Samoa and Nino3.4 or Nino3 SST each instrument. Here we report the δ O results of
53 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Figure 2: (top): Sea surface temperature (SST) (NCEP) and sea surface salinity (SSS; Delcroix et al., 2011) for the grids containing Ta’u, American Samoa. (bottom): Ta’u coral 18O from the Ta’u-1 Porites coral core and “pseudo-coral 18O” calculated from SST and SSS. The very strong (VS) El Niño events of 1982/93 and 1997/98 are indicated by gray shading. During these events the SPCZ collapsed onto the equator (so-called zonal SPCZδ events). Annual average 18O and pseudo-coral 18O,δ R = 0.77 (p<0.001). This correspondence demonstrates that coral 18O at Samoa is accurately recording surface ocean conditions. δ δ δ 18 result in elevated SSS in the Samoa region today (Gouriou over the last ~30 years (see Fig. 1), large El Niño events shiftComparing in the thedecadal Ta’u mean coral correlation δ O record in to the the late timing 1920s. of equatorial ENSO variability indicates a 18 striking phase and Delcroix 2002; Hasson et al., 2013)(see Fig. 2). There 18 has been a close relationship between SPCZ movements, Runninghighlights correlations the abruptness between of the the phase Ta’u shift δ O in results the 1920s and ElENSO Niño and events the easternand southwest extent ofduring the westernLa Niña Pacificevents Nino3.4 SST and the FPM equatorial composite coral δ O (Gouriouwarm pool and with Delcroix the SPCZ2002, shiftingVincent et northeast al., 2009). during This ERSST) at American Samoa shows no change in phasing with(see Nino3.4 Fig. 3B SST and in D). the Sea 1920s surface (not temperatureshown) pointing (using to a change in the timing of interannual surface salinity SPCZ re-positioning results in more saline conditions on 18 relativelyaverage in salty the Samoawater into region the duringregion. El Niño as the SPCZ relationship between El Niño events and more saline shifts northeast and the westward flowing SEC advects variability. The Ta’u coral δ O series indicates that the To evaluate interannual and lower frequency changes in back to 1927, when there was an abrupt change. Prior to 18 conditions in this central18 region of the SPCZ existed only 18 that on average fresher conditions occurred during El Ta’u-1 coral δ O for comparison to equatorial indices of 1927, the Ta’u coral δ O record contains distinct evidence ENSO, we filtered the monthly coral18 δ O series in two This is exactly the situation which occurs today north ways. Our first approach was to 24 month18 high-pass ofNiño Samoa events between in Samoa/American 7°S and 8°S near Samoa the island(Fig. 3A groups and C). of filterdetrending and then was detrend accomplished the coral using δ OSingular series due Spectrum to the presence of a significant secular coral δ O trend. The 18 eventsTokelau in (8°S, the region 172°W) extending and Tavalu NW-SE (7°S, including 179°E) the (see Tavalu Fig. timeAnalysis series to isolatewas then and compared then remove to 7 the month first principalrunning and4). Surface Tokelau salinity Island isgroups significantly at 7-8°S. lower At theduring same El time,Niño component (the secular trend). This filtered Ta’u δ O average filtered Nino3.4 SST anomalies (see Fig. 3A). The Thesesurface observationssalinity increases indicate at Fiji, thatTonga the and mean Samoa position (Fig. 4). second filtering approach was to apply only18 a 24 month high-pass filter (leaving the trend in place) to facilitate18 current position during at least the ~50 year period direct comparison to equatorial coral δ O records (see priorof the to SPCZthe late must 1920s have combined been shifted possibly southwest with a reduced of its etFig. al., 3C). 2013) We useand a Maiana composite (Urban average et al., of 2000) three coralas a coral δ O latitudinal migration to the northeast during El Niño. records18 from Fanning (Cobb et al., 2013), Palmyra (Cobb This reorganization would explain the fresher conditions see Linsley et al., 2015). 18 δ O-based index of equatorial ENSO state (termed FPM; during El Niño recorded in our Ta’u coral δ O record in
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 54 Year 1800 1850 1900 1950 2000 -0.4 3 18 18 Tau-1 coral δ O, 24 month high pass filtered - trend Samoa cora δ O and Nino3.4 SST -0.3 Nino3.4 SST 7 month running average A Nino3.4 SST (°C) 2 -0.2
-0.1 1 0 O 24mo-trend 18
δ 0.1 0
0.2
Tau-1 -1 0.3
0.41 -2 R value (Tau-1 d18O vs Nino3.4 SST)(25 month window) 0.5 B 0
R value -0.5
(25 month window) -1 1800 1850 1900 1950 2000 Year
1800 1850 1900 R= -0.61 1950 R= 0.19 2000 -1 -0.8 18 18 18 Fanning, Palmyra, Maiana (FPM) coral δ O Samoa coral δ O and Equatorial Pacific coral δ O FPM coral δ -0.8 composite (7 month running average). -0.6 18 C Tau-1 coral δ O (24 month high pass filtered) -0.6 -0.4
O (per mil) -0.4 -0.2 18 18 δ O (per ml) -0.2 0
0 0.2
0.2 0.4 Tau-1 coral
0.4-1 0.6 Tau-1 d18O and FPM coral d18O -0.5 (running correlation)
0 D R value 0.5
(25 month window) 1 1800 1850 1900 1950 2000 Year R= 0.54 R= -0.13 Figure 3: Comparison of Ta’u (American Samoa) coral 18O to: (A) Nino3.4 SST and (C) an equatorial Pacific composite coral 18O record from Fanning, Palmyra and Maiana (FPM)(Linsley et al., 2015). Panels B and D show the 25 month running correlation between the series. Horizontal gray bars in B and D indicateδ average correlation (R value) across the interval. In panel A, arrows δindicate El Niño events where it was distinctly cooler and saltier at Samoa. Note that on average after ~ 1927, warmer conditions in the Niño3.4 area (El Niño) occurred when it was cooler and saltier at Samoa. Before 1927, the opposite pattern is observed; fresher and warmer conditions at Samoa corresponded with El Niño conditions on the equator back to ~ 1872 AD (see panel B) this period prior to 1927. This is the opposite response to the higher salinity conditions that occur during El Niño events at Ta’u beginning at ~ 1930. The abruptness of the in the mid-1920a as the AMO changed from a negative to shift in El Niño response in the late 1920s suggests a rapid positive phase and the ITCZ in the Atlantic shifted north axis(e.g.; also Knight shifted et north al., 2006; in the Zhang mid-1920s and Delworth,as our Ta’u 2006; coral Based on observational data in the Atlantic, the timing García-García18 and Ummenhofer 2015). If the SPCZ central reorganization of climate patterns in the South Pacific. δ O indicates, this would point to a coordinated ITCZ of this abrupt change in SPCZ position occurred during change in both the Atlantic and Pacific basins. However, a phase change of the Atlantic Multidecadal Oscillation the AMO also changed phase in the late 1960s when our (AMO) when SST in the North Atlantic abruptly warmed Ta’u results do not indicate a phase change between SPCZ
Figure 4: Monthly surface salinity from near Fiji, American Samoa, Tonga, and near Tokelau and Tavalu (north and northwest of Samoa respectively) (data from Delcroix et al., 2011). Note the strong freshening in the regions of the Tokelau and Tavalu Islands during El Niño (tan bars) when the SPCZ shifts northeast. Fiji and Tonga experience higher salinities during El Niño whereas Samoa surface salinity has a more intermediate response. Blue bars are La Niña events.
55 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 impacts of the Atlantic multidecadal oscillation, Geophts. shiftedvariability from and a positive equatorial phase ENSO. to a negative The lack phase of asuggests change Res.Knight, Lett., J. vol R. 33, C. K.L17706, Folland, doi:10.1029/2006GL026242 J. A. Scaife, 2006: Climate thatin SPCZ-ENSO there was somethingphasing in climaticallythe late 1960s different when in the the AMO late changed1920s and/or phase, that although the SPCZ this and phase ITCZ change are not was connected gradual Linsley, B. K., H. C. Wu, E. P. Dassié, and D. P. Schrag, causally. The 1920’s were also a time when the PDO oscillation2015: Decadal and upper changes ocean in heat South content, Pacific Geophys. sea surface Res. Lett.,temperatures 42, doi:10.1002/2015GL063045. and the relationship to the Pacific decadal and appears to have started in the early 1920s. Other clues to significant tropical-subtropical re-organization in the late 1920s at the same time as the SPCZ and Atlantic ITCZ shifted north are a 1920s shift to weaker Pacific trade Linsley, B. K., H. C. Wu, T. Rixen, C. D. Charles, A. L. inwinds the late(Thompson 1920s will et al.,require 2014). future Further work. interpretation of Gordon, M. D. Moore, 2017: SPCZ Zonal Events and these preliminary observations of SPCZ position change 43,Downstream doi:10.1002/2016GL070985. Influence on Surface Ocean Conditions in References the Indonesian Throughflow Region, Geophys. Res. Lett., Cai W., M. Lengaigne, S. Borlace, M. Collins, T. Cowan, M. J. Newman, M., M.A. Alexander, T.R. Ault, K.M. Cobb, C. McPhaden, A. Timmermann, S. Power, J. Brown, C. Menkes, Deser, E. Di Lorenzo, N.J. Mantua, A.J. Miller, S. Minobe, A. Ngari, E. M. Vincent, M. J. Widlansky, 2012: More
Nature, 488, 365–369, doi:10.1038/nature11358. revisited.H. Nakamura, J. Climate, N. Schneider, vol. 29, 4399-4427,D.J. Vimont, JuneA.S. Phillips,2016. J.D. extreme swings of the SPCZ due to greenhouse warming, Scott, C.A. Smith, 2016: The Pacific decadal oscillation,
Schmidely-Leleu, 1995: Climate trends in the south-west windChen, burstsD. , T. Lian, on El C. NiñoFu, M. diversity, A. Cane, Y.Nature Tang, R.Geoscience, Murtugudde, vol Salinger, M. J., B. B. Fitzharris, J. E. Hay, P. D. Jones, and J. P. X. Song, Q. Wu, L. Zhou, 2015: Strong influence of westerly Thompson,Pacific, Int. J.D. Climatol., M., J. E. Cole, 15, 285–302. G. T. Shen, A. W. Tudhope, G. 8, May 2015, doi:10.1038/NGEO2399. M. Meehl, 2014: Early twentieth-century warming linked
Cobb, K. M., N. Westpha, H.R. Sayani, J.T. Watson, E. Di published online December 22, 2014, doi: 10.1038/ Lorenzo, H. Cheng, R.L. Edwards, and C.D. Charles, 2013: Ngeo2321.to tropical Pacific wind strength, Nature Geoscience, doi:10.1126/science.1228246.Highly Variable El Niño-Southern Oscillation Throughout the Holocene, Science, 339, no. 6115, pp. 67-70,. Delcroix, T., G. Alory, S. Cravatte, T. Correge, and M. J. of the American Meteorological Society, 78, 2771-2777. McPhaden, 2011: A gridded sea surface salinity data set Trenberth, K. E. (1997) The Definition of El Niño. Bulletin
2008), Deep Sea Res., Part I, 58(1), 38–48, doi:10.1016/j. mean climate change on climate variability from a 155- dsr.2010.11.002.for the tropical Pacific with sample applications (1950– Urban, F. E., J.E. Cole, J. T. Overpeck, 2000: Influence of
Evans. J.P. K. Bormann, J. Katzfey, S. Dean, R. Arritt, 2015: year tropical Pacific coral record, Nature, 407, 989-993. review, 1994: Mon. Weather Rev., 122, 1949–1970. Vincent, D., The South Pacific convergence zone (SPCZ): A s00382-015-2873-x.Regional climate model projections of the South Pacific Vincent, E. M., M. Lengaigne, C. E. Menkes, N. C. Jourdain, P. Convergence Zone, Clim. Dynamics, DOI 10.1007/ Marchesiello, and G. Madec 2009: Interannual variability
Multidecadal variability of the continental precipitation for tropical cyclone genesis, Clim. Dyn., 36(9-10), 1881– García-García, D., and C. C. Ummenhofer (2015), 1896,of the doi:10.1007/s00382-009-0716-3.South Pacific Convergence Zone and implications Res. Lett., 42, 526–535, doi:10.1002/2014GL062451. annual amplitude driven by AMO and ENSO, Geophys. multidecadal oscillation on India/Sahel rainfall and of sea surface salinity and temperature in the South AtlanticZhang, R.,hurricanes, T. L. Delworth, Geophys. 2006:Res. Lett., Impact vol. 33, of L17712, Atlantic Gouriou Y, T. Delcroix, 2002: Seasonal and ENSO variations doi:10.1029/2006GL026267.
Pacific Convergence Zone during 1976–2000. J Geophys Res Oceans 107(C12):8011. doi:10.1029/2001jc000830. Hasson, E.A., T. Delcroix, R. Dussin, 2013: An assessment of the mixed layer salinity budget in the tropical Pacific Ocean. Observations and modelling (1990–2009), Ocean Dynamics DOI 10.1007/s10236-013-0596-2.
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 56 Summer North Atlantic Oscillation (SNAO) variability on decadal to palaeoclimate time scales doi: 10.22498/pages.25.1.57 Hans W. Linderholm1 and Chris K. Folland1, 2, 3, 4 1 University of Gothenburg, Sweden 2 Met Office Hadley Centre, UK 3 University of East Anglia, UK 4 University of Southern Queensland, Australia The summer North Atlantic Oscillation (SNAO) climate in the North Atlantic region has been highlighted East Asia (Linderholm et al., 2011). overThe influencethe past offew the decades. North Atlantic Although Oscillation most prominent (NAO) on noted, e.g. eastern North America (Hardt et al., 2010) and variability that persist throughout the year, although variability can be linked to variations in North Atlantic during winter, the NAO is one of the few modes of On interannual to multidecadal timescales, SNAO through the seasons (Barnston and Livezey, 1987). This indicate an association between the Atlantic Multidecadal isthere related are to systematic seasonal variations differences of the in North its configuration Atlantic jet surface temperature (SST). Observations and models stream which on average moves northwards in summer relative to winter. Consequently, the positive and negative Oscillation (AMO) ( Kerr, 2000) and the SNAO for periods greater than 10 years (F09) such that a cold (warm) phase positions during summer. Until recently, most studies of of the AMO corresponds a positive (negative) phase of nodes of the dipole NAO pattern have more northerly the SNAO, clearly seen in Sutton and Dong (2012). change, particularly related to the large reduction in the link between the NAO and climate have focused on seaRecently, ice coverage, the potential on mid-latitude influence circulation of Arctic patterns climate haswinter, also but been after directed a thorough to summer. study of the summer NAO (SNAO) by Folland et al. (2009, henceforth F09), attention suggestedhas been that studied winter (e.g. sea Overlandice concentration and Wang, conditions 2010; located over the British Isles/Scandinavia and Greenland Francis and Skific, 2015). For instance, Wu et al. (2013) During summer the NAO pattern has its pressure centres atmospheric circulation over northern Eurasia. Using observations,west of Greenland Knudsen influences et al. (2015) the found following a link between summer (Hurrell and Folland, 2002). Due to the lack of data from anomalous Arctic sea ice melt and changes in midlatitude its northern node, SNAO has largely been defined until atmospheric patters during summer, as did Screen levelnow from (PMSL) the variabilityanomalies of (PMSLA) the southern over node.the extratropical F09 defined (2013) using an atmospheric general circulation model. European–Norththe SNAO as the first Atlantic eigenvector sector of(25–70°N, pressure 70°W–50°E) at mean sea Petrie et al. (2015), using a fully coupled climate model, found that sea ice loss together with increased SST in the interannual to decadal variability as does the winter Labrador Sea affects the summer atmospheric circulation in July and August (JA). The SNAO time series shows large over the North Atlantic region.
EuropeanNAO, but thesummer correlation storm tracksbetween (Dong them et is al., very 2013). low. In The its Within an ongoing International CLIVAR Climate of the SNAO phase is strongly related to changes in Atlantic and 20th conditions over Northern Europe, yielding sunny, warm and a project supported by the Swedish Research council, andpositive dry conditionsphase, the there.SNAO Accordingly,is associated the with positive anticyclonic phase Century (C20C) Project (Kinter and Folland, 2011) on decadal to multicentury timescales, mechanisms to Scandinavia in particular, and a northerly position of behindstudies haveits variability been underway and itsto describepotential SNAO predictability. variability of the SNAO is related to summer droughts from the UK storm track moves ~10° further south, giving cloudy, wet to include June in addition to July and August, and new andthe maincooler storm conditions track. over In the this negative region. SNAOThe relationship phase, the Recent work has extended the definition of the SNAO with surface climate is surprisingly strong for southern to include data from the whole Arctic. This work, to be Europe and more or less the opposite, especially in the reporteddata sets elsewhere, allow this definitiondoes not change to be extended the fundamental spatially eastern Mediterranean (Chronis et al., 2011). Climatic aligned to important aspects of seasonal forecasting spatial or temporal character of the SNAO but it is better influences outside northwestern Europe have also been 57 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 wards more anticyclonic, dry, conditions arising from an-
including the use of CMIP5 models, to attempt to explain thisthropogenic large short warming term change(e.g. F09). in climate, Research particularly is underway, sea surface temperatures in the Atlantic and the global trop- -
ics and possible influences of changing Arctic sea ice ex scalestents (Screen, are summarised 2013; Petrie annually et al., 2015).in State Other of the aspects Climate of the behaviour of the SNAO on interannual to century time
SNAOpublications variability (e.g. Allanduring and the Folland, last millennium 2016). Several studies have shown that tree growth variations - culation patterns (e.g. Seftigen et al., 2013), suggesting theacross suitability Europe ofare using linked tree-ring to SNAO-like data to atmospheric reconstruct cirthe
tree-ring data from western Norway and northern UK to SNAO before the observational record. Indeed, F09 used
aproduce tree-ring a reconstruction network with much of the larger JA SNAO spatial back distribution, to 1706 CE, theverified reconstruction by long instrumental was extended records back tofrom 1441 the CE, UK. provid Using- ing opportunities to study e.g. associations between the Figure 1: (Top Panel) Differences in pressure at mean sea level over the North Atlantic and Europe in July and August between the two decades 1997-2016 and the two decades climateSNAO and in East European/Sahel Asia in a long-term drought context (Linderholm (Linderholm et al., 1966-1985, together with significances of these differences 2009) and associations between the SNAO and summer- at the 5% level. (Bottom Panel) Variations in the July and struction where the target season was extended to JJA. August SNAO, 1850-2016 Thiset al., should 2013). also Here help we frompresent a tree-ring a preliminary perspective new reconas the
temperature or precipitation in June as well. This recon- research. Here we confine ourselves to the JA SNAO as struction,growth of treesbased in only northwestern on tree-ring Europe data fromis influenced the south by- adiscussed key component in F09 and of showJA atmospheric a key result circulation that indicates change. that over the last 5 decades the JA SNAO emerges naturally as mean sea level using the NCEP Reanalysis between the ern node region of the SNAO (i.e. UK and Fennoscandia), twoFig. most 1 (top recent panel) decades shows 1997-2016 the difference and the in two pressure decades at above,which it extends is also compared back to 1200 to northeastern CE, is shown Canadian in Fig. sum 2. In- 1966-1985. These periods have been chosen to illustrate merlight seaof the ice potentialcover (SIC) influences variations of inferredArctic sea from ice corallineas noted - - tom panel). This shows that the last half century contains the character of a large decline in the JA SNAO (Fig 1, bot algae (Halfar et al., 2013), and a multi-proxy reconstruc thetion Little of the Ice AMO Age (LIA) (Mann coincided et al., 2009). with high On SIC multidecadal (note that the largest coherent fluctuation of the SNAO since 1850 timescales, a sustained period of negative SNAO during similarwith a large to its decline average in level its value in the since late thenineteenth 1970s. However century, sothe that recent the relativelyvery positive negative levels level of the of the1970s SNAO in particuis quite- SIC is inverted in Fig. 2). Also the positive SNAO in the andtwentieth summer century SIC during corresponds the last to six a centuriessignificant is decrease evident. the difference pattern appears to be very like the nega- Thisin SIC. may However, be because no stable none association actually exists, between despite the an SNAO ap- lar are the more unusual. Fig 1 (top panel) shows that eigenvector analysis over the much longer period 1881- 2003tive pattern and explaining of the SNAO. about Thus 28% the of SNAO, the mean defined July fromand Au an- theparent variation influence, of drivers where not reduced studied Arctic here seais more ice extentsimpor- gust variance over this period, dominates the changing favour the negative SNAO implied by Screen (2013), or pattern of interdecadal July and August decadal pressure at mean sea level since the 1970s over the North Atlantic tant. To better assess the potential influence of Arctic sea and Europe. In fact the centres of difference over Green- isice in on general the SNAO (except in afor long-term the 1200s) context, quite similar additional to that SIC proxies are needed. The long-term evolution of the AMO at the 0.1% level. The negative centre over the UK has, forland instance, and near led the to Unitedmarkedly Kingdom wetter are late both summers significant over of the SNAO: negative (positive) multidecadal phases of England and Wales in the most recent decade in contrast the the AMO correspond to periods of negative (positive) to expected long term changes found in many papers to- SNAO. Our tentative comparison suggests that both long- term changes in the AMO and SIC are of opposite signs in their apparent influences on the recent shorter-term
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 58 on the Eastern Mediterranean. J. Clim, 24: 5584–5596. 2011: The Summer North Atlantic Oscillation Influence
Variability of the North Atlantic summer storm track: mechanismsDong, B., R.T. andSutton, impacts T. Woollings, on European and K.climate. Hodges, Environ. 2013: Res. Lett. 8: 034037.
-
Folland, C.K., J. Knight, H.W. Linderholm, D. Fereday, S. In 1103.eson, and J.W. Hurrell, 2009: The Summer North Atlantic Oscillation: past, present and future. J. Clim, 22: 1082–
Arctic warming to mid-latitude weather patterns. Phil. Figure 2: A preliminary reconstruction of the JJA SNAO, Francis, J., and N. Skific, 2015: Evidence linking rapid based on tree-ring data from Fennoscandia and the UK, (upper panel) compared to inferred (inverted) sea ice variability in Trans. R. Soc. A 373; 20140170 the eastern Canadian Arctic (middle panel), derived from coralline alge and representing the region 85-60°W, 55-73°N, byHalfar, multicentury J., W.H. Adey, annual-resolution A. Kronz, S. Hetzinger,record from E. crustose Edinger, (Halfar et al., 2013, data available at www.ncdc.noaa.gov/ corallineand W.W. algalFitzhugh, proxy. 2013: PNAS, Arctic 110, 19737–19741.sea ice decline archived paleo-search/study/15454), and a reconstruction of the AMO (lower panel, Mann et al. (2009)). All records have been - z-scored. Thick lines represent 30-year variability. Note that positive algal proxy anomalies correspond to below normal wards, 2010: The seasonality of east central North Amer- sea ice coverage. Hardt, B., H.D. Rowe, G.S. Springer, H. Cheng, and R.L. Ed- leothems from southern West Virginia. Earth Plan. Sci. - Lett.,ican precipitation295: 342-348 based on three coeval Holocene spe behaviour of the SNAO. Still, this apparent contradic- - structiontion may beof duetemperatures to the data mainly used here. based For on instance, terrestrial the mer circulation over the North Atlantic. CLIVAR Exch., 25: proxiesAMO index only. used The here recent was increase derived infrom the a spatiotemporalgridded recon 52–54.Hurrell JW. and Folland CK. 2002: A change in the sum representation of palaeoclimate proxies, e.g. within the PAGES 2k initiative, provides new the opportunities for Kerr, R. A., 2000: A North Atlantic climate pacemaker for improving the multicentury reconstruction of indices the centuries. Science, 288, 1984–1985.
Acknowledgementslike the AMO and different Arctic sea ice parameters. The paper contributes to the CLIVAR International Cli- Kinter, J. and C.K Folland, 2011: The International CLIVAR Climate of the 20th Century Project: Report of the Fifth the PAGES 2k Network. Past Global Changes (PAGES) is Workshop. CLIVAR Exchanges, 57, 39-42. DOI: mate of the Twentieth Century Plus (C20C+) project and- - mersKnudsen, of unusual EM. Orsolini, Arctic sea YJ. ice Furevik, melt. J Geophys T. and Hodges, Res-Atmos KI. supported by the US and Swiss National Science Founda 2015: Observed anomalous atmospheric patterns in sum tions. HL and CF were funded by the Swedish Research (GA01101)Council (VR, and grant the 2012- Climate 5246) Science and toCF Service by the JointPartner UK- 120: 2595-2611. DOI: shipBEIS/Defra (CSSP-China). Met Office We Hadley thank YochananCentre Climate Kushnir Programme for com- multicentury perspective on the summer North Atlantic ments on an earlier version of the manuscript. Linderholm, H.W., C.K., Folland, and A., Walther, 2009: A
References Oscillation (SNAO) and drought in the eastern Atlantic Region, J. Quat. Sci., 24: 415–425. 1. Sea level Pressure. State of the Climate 2015. Bull. - Amer.Allan, R.,Meteor. and C.K. Soc. Folland, 97, S36-38. 2016: Atmospheric circulation. Linderholm HW. T. Ou, J-H. Jeong, C.K. Folland, D. Gong, theH. Liu, summer Y. Liu, andEast D. Asian Chen, Monsoon. 2011: Interannual J.Geophys. Teleconnec Res, 116: - D13107tions between the Summer North Atlantic Oscillation and sonality and persistence of low-frequency atmospheric circulationBarnston, A.G. patterns. and R.E. Mon. Livezey, Wea. Rev.,1987: 115, Classification, 1083–1126. sea - Chronis, T., D.E. Raitsos, D. Kassis, and A. Sarantopoulos, Linderholm, H.W., A. Seim, T. Ou, J-H. Jeong, Y. Liu, X. Wang, G. Bao, and C. Folland, 2013: Exploring teleconnec tions between the summer NAO (SNAO) and climate in
59 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 East Asia over the last four centuries – a tree-ring per spective. Dendrochronologia 31: 297-310.
2009:Mann, Global M.E., Z.signatures Zhang, S. and Rutherford, dynamical R.S.origins Bradley, of the M.K. Lit- tleHughes, Ice Age D. and Shindell, Medieval C. Ammann, Climate Anomaly. G. Faluvegi, Science, and F.326, Ni, 1256-1260.
- pheric circulation changes are associated with the recent lossOverland, of Arctic J.E., sea and ice, M. Tellus Wang, A, 62: 2010: 1‐9. Large‐scale atmos
Petrie, RE., Shaffrey L. C. and R. T. Sutton, 2015: Atmos-
Atmosphere Simulation. J. Clim, 28 9606-9622. pheric Impact of Arctic Sea ice Loss in a Coupled Ocean– - ropean summer precipitation. Env. Res. Lett., 8, doi:10.1088/1748-9326/8/4/044015.Screen, J.A. 2013: Influence of Arctic sea ice on Eu
- son, 2013: Reconstructed drought variability in east-cen- tralSeftigen, Sweden K., H.W.since Linderholm, the 1650s. InternationalI. Drobyshev, andJournal M. Niklas of Cli- matology 33: 2449-2458. on a shift in European climate in the 1990s, Nat. Geosci., 5,Sutton, 788–792, R. T. doi:10.1038/ngeo1595.and B. Dong, 2012: Atlantic Ocean influence
Relationship between Winter Sea ice and Summer Atmos- phericWu, B.Y., Circulation R.H. Zhang, over R. D'Arrigo,Eurasia. J and Climate, J.Z. Su, 26, 2013: 5523-5536. On the
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 60 A last millennium perspective on North Atlantic variability: exploiting synergies between models and proxy data doi: 10.22498/pages.25.1.61 Pablo Ortega1, Jon Robson1, Paola Moffa-Sanchez2, David Thornalley3, Didier Swingedouw4 1 NCAS-Climate, Department of Meteorology, University of Reading, Reading, UK 2 School of Earth and Ocean Sciences, Cardiff University, Cardiff, UK 3 Department of Geography, University College London, London, UK 4 EPOC/CNRS, Université de Bordeaux, Bordeaux, France
Introduction The North Atlantic is a key region for decadal prediction 2001), which have all been linked with widespread climate over the observed period. This variability, which is impactsthe Atlantic over Multidecadal the surrounding Variability continents. (AMV) (Enfield Modelling et al., thoughtas it has experiencedto be intrinsic significant to the multi-decadalregion, can potentially variability studies suggest that all these modes interact with the modulate, either by amplifying or mitigating, the global warming signal from anthropogenic greenhouse areAMOC complex (Gastineau and remain and Frankignoul, to be disentangled. 2012; Hátún Also etto al.,be Atlantic contributed to the recent hiatus period between determined2005; Knight are et al.,the 2005) underlying but the mechanisms exact interrelationships responsible 1998emissions. and 2012, For example, by triggering studies an suggestatmospheric that theresponse North different climate models showing different key drivers McGregor et al., 2014). The subpolar North Atlantic is (Menaryfor the decadal et al., 2015a). and centennial Similarly, AMOC the exact modulations, impact of with the which impacted on the eastern tropical Pacific (e.g. 2 sink, and therefore of great importance natural external forcings (e.g. volcanic aerosols, solar for the global carbon cycle (Perez et al., 2013). irradiance) on the variability of these different large- also a major CO scale climate modes still remains unclear.
A unique opportunity to deepen our understanding whichOne of theis associated key players with in the sinking North Atlanticdue to regiondeep wateris the The study of the last millennium climate provides us Atlantic Meridional Overturning Circulation (AMOC), with an ideal framework to investigate natural climate the primary control of the poleward heat transport in the variability and associated mechanisms within the North formation in the Labrador and Nordic Seas. The AMOC is Atlantic. It is particularly interesting because it provides important climate impacts, and plays an active role in a long-term context of naturally forced variability which variousAtlantic feedbackregion. Therefore, mechanisms the with,AMOC for is example,associated sea with ice is useful (i) to assess whether current or future changes (Mahajan et al., 2011) and the atmospheric circulation in these variables are unprecedented, (ii) to robustly test the effects of natural forcings on their variability (e.g. by exhibited abrupt variations in the past (e.g. the last glacial increasing the sample size of major volcanic events), and period,(Gastineau Rahmstorf, and Frankignoul, 2002) and could 2012). experience The AMOC a major has (iii) to better characterise the typical timescales of the slowdown in the future due to the combined effect of surface warming and Greenland ice sheet melting on deep water formation (Bakker et al., 2016). The possibility Thekey variablesavailability at play of data(e.g. AMOC,to undertake AMV, SPG). these analyses of such a shutdown has stimulated considerable is rapidly increasing thanks to joint efforts from the international efforts to observe and reconstruct the modelling and paleoclimate data communities. The Paleoclimate Modelling Intercomparison Project (PMIP) variability will we be able to detect and anticipate an is now entering its fourth phase, and includes a set of past AMOC changes. Only by understanding its natural coordinated "tier 1" experiments for the last millennium
Decadalanthropogenic modulations impact are on thealso AMOC. found in other large-scale modes of climate variability, such as the North Atlantic Additional(Jungclaus etsensitivity al., 2016), withexperiments all models using,to explore for the firstthe uncertaintytime, the same in "default"external externalforcings forcingare also configuration. envisaged. The ultimate purpose of this exercise is to evaluate Oscillation (NAO) (Stephenson et al., 2000), the Subpolar Gyre strength (SPG) (Häkkinen and Rhines, 2004) and 61 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 the skill of models against well-documented climatic epochs, in order to reduce the uncertainty for future termedstrengthening the Little of Icethe Age AMOC (LIA). during These the results last two are centuriestherefore also(Fig. in1b), stark following contrast a minimum with the Rahmstorfduring the etcold al. interval(2015) surfaceclimate projections.temperature Additionally, (SST) synthesis the OCEAN2Kdataset (McGregor initiative reconstruction. Reconstructions of the vigour of the etwithin al., 2015), the PAGES2k including network 29 peer-reviewed has prepared and a firstpublicly sea available reconstructions from marine-archives in the changes across the last millennium. Interestingly, there is Atlantic ocean, all of them covering, at least partly, the evidenceNordic Seas of a Overflows potential anti-phase (Fig. 1c) show relationship multi-centennial between last 2000 years. Phase 2 of the OCEAN2K initiative aims the overflows East and West of Iceland, with the Denmark studyto advance paleoceanographic this field by reconstructions addressing different related topics,to the Strait Overflow Water (DSOW) strengthening when the dynamicalspecifically overturning two working changes groups in the North will compile Atlantic, andone Iceland Scotland Overflow Water (ISOW) is weaker, and Greenlandvice versa (Moffa-Sanchez Scotland Ridge et through al. 2015). the This last configuration millennium. model-data comparisons. suggests a constant flow of deep dense waters over the specifically focused on the proxy data, and the other in centennial variability, the inferred near-constancy of the Our current knowledge of the last millennium from If we assume that the AMOC does exhibit significant observations and paleoclimate records Because of the dynamic and large-scale nature of the variabilityNordic Overflows as suggested possibly by implicates Moffa-Sanchez changes et inal. Labrador (2014b) forSea theWater LIA, formation which would as a key parallel driver its of importantcentennial role AMOC in recent decadal changes. continuousAMOC, robust measurements observations of of its its strength variability date require back toextensive 2004, when (and the costly) RAPID measurement observing array arrays. at 26°N The firstwas We turn now our attention to other major contributors to North Atlantic climate variability in the last millennium. weakening of about 0.5 Sv per year (Smeed et al., 2014). The role of the atmosphere has been invoked to explain Andeployed. obvious The question first decadeis whether of observationsthis decline is exhibits linked to a another important centennial-scale climate event: the multi-decadal variability. Different approaches have been reconstruction (Trouet et al., 2009) shows persistent the effect of global warming or instead reflects natural strongMedieval positive Climate phases Anomaly during (MCA).this period, A bi-proxy followed NAOby a shift towards more negative phases that could have partly considered to reconstruct the AMOC changes back in presenttime and important give a longer uncertainties, context to this which trend; can however, contribute the connection of these indirect estimates with the AMOC can changescontributed are toless the evident MCA-LIA in a transition more recent (Fig. annually- 1d, light resolvedgreen line). reconstruction However, these based remarkable on multiple multi-centennial proxy records to conflicting conclusions. For example, Rahmstorf et al. A(2015) drawback uses AMOCof this covariances reconstruction with is SSTs that to it produce employs an a for prediction purposes, this recent reconstruction griddedAMOC reconstruction surface temperature for the lastreconstruction millennium (Mann (Fig. 1a). et (Fig. 1d, dark green line, Ortega et al., 2015). Of relevance al., 2008) mostly based on indirect proxy evidence from phases peaking 2 years after the eruptions. Mid-sized volcanicsuggests eruptionsthat volcanic can aerosolsalso impact can theinduce ocean positive and act NAO as been weakening since the beginning of the 20th century, a pacemaker of the intrinsic oceanic variability, as shown continentalwhich Rahmstorf areas. et This al (2015) index suggestssuggest is that a consequence the AMOC has of for two independent proxy reconstructions of the AMV Greenland ice sheet melting. A similar centennial trend is found in Dima and Lohmann (2010), which uses SST North of Iceland (Swingedouw et al., 2015). Likewise, observations to make inferences about the large-scale and the AMOC-driven changes in the nutrient supply example, the occurrence of Great Salinity Anomalies (Belkindecadal etclimate al., 1998). fluctuations All of these can beprocesses associated (forced with, and for independentcirculation. However, reconstructions other studies of the basedocean on circulation different unforced) can have different impacts on the variability of basedtechniques on sea contradict level data these(McCarthy results. et al., For 2015) instance, and deep two the major large-scale ocean modes in the North Atlantic. Labrador Sea densities (Robson et al., 2016) show no Indeed, the available reconstructions highlight prominent major long-term trends during the industrial period.
centennial changes in the AMOC (Fig. 1a), multidecadal changes in the AMV (Fig. 1e) and decadal changes in the On longer time-scales, rather than aiming to reconstruct DisentanglingSPG strength (Fig. the 1f). interplay between these different morethe AMOC easily asrealized. a whole, Proxy-based investigation reconstructions of individual of modes of variability and the wider climate system is surface and deep components of the AMOC may be still not possible due to the uncertainties and sparsity the surface ocean circulation near the North Icelandic of current reconstructions. Yet, paleoclimatology is a shelfthe Florida (Wanamaker Current et transportal., 2012) (Lundare both et suggestive al., 2006) of and a
growing field and the spatial distribution and number
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 62 have been largely inferred from continental records and therefore rely on atmospheric teleconnections that are still not totally understood. In addition, extending the current network of terrestrial records is also important to better constrain the concomitant atmospheric changes and continental impacts.
Insights from climate models Climate models provide a complementary source of information for the last millennium, allowing us to test different hypotheses, such as the external forcing conditions, and their effect on the major climate excursions (e.g. MCA, LIA, industrial global warming). Their horizontal and vertical resolution, as well as the representation of key physical processes (e.g. ocean eddies, aerosol-cloud and sea-ice interactions), are being continuously improved, offering unique access
aspect to most atmosphere-ocean general circulation to the complexity of the climate system. One common
models (AOGCMs) is that they naturally generate decadal diversityfluctuations in the in the mechanisms North Atlantic that lead under to fixedsuch externaldecadal forcing conditions. However, there is considerable suggest that multi-decadal oscillations (particularly linkedvariability. to the For AMV) example, can emerge studies in the with absence idealized of interactive models
DelSole, 2017). More generally, the preferential time- scaleocean of dynamics the internal (Clement variability et al., is associated 2015; Srivastava with ocean and adjustment processes that are strongly model dependent (Menary et al., 2015a), suggesting an important sensitivity to model biases (Menary et al., 2015b). Encouragingly, a multi-model comparison in control simulations (Ba et al., 2014) reports reasonable consistency in terms of
Figure 1: Last-millennium paleo-climate evidence for the North Atlantic: a) Surface temperature-derived AMOC andthe majormost ofinteractions them exhibit in thea lagged North relationship Atlantic; 8 outbetween of 10 reconstruction (Rahmstorf et al., 2015), b) Estimates of the models show a close link between AMV and the AMOC Florida current (blue line; Lund et al., 2006) and northward- of the models in Ba et al. (2014) appear to support a flowing surface transport across the North-Icelandic shelf the SPG changes and those in the AMOC. However, none (light blue line; Wanamaker et al., 2012), c) sortable silt )-derived Denmark Strait Overflow Water (DSOW) and at decadal timescales, a result inconsistent with other Iceland-Scotland Overflow Water (ISOW) flow speed (Moffa- significant relationship between the AMOC and the NAO Sanchez et al., 2015), d) reconstructed NAO evolution (dark green line, Trouet et al., 2009; light green lines, Ortega et al., 2014).studies Basupporting et al. (2014) a driving also rolenoted of thethat NAO salinity-driven on decadal 2015), e) estimates of AMV (light pink line, Mann et al., 2009; densityAMOC variability anomalies (e.g. seem Ortega to play et aal., dominant 2011; Mecking role in Northet al., dark pink line Gray et al., 2004) and f) SST-derived changes in the Subpolar Gyre Strength (Moffa-Sánchez et al., 2014a). All salinity contribution might be over-represented due to panels show anomalous values with respect to the common importantAtlantic convection, cold model and biases,therefore, which on thecould AMOC. potentially Yet, the period 1572-1787. All data were decadally smoothed, except compromise the realism of their described inter- for the Florida Current record, which is centennially resolved, relationships. and the NAO and overflow reconstructions, instead smoothed at 30 years to better highlight the multi-decadal changes. To date, only a limited number of studies have of high-resolution proxy records is continuously systematically assessed the effect of external forcings on increasing, especially for the last millennium, which should soon allow more reliable reconstructions. In particular, the production of new subdecadally resolved these modes of climate variability. For example, Gómez- Navarro and Zorita (2013) found no evidence of coherent about the past changes in the ocean. Until now, these variabilitychanges in wasNAO internally variability driven. across This,a large however, ensemble might of last be marine proxies is necessary to provide first-hand insights millennium AOGCM simulations, suggesting that all NAO
63 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 due to well-known limitations in the previous generation representation of the stratosphere, or to a simplistic implementationof AOGCMs (PMIP3 of andthe older),radiative either forcings. due to a Indeed, coarse the CNRM-CM5 model, which has a highly resolved stratosphere, and was not included in the previous to volcanic eruptions. Volcanic forcing is also found to exciteanalysis, an shows heterogeneous a consistent range strong of positiveresponses NAO of responseboth the simulations in Swingedouw et al. (2017). Thus, in light ofAMOC these and large AMV, model as uncertainties, shown for several proxy lastrecords millennium provide essential information to assess the degree of realism of models, and thus identify the most reliable ones.
Combining model and paleoclimate data There are multiple ways in which model simulations Figure 2: Spatial correlations between a selection of North Besides the obvious use of paleoclimate records as a Atlantic climate indices and the SST fields in two 300-yr long referenceand proxy benchmark reconstructions for climate can benefit models, from models each can other. also preindustrial control runs with HadGEM3-GC2 (Ortega et al., prove extremely useful (i) for the climatic interpretation 2016; top) and HiGEM (Shaffrey et al., 2009; bottom). All data of proxies (e.g., Bakker et al., 2015), (ii) to evaluate the were low-pass filtered at 10 years to highlight the decadal validity of different reconstruction techniques (e.g., variability. In-phase correlations are shown for the AMV and Moreno-Chamarro et al., 2017), and (iii) to guide future SPG strength indices. For the AMOC index, SSTs are delayed paleo-oceanographic efforts to new regions and variables by 6 years (the lag with maximum correlations). Significant values at the 95% confidence level are highlighted with thick with relevant climate information. grey contours. Yellow stars indicate the location of the SST reconstructions compiled by the OCEAN2K community The latter point can be addressed with model-derived (McGregor et al., 2015) and green circles the position of other temperature records also available (Risebrobakken et al., 2003; variability between the large-scale climate modes and Cage and Austin, 2010; Wurtzel et al., 2013; Moffa-Sánchez otherocean more fingerprints easily observed (Zhang, climate 2008), variables. highlighting These, co- et al., 2014a,b; Hoogakker et al., 2015; Miettinen et al., 2015). however, need to be considered with caution, as important differences can emerge from the various models, and Although the comparison of the co-variability patterns in these two high-resolution models is largely consistent, broader multi-model comparisons are still necessary also at different timescales (Muir and Fedorov, 2015). In Fig. 2 we explore the potential of ocean fingerprints millennium simulations are also required in order to test for the identification of suitable proxies to produce ifto andevaluate how thewhich inclusion fingerprints of forced are centennialrobust. Coarser variability last- separate distinct reconstructions of the AMV, AMOC and SPG strength. The figure depicts the correlation of impacts the simulated covariances - we suspect that it does, since previous studies with transient simulations controlthe simulated experiments SST fields with differentwith the oceanlarge-scale and atmosphere variability components.(AMOC, AMV, AllSPG) data in istwo decadally 300-yr high-resolutionsmoothed with 10-yearAOGCM exhibit a clear "warming hole" SST response to long-term running means to focus on the multidecadal co-variability. Interestingly, despite some apparent differences between the models, robust features are also discerned. The impact forced AMOC changes (Drijfhout et al., 2012; Rahmstorf climateet al., 2015), variables not that present are well in captured Fig. 2. The by proxies, multi-model such leads by 6 years, with both simulations showing an area asassessment sea level height.could be As extended an alternative to other to theAMOC-sensitive use of ocean of themaximum AMOC on correlations SSTs is most in pronounced the eastern when SPG, thefor AMOCwhich some SST-sensitive proxy records are available (green 13C records may also be fingerprints,used to infer proxiespast changes of deep-ocean in the deep flow ocean speed, circulation. based on with the AMV show a larger-scale structure that extends sortableAlthough silt informative, measurements, the interpretation and δ of some of these todots the and subtropical yellow stars North in Fig. Atlantic. 2). By Thus,contrast, the correlations addition of existing proxy records near West Africa and the Brazilian carbon-isotope enabled climate models highlight that coast could help to disentangle AMV variability from that proxies13C variability with regards cannot toalways AMOC be isinterpreted not straightforward, in terms of a coherent SST dipole between the Gulfstream and the δ Gulfof the of AMOC. Saint AsLawrence, for the SPG which strength, could both have models encouraging exhibit circulationchanging AMOC is not (Bakker always obvious, et al., 2015; and Blaschek sholud be et supported al., 2015). potential for reconstruction purposes due to the high throughLikewise, the the linkuse ofof localhigh-resolution flow speed toclimate the largescale models, availability of proxy records in both regions. and where possible, with related observational data.
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 64 paleoclimate records to better constrain past climate S0079-6611(98)00015-9 A final approach to combining model simulations and Oceanogr. 41, 1–68. doi:http://dx.doi.org/10.1016/ althoughevolution isthis the showed use of data limited assimilation reliability techniques. in the ANorth first Atlanticattempt oceanhas been (Goosse made et with al., 2010).a simplified Recently, climate the model,launch ocean-seaBlaschek, M.,ice Renssen, model H., Kissel,compared C., Thornalley, to proxy-based D., 2015. of a last millennium climate reanalysis (LMR) project reconstructionsHolocene North 1503–1524.Atlantic Overturning in an atmosphere- doi:10.1002/2015PA002828.Received reanalysis(Hakim et al.,covering 2016) thehas lastfostered millennium, research providing on this topic, key Cage, A.G., Austin, W.E.N., 2010. Marine climate variability informationand will hopefully about both lead externally-forced soon to the first and AOGCM-based internally- during the last millennium: The Loch Sunart record, driven changes in the North Atlantic. dx.doi.org/10.1016/j.quascirev.2010.01.014 The latest advances by the paleoclimate and modelling Scotland, UK. Quat. Sci. Rev. 29, 1633–1647. doi:http:// communities thus present us with a unique opportunity Clement, A., Bellomo, K., Murphy, L.N., Cane, M.A., to further our understanding of the main processes that Mauritsen, T., Rädel, G., Stevens, B., 2015. The Atlantic shaped climate variability in the North Atlantic over the last millennium. To this end, exploiting model-data circulation. Science (80-. ). 350, 320–324. synergies will be essential because it will help to improve Multidecadal Oscillation without a role for ocean reconstructions, and to identify the most reliable climate Dima, M., Lohmann, G., 2010. Evidence for Two model simulations. Changes over the Last Century. J. Clim. 23, 5–16. Acknowledgements doi:10.1175/2009JCLI2867.1Distinct Modes of Large-Scale Ocean Circulation We are grateful to David Lund (University of Connecticut) with us, and to all the researchers that made their data availablefor sharing online. the data The of GC2 the Floridamodel dataCurrent used reconstruction in this study Drijfhout, S., van Oldenborgh, G.J., Cimatoribus, A., 2012. Is a Decline of AMOC Causing the Warming Hole keeps the ownership rights under Crown Copyright. JCLI-D-12-00490.1above the North Atlantic in Observed and Modeled was kindly provided to us by the UK Met Office, which Warming Patterns? J. Clim. 25, 8373–8379. doi:10.1175/
ANR-13-BS06-0003).Finally, we acknowledge the support from NERC (project DYNAMOC, NE/ M005127/1) and ANR (HAMOC project, Enfield, D.B., Mestas-nuñez, A.M., Trimble, P.J., 2001. The References researchAtlantic Multidecadalarticle resubmitted Oscillation to Geophysical and its Relationship Research Letters.to Rainfall Atlantic and River28, 2077–2080. Flows in the Continental U . S . A
Ba, J., Keenlyside, N.S., Latif, M., Park, W., Ding, H., ALohmann, multi-model K., Mignot, comparison J., Menary, of M., Atlantic Otterå, O.H.,multidecadal Wouters, atmospheric response to the natural variability of the variability.B., Salas y Melia, Clim. D., Dyn. Oka, A.,43, Bellucci, 2333–2348. A., Volodin, doi:10.1007/ E., 2014. AtlanticGastineau, meridional G., Frankignoul, overturning C., circulation. 2012. Cold-season Clim. Dyn. s00382-014-2056-1 39, 37–57. doi:10.1007/s00382-011-1109-y
Bakker, P., Govin, A., Thornalley, D.J.R., Roche, D.M., modes in simulations over the past millennium: No long- termGómez-Navarro, response to J.J., external Zorita, forcing. E., 2013. Geophys. Atmospheric Res. Lett.annular 40, overturningRenssen, H., circulation: 2015. The Toward evolution a more of deep-ocean direct model- flow 3232–3236. doi:10.1002/grl.50628 dataspeeds comparison. and δ13C under Paleoceanography large changes in30, the 95–117. Atlantic doi:10.1002/2015PA002776
temperatureGoosse, H., Crespin, changes E., over de Montety, the past A.,600 Mann, years M.E., using Renssen, climate modelH., Timmermann, simulations with A., 2010.data assimilation. Reconstructing J. Geophys. surface Bakker, P., Schmittner, A., Lenaerts, J.T.M., Abe-Ouchi, A., Bi, Res. Atmos. 115, n/a-n/a. doi:10.1029/2009JD012737 D., van den Broeke, M.R., Chan, W.-L., Hu, A., Beadling, R.L., Marsland, S.J., Mernild, S.H., Saenko, O.A., Swingedouw, D., Gray, S.T., Graumlich, L.J., Betancourt, J.L., Pederson, G.T., warmingSullivan, A.,and Yin, Greenland J., 2016. melting.Fate of theGeophys. Atlantic Res. Meridional Lett. 43, 2004. A tree-ring based reconstruction of the Atlantic 12,212-252,260.Overturning Circulation: doi:10.1002/2016GL070457 Strong decline under continued Lett. 31, L12205. doi:10.1029/2004GL019932 Belkin, I.M., Levitus, S., Antonov, J., Malmberg, S.-A., 1998. Multidecadal Oscillation since 1567 A.D. Geophys. Res. ‘Great Salinity Anomalies’ in the North Atlantic. Prog.
Hakim, G.J., Emile-Geay, J., Steig, E.J., Noone, D., Anderson,
65 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 D.M., Tardif, R., Steiger, N., Perkins, W.A., 2016. The last temperature variations over the past two millennia. Proc. Natl. Acad. Sci. U. S. A. 105, 13252–13257. doi:10.1073/ pnas.0805721105 doi:10.1002/2016JD024751millennium climate reanalysis project: Framework and first results. J. Geophys. Res. Atmos. 121, 6745–6764.
North Atlantic circulation during the 1990s. Science 304, GlobalMann, M.E.,signatures Zhang, and Z., Rutherford, dynamical S.,origins Bradley, of theR.S., Little Hughes, Ice 555–559.Häkkinen, doi:10.1126/science.1094917 S., Rhines, P.B., 2004. Decline of subpolar AgeM.K., and Shindell, Medieval D., Ammann,Climate Anomaly. C., Faluvegi, Science G., Ni,326, F., 1256–2009. 1260. doi:10.1126/science.1177303 gyreHátún, on the H., thermohaline Sandø, A.B., circulation. Drange, Science H., Hansen, (80). 309, B., 1841–1844.Valdimarsson, H., 2005. Influence of the Atlantic subpolar climateMcCarthy, variability G.D., Haigh, revealed I.D., by Hirschi, sea-level J.J.-M., observations. Grist, J.P., NatureSmeed, 521, D. a., 508–510. 2015. Ocean doi:10.1038/nature14491 impact on decadal Atlantic inHoogakker, the Labrador B.A.A., Sea. McCave, J. Geol. I.N., Elderfield,Soc. 172, H., 272–277. Hillaire- doi:10.1144/jgs2013-097Marcel, C., Simstich, J., 2015. Holocene climate variability McGregor, S., Timmermann, A., Stuecker, M.F., England, M.H., Merrifield, M., Jin, F.-F., Chikamoto, Y., 2014. Recent Jackson, L.C., Peterson, K.A., Roberts, C.D., Wood, R.A., 892.Walker circulation strengthening and Pacific cooling 2016. Recent slowing of Atlantic overturning circulation amplified by Atlantic warming. Nat. Clim. Chang. 4, 888– as a recovery from earlier strengthening. Nat. Geosci 9, 518–522. M.-S.,McGregor, Sicre, H. M.-A.,V, Evans, Phipps, M.N., Goosse,S.J., Selvaraj, H., Leduc, K., Thirumalai, G., Martrat, B., Addison, J.A., Mortyn, P.G., Oppo, D.W., Seidenkrantz, cooling trend for the pre-industrial Common Era. Nat. Jungclaus, J.H., Bard, E., Baroni, M., Braconnot, P., Cao, GeosciK., Filipsson, 8, 671–677. H.L., Ersek, V., 2015. Robust global ocean KleinJ., Chini, Goldewijk, L.P., Egorova, K., Krivova, T., Evans, N., M.,LeGrande, González-Rouco, A.N., Lorenz, J.F., S.J.,Goosse, Luterbacher, H., Hurtt, J., G.C., Man, Joos, W., F.,Meinshausen, Kaplan, J.O., M., Khodri, Moberg, M., Mecking, J. V., Keenlyside, N.S., Greatbatch, R.J., 2014. Stochastically-forced multidecadal variability in the North Atlantic: A model study. Clim. Dyn. 43, 271–288. Schmutz,A., Nehrbass-Ahles, W., Schurer, C., A., Otto-Bliesner,Shapiro, A.I., Sigl, B.I., M., Phipps, Smerdon, S.J., doi:10.1007/s00382-013-1930-6 J.E.,Pongratz, Solanki, J., S.K., Rozanov, Timmreck, E., Schmidt, C., Toohey, G.A., M., Usoskin, Schmidt, I.G., H.,
Wood, R.A., 2015. A Mechanism of Internal Decadal Wagner, S., Wu, C.-Y., Yeo, K.L., Zanchettin, D., Zhang, Menary, M.B., Hodson, D.L.R., Robson, J.I., Sutton, R.T., ExperimentalQ., Zorita, E., Design 2016. for The the PMIP4 PMIP4 contribution past1000 simulations. to CMIP6 Climate Model. J. Clim. 28, 7764–7785. doi:10.1175/ Geosci.- Part 3:Model the LastDev. Millennium, Discuss. 2016, Scientific 1–34. Objectivedoi:10.5194/ and JCLI-D-15-0106.1Atlantic Ocean Variability in a High-Resolution Coupled gmd-2016-278
M.E., 2005. A signature of persistent natural thermohaline CMIP5Menary, model M.B., biases Hodson, on D.L.R.,the simulation Robson, of J.I., North Sutton, Atlantic R.T., circulationKnight, J.R., cycles Allan, in R.J., observed Folland, climate. C.K., Vellinga,Geophys. M., Res. Mann, Lett. decadalWood, R.A., variability. Hunt, J.A.,Geophys. 2015. Res. Exploring Lett. 42, the 5926–5934. impact of 32, 1–4. doi:10.1029/2005GL024233 doi:10.1002/2015GL064360
Lund, D.C., Lynch-Stieglitz, J., Curry, W.B., 2006. Gulf Stream density structure and transport during the A., 2015. Exceptional ocean surface conditions past millennium. Nature 444, 601–604. doi:10.1038/ onMiettinen, the SE A., Divine,Greenland D. V, shelfHusum, during K., Koç, the N., Jennings,Medieval nature05277 Climate Anomaly. Paleoceanography 30, 1657–1674. doi:10.1002/2015PA002849
ArcticMahajan, Surface S., Zhang, Air Temperature R., Delworth, and T.L., Sea 2011. Ice Impact Variability. of the J. Barker, S., 2014a. Solar forcing of North Atlantic surface Clim.Atlantic 24, Meridional 6573–6581. Overturning doi:10.1175/2011JCLI4002.1 Circulation (AMOC) on temperatureMoffa-Sánchez, and P., salinity Born, A.,over Hall, the I.R.,past Thornalley, millennium. D.J.R., Nat. Geosci 7, 275–278. doi:10.1038/ngeo2094 reconstructionsMann, M.E., Zhang, of hemispheric Z., Hughes, and M.K., global Bradley, surface R.S., D.J.R., Yashayaev, I., 2014b. Surface changes in Miller, S.K., Rutherford, S., Ni, F., 2008. Proxy-based Moffa-Sánchez, P., Hall, I.R., Barker, S., Thornalley,
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 66 the eastern Labrador Sea around the onset of the climatic trends in the North Atlantic since 2005. Nat. Little Ice Age. Paleoceanography 29, 160–175. Geosci. 9, 513–517. doi:10.1002/2013PA002523 Shaffrey, L., Stevens, I., Norton, W.A., Roberts, M.J.,
Stewart, C., 2015. Changes in the strength of the Nordic M.J., Demory, M.E., Donners, J., Clark, D.B., Clayton, Moffa-Sanchez, P., Hall, I.R., Thornalley, D.J.R., Barker, S., A.,Vidale, Cole, P.L., J.W., Harle, Wilson, J.D., S.S., Jrrar, Connolley, A., Stevens, W.M., D.P., Davi,Woodage, T.M., 123, 134–143. doi:10.1016/j.quascirev.2015.06.007 Seas Overflows over the past 3000 years. Quat. Sci. Rev. Resolution Global Environment Model—Model DescriptionMartin, G.M., and 2009. Basic U.K. Evaluation. HiGEM: J. TheClim. New 22, 1861–1896. U.K. High- Montoya, M., 2017. Assessing reconstruction techniques doi:10.1175/2008JCLI2508.1 Moreno-Chamarro, E., Ortega, P., González-Rouco, F., last millennium. Clim. Dyn. 48, 799–819. doi:10.1007/ s00382-016-3111-xof the Atlantic Ocean circulation variability during the Williams, E., Rayner, D., Johns, W.E., Meinen, C.S., Baringer, Smeed, D.A., McCarthy, G., Cunningham, S.A., Frajka- decline of the Atlantic meridional overturning circulation temperatures on decadal to centennial timescales: the M.O., Moat, B.I., Duchez, A., Bryden, H.L., 2014. Observed NorthMuir, L.C., Atlantic Fedorov, versus A. V, an 2015. interhemispheric How the AMOC seesaw. affects ocean Clim. 29-2014 Dyn. 45, 151–160. doi:10.1007/s00382-014-2443-7 2004 to 2012. Ocean. Sci. 10, 29–38. doi:10.5194/os-10- Srivastava, A., DelSole, T., 2017. Decadal predictability without ocean dynamics. Proc. Natl. Acad. Sci. . governing the predictability of the Atlantic meridional doi:10.1073/pnas.1614085114 overturningOrtega, P., Hawkins,circulation E., in a Sutton, coupled R., GCM. 2011. Clim. Processes Dyn. 37, 1771–1782. doi:10.1007/s00382-011-1025-1 Stephenson, D.B., Pavan, V., Bojariu, R., 2000. Is the North
V., Raible, C.C., Casado, M., Yiou, P., 2015. A model-tested Atlantic Oscillation a random walk? Int. J. Climatol. 20, 1–18. Ortega, P., Lehner, F., Swingedouw, D., Masson-Delmotte, doi:10.1002/(SICI)1097-0088(200001)20:1<1::AID- millennium. Nature 523, 71–4. doi:10.1038/nature14518 JOC456>3.0.CO;2-P North Atlantic Oscillation reconstruction for the past volcanicSwingedouw, eruptions D., Mignot, on the J., main Ortega, climate P., Khodri, variability M., Menegoz, modes. Mechanisms of decadal variability in the Labrador Sea Glob.M., Cassou, Planet. C.,Change Hanquiez, Submitted. V., 2017. Impact of explosive andOrtega, the wider P., Robson, North J.I.,Atlantic Sutton, in a R.T.,high-resolution Martins, A., climate 2016.
Masson-Delmotte, V., Butler, P.G., Khodri, M., Séférian, model. Clim. Dyn. Published Online. R.,Swingedouw, 2015. Bidecadal D., Ortega, North P., Atlantic Mignot, ocean J., Guilyardi, circulation E., variability controlled by timing of volcanic eruptions. Perez, F. F., H. Mercier, M. Vazquez-Rodriguez, P. 2 uptake reduced by weakening Nat. Commun. 6, 6545. doi:10.1038/ncomms7545 Lherminier,of the meridional A. Velo, overturning P. C. Pardo, circulation, G. Roson, Nat.and Geosci.A. F. Rios 6, (2013),146–152. Atlantic Ocean CO Trouet, V., Esper, J., Graham, N.E., Baker, A., Scourse,
Atlantic oscillation mode dominated the Medieval during the past 120,000 years. Nature 419, 207–214. ClimateJ.D., Frank, Anomaly. D.C., Science 2009. Persistent324, 78–80. positive doi:10.1126/ North doi:10.1038/nature01090Rahmstorf, S., 2002. Ocean circulation and climate science.1166349
A., Rutherford, S., Schaffernicht, E.J., 2015. Exceptional Rahmstorf, S., Box, J.E., Feulner, G., Mann, M.E., Robinson, changesWanamaker, in theA.D., North Butler, Atlantic P.G., Scourse, meridional J.D., Heinemeier, overturning J., overturning circulation. Nat. Clim. Chang. 5, 475–480. circulationEiríksson, J., during Knudsen, the K.L., last Richardson,millennium. C. Nat. a., 2012. Commun. Surface 3, doi:10.1038/nclimate2554twentieth-century slowdown in Atlantic Ocean 899. doi:10.1038/ncomms1901
Risebrobakken, B., Jansen, E., Andersson, C., Mjelde, E., Wurtzel, J.B., Black, D.E., Thunell, R.C., Peterson, L.C., Tappa, E.J., Rahman, S., 2013. Mechanisms of southern Caribbean paleoclimatic and paleoceanographic changes SST variability over the last two millennia. Geophys. Res. inHevrøy, the K.,Nordic 2003. Seas. A high-resolution Paleoceanography study of18, Holocene 1017. Lett. 40, 5954–5958. doi:10.1002/2013GL058458 doi:10.1029/2002PA000764
the Atlantic meridional overturning circulation. Geophys. Zhang,Res. Lett. R., 2008.35, L20705. Coherent doi:10.1029/2008GL035463 surface-subsurface fingerprint of Robson, J., Ortega, P., Sutton, R., 2016. A reversal of
67 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 Reconciling disparate views on decadal climate variability from proxies and models doi: 10.22498/pages.25.1.68 Toby R. Ault Dept. of Earth and Atmospheric Sciences, Cornell University, USA
Introduction Researchers studying decadal variability over the 2004) exhibit variable seasonal sensitivity to temperature instrumental period are often confronted with two and precipitation depending on the region. In the US Southwest, for example, the PDSI is highly sensitive compared to the timescales of interest, sampling at best onlymajor a obstacles. few realizations First, the ofobservational decadal-scale record phenomena is short depends more strongly on summer temperature. These (Meehl et al., 2009). Second, most climate variables to winter moisture, while in the Pacific Northwest, it include long-term trends driven by human activity (e.g., of tree growth on different environmental factors during land use change, aerosol pollution, and of course the seasonal dependencies reflect, in part, the dependence impact of greenhouse gas emissions), which sometimes consistent with basic dendroclimatological theory mask decadal variability from natural causes. The the seasonal cycle (St George and Ault, 2014), a finding climate research community therefore often turns to both paleoclimate archives of past changes, as well as relatively(Fritts, 1976). straightforward On interannual because timescales, data are diagnosing annually multi-century integrations of general circulation models resolvedthe filtering and effects overlap of treewith growththe instrumental on climate inputperiod. is (GCMs). Both types of data can provide insights into the amplitudes, patterns, and plausible mechanisms of decadal time horizons, and it remains a possibility that internal decadal variability, which could ultimately help treesHowever, grow this in response problem to has different not been climate widely factors studied across on inform and evaluate predictions of near-term climate evolution. In principle, proxy and GCM data should yield a consistent view of the climate system on these timescales. (ii)timescales Forward (e.g., models Franke of et paleoclimateal., 2013). archives might In practice, current paleoclimate data-model comparisons be biased by spatial and temporal patterns in GCMs. of decadal variability must contend with at least one of the challenges delineated below. To address these climate information, one might be tempted to simply concerns, I submit several heuristic recommendations runGiven GCM the output tendency through for proxies“forward to models” redden of and various filter to help to identify fundamental similarities—and critical proxy systems and compare the resulting output with differences—between paleoclimate and climate model actual archives. Caution would be recommended for perspectives on decadal variability of the last millennium. such an approach because models themselves exhibit systematic geographic and frequency biases. Consider (i) Paleoclimate archives filter climate variability in a case in which a forward model of tree-ring growth is ways that are difficult to quantify. run to predict annual ring-width anomalies as a function Most paleoclimate archives “redden” climate information of monthly temperature and precipitation (e.g. the by storing information from one time period to the “Vaganov-Shashkin-Lite" model of Tolwinski-Ward et
Dee et al., 2015). This reddening, in turn, has the effect output from a GCM with a wet bias (as is common for the next (e.g., Matalas, 1962; Evans et al., 2013; Ault 2013; Americanal., 2011; VS-Lite).Southwest), If this VS-Lite model would were produce to be run simulations with raw relative to their climatic drivers. Consequently, the mere where tree growth is never limited by the availability presenceof amplifying of high decadal amplitude fluctuations decadal variability in proxy in recordsa given of soil moisture, even during the “driest” year. Similar paleoclimate time series cannot be taken as evidence considerations apply to other types of proxy systems, of correspondingly energetic climatic variability (the and although standard bias-correction techniques are details of this effect are considered extensively in Ault et available for removing systematic model errors (e.g. al., 2013 and also Dee et al., in revision). Maurer et al., 2007), these tools have not been widely adopted for paleoclimate model-data comparisons. In addition to reddening the spectrum of underlying climate variables, many paleoclimate archives (iii) Climate teleconnections are not necessarily preferentially record information from certain seasons. stable through time. There are inherent biases in the structure of GCM teleconnections linking remote climate ring reconstructions of North American PDSI (Cook et al., For example, St. George et al. (2010) showed that tree- variations (e.g., in the Pacific basin) to the locations Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 68 where there are paleoclimate records (e.g., the American help isolate climate, as opposed to non-climate, sources of decadal variability. areSouthwest). not well Forsimulated example, by Coats some et models al. (2013) in the found American that El An example of how a few of these principles can be Southwest,Niño/Southern and Oscillation(b) are not teleconnections always stable inin allGCMs: models (a) from one century to the next. These considerations extent applied is shown in Fig. 1 (adapted from Ault et al., 2013). BliesnerHere NINO3.4 et al., spectra 2016) from are reconstructionscompared against (Emile-Geay the null to decadal timescales and observations data; Newman et et al., 2013) and last millennium model output (Otto- notal., (2016)be representative argued that of the decadal spatial variability pattern ofin thethat Pacific basin external boundary conditions (as in Ault et al., 2013). overDecadal the Oscillationlast millennium, (PDO) andduring hence the the20th teleconnections Century might distribution of ENSO variations with no changes to the driven by this climate mode may have been different generate the null distribution (see Ault et al., 2013 and in the past. Consequently, both GCMs and proxies may NewmanHere a linear et al., 2011 inverse for modeldetails). (LIM) At the haslongest been resolvable used to be susceptible to aliasing by changes in the large-scale timescales (centuries), the null hypothesis can be rejected structure of processes that generate decadal variability. for the reconstructions, but not for the model runs. At higher (interannual) frequencies, the reconstructions Suggestions to improve our understanding of decadal are well within the null distribution, whereas the model variability in proxies and models. oscillations are not (because this version of the model The list of considerations above implies at least four key principles should be followed when attempting to in comparison to observations). characterize decadal variability in a given system or produces ENSO fluctuations that are too high amplitude region using paleoclimate data and climate model output. While the null hypothesis can be rejected for the These include: centennial timescales in the reconstruction, and the 1. Comparisons are likely to be most meaningful if interannual ones in the model, it cannot be rejected for reconstructed phenomena are compared with the amplitudes of multidecadal (50-100 year) variations in either data type. This approach could help identify the or regional variations. Reconstructions of large- timescales that require the greatest attention by both scalemodel climate phenomena modes (e.g., tend Fig. to 1), rely as opposedon networks to local of paleoclimate and climate modeling research communities paleoclimate archives, often from different proxy to understand the processes responsible for generating types (e.g., Emile-Geay et al., 2013). Accordingly, such low-frequency variability.
as well as differences in spatial scales between model gridsnetworks and individualcan minimize sites. the Moreover, effects of if teleconnectionproxy filtering patterns change through time, a large-scale network of sites will be better suited to “see” the same phenomena even if its spatial imprint varies. 2. Decadal variability inferred from both paleoclimate and GCM sources should be evaluated against an
univariate setting, such a null hypothesis is usually appropriately defined null hypothesis. In a simple, more complicated systems, or for multivariate cases, athe more spectrum sophisticated generated method by an for AR(1) generating processes. the nullFor distribution might be needed. 3. Methodologies for comparing decadal variability in proxies and climate models should employ time Figure 1: Power spectra of NINO3.4 time series derived from a LIM (black lines with gray shading), multi-proxy series analysis and spectral techniques alike. While paleoclimate reconstructions (green; Emile-Geay et al., the former can help isolate the role of external 2013), and the CESM Last Millennium Ensemble inner forcings if the temporal evolution of those forcings quartile range (IQR) (red; Otto-Bliesner et al., 2016). The is known, the latter can identify timescales at which vertical dashed line marks the middle of the 2-7 year models and proxies exhibit fundamentally different peak typically associated with ENSO in observations amplitudes of variability. 4. Acknowledgements models of paleoclimate archives to characterize I would like to thank Scott St. George for helpful theFinally, imprint researchers of proxy should systems consider on the usingcontinuum forward of variability encoded in existing records (e.g., Dee et AGS 1602564. suggestions. This work was partially funded by NSF Grant
al., 2015; Dee et al., in revision). Such analyses will
69 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 References N., Kimoto, M., Kirtman, B., Navarra, A., Pulwarty, R., Smith,A. M., D., Hawkins, Stammer, E., D., Hegerl, and Stockdale, G., Karoly, T. (2009). D., Keenlyside, Decadal (2013).Ault, T. R., The Cole, continuum J. E., Overpeck, of hydroclimate J. T., Pederson, variability G. T., George, in westernS. S., Otto-Bliesner, North America B., Woodhouse, during the C. last A., andmillennium. Deser, C. Journal of Climate. Prediction: Can It Be Skillful? Bulletin of the American Newman,Meteorological M., Alexander, Society, 90(10):1467+. M. A., Ault, T. R., Cobb, K. M., Coats, S., Smerdon, J. E., Cook, B. I., and Seager, R. (2013). Deser, C., Di Lorenzo, E., Mantua, N. J., Miller, A. J., Minobe,
North America in CMIP5/PMIP3 model simulations. GeophysicalStationarity Research of the tropical Letters, 40(18):4927–4932. pacific teleconnection to S., Nakamura, H., Schneider, N., Vimont, D. J., Phillips, A. S., Scott, J. D., and Smith, C. A. (2016). The Pacific Decadal Cook, E. R., Woodhouse, C. A., Eakin, C. M., Meko, D. M., Landrum,Oscillation, L., Revisited. Stevenson, Journal S., Rosenbloom, of Climate, 29:4399–4427.N., Mai, A., and and Stahle, D. W. (2004). Long-term aridity changes in the Strand,Otto-Bliesner, G. (2016). B. L.,Climate Brady, Variability E. C., Fasullo, and Change J., Jahn, since A., western United States. Science, 306(5698):1015–1018. 850 CE: An Ensemble Approach with the Community Earth System Model. Bulletin of the American Meteorological Dee, S., Emile-Geay, J., Evans, M. N., Allam, A., Steig, E. J., and Society, 97(5):735–754. Thompson, D. (2015). PRYSM: An open-source framework for PRoxY System Modeling, with applications to oxygen- St. George, S. and Ault, T. R. (2014). The imprint of climate isotope systems. Journal of Advances in Modeling Earth Systems, pages n/a–n/a. Reviews, 89:1–4. within Northern Hemisphere trees. Quaternary Science St George, S., Meko, D. M., and Cook, E. R. (2010). The Emile-Geay, J. (under revision for Earth and Planetary seasonality of precipitation signals embedded within the ScienceDee, S., Parsons,Letters.). L., Improved Loope, G., spectralAult, T., Overpeck,comparisons J., and of paleoclimate models and observations via proxy system 988. modeling: implications for multi-decadal variability. North American Drought Atlas. Holocene, 20(6):983–
Emile-Geay, J., Cobb, K. M., Mann, M. E., and Wittenberg, climateTolwinski-Ward, controls on S., interannual Evans, M., variation Hughes, in M.,tree-ring and SST variability over the Past Millennium. Part 2: width.Anchukaitis, Climate K. Dynamics,(2010). An pages efficient 1–21. forward model of the ReconstructionsA. T. (2013). Estimatingand Implications. Central Journal Equatorial of Climate. Pacific
Evans, M. N., Tolwinski-Ward, S. E., Thompson, D. M., and Anchukaitis, K. J. (2013). Applications of proxy system
Science Reviews (accepted). modeling in high resolution paleoclimatology. Quaternary
Broennimann, S. (2013). Spectral biases in tree-ring climateFranke, proxies. J., Frank, Nature D., Climate Raible, Change, C. C., 3(4):360–364. Esper, J., and
Press. Fritts, H. C. (1976). Tree Rings and Climate. Academic
A. (1965). Tree-Ring Characteristics Along a Vegetation GradientFritts, H. in C., Northern Smith, D. Arizona. G., Cardis, Ecology, J. W., 46(4):393–401.and Budelsky, C.
MATALAS, N. C. (1962). STATISTICAL PROPERTIES OF TREE RING DATA. International Association of Scientific Maurer,Hydrology. E. P., Bulletin, Brekke, 7(2):39–47. L., Pruitt, T., and Duffy, P. B. (2007). climate change impact studies. Eos, Transactions AmericanFine-resolution Geophysical climate Union, projections 88(47):504–504. enhance regional
Meehl, G. A., Goddard, L., Murphy, J., Stouffer, R. J., Boer, G., Danabasoglu, G., Dixon, K., Giorgetta, M. A., Greene,
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 70 The third phase of the PAGES 2k Network doi: 10.22498/pages.25.1.71 PAGES2k Coordinators* PAGES IPO, Falkenplatz 16, Bern, Switzerland
Introduction The past 2000 years (the “2k” interval) provides critical Along with the many products and publications of the context for our understanding of recent anthropogenic regional groups, the initiative spawned a number of forcing of the climate, as well as baseline information about Earth’s natural climate variability. It also product was the coordinated publication of temperature provides opportunities to improve the interpretation of reconstructionssuccessful, large from network-wide seven continents, projects. which The informed first key paleoclimate proxy observations, and to perform out- of-sample evaluation of the climate and earth system Panel on Climate Change (PAGES 2k Consortium, 2013). models that are used to generate projections of future the Fifth Assessment Report of the Intergovernmental climate change. spanning the 2k interval were systematically compared. TheFor the results first time, showed continental-scale that there temperature were no historiesglobally- Achievements of the PAGES 2k Network In 2008 PAGES initiated the 2k Network, to coordinate worldwide “Medieval Warm Period” or “Little Ice Age” and integrate regional efforts to assemble existing consistent temperature fluctuations consistent with a proxy data and generate climate reconstructions. Nine cooling trend during the Common Era, which culminated regional groups were established during the course of in(Figure a cold 1). interval However, from there 1580 was to 1880 a near-global CE. In contrast, long-term the the initiative, spanning eight continents and the global period from 1971 to 2000 CE was the warmest during ocean. Phase 1 (2008-2013) focused on generating the last 1400 years. Through this publication, the 2k regional temperature reconstructions. During Phase Network received considerable attention outside the 2 (2014-2016), as a natural step forward, a number of trans-regional groups emerged from amongst the it remarks a milestone in making 2k paleoclimate science community, focusing on topical challenges such as accessiblefield of paleoclimate and understandable and amongst to the a widergeneral audience. public and It methods development, data-model comparison, database currently ranks in the 99th percentiles of Earth and construction and large-scale climate. Planetary Sciences articles of the same age and document
Figure 1: 30-year-mean temperatures for the seven PAGES 2k Network regions, standardized to have the same mean (0) and standard deviation (1) over the period of overlap among records (1190–1970 CE). North America includes a shorter tree-ring based and a longer pollen-based reconstruction. Adapted from PAGES2k Consortium (2013).
71 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 In collaboration with the Paleoclimate Modelling Intercomparison Project (PMIP), these regional reconstructions were compared with transient simulations of the last millennium (850 to 1850 CE).
the general tendencies, but temperature changes in differentThe resulting regions publication correlated more identified closely a with consistency each other in within the simulations than within the reconstructions (PAGES2k-PMIP3 group, 2015). Subsequent efforts by the Figure 2: Global SSTs over the last 2000 years: A cooling over the past two millennia was reversed only in the most cooling trend also occurred over the global ocean during recent two centuries. Fifty-seven previously published Ocean2k Working Group found that a robust long-term and publicly available marine sea surface temperature al., 2015). Comparison with climate model simulations reconstructions were combined and compiled into 200-year suggestedthe pre-industrial that this Common trend was Era (Figuredriven 2;by McGregorclusters of et brackets, represented by the boxes. The thin horizontal lines volcanic eruptions. dividing each box are the median of the values in that box. The thick blue line is the median of these values weighted for differences in the region of the global ocean in which they Combination of the continental reconstructions with were found. Modified from McGregor et al. (2015). temperature histories from ocean basins over the last
industrial-era warming, which could already be detected type in terms of citations, Mendeley readers and tweets, in500 the years mid-19th yielded century the curious over finding the tropical of an early oceans onset and of and it received 27 mass media mentions (www.scopus. com). Northern Hemisphere continents (Figure 3; Abram et
Figure 3: Onset of industrial-era warming in regional temperature reconstructions. Regional reconstructions since 1500 CE (coloured lines) with 15-yr (thin black lines) and 50-yr (thick black lines) Gaussian smoothing, shown alongside the median time of onset for sustained, significant industrial-era warming assessed across 15–50-yr filter widths (vertical red bars). Grey 1 °C scale bar denotes the y-axis scale of each regional temperature reconstruction. Modified from Abram et al. (2016).
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 72 Figure 4: Quality screened proxy records assembled in the PAGES2k temperature database at the end of Phase 2 (PAGES 2k Consortium, in press). Number of records per archive is indicated in the legend. Background image from http://visibleearth.nasa. gov. al., 2016). This trans-regional effort highlighted that the effect of greenhouse-gas forcing on temperatures started earlier than is suggested by instrumental data alone. Additionally, a spatial reconstruction of precipitation 2kResearch members. is organized This bottom-up as a linked concept network of oftrans-regional well-defined projects initiatedand targeted and conductedmanuscripts, by communityidentified and members led by years, in comparison with climate model simulations, has successfully emerged during Phase 2 and is expected suggestsacross the that Northern models do Hemisphere not yet accurately for the simulate past 1200long- to further stimulate collaboration within the 2k Network. term hydroclimate variability (Ljungqvist et al., 2016). systematically organize temperature-sensitive proxy data alongThe 2k regional projects boundaries. focus on specificAlong with scientific the products questions of andFinally, metadata a community-wide covering the effort 2k interval of Phase into 2 has a common been to thealigned individual with Phase projects, 3 goals, one ratheror more than community being defined wide database product, to facilitate future assessments of projects are envisaged for Phase 3. temperature variability during this period. As mentioned above, an enduring element from earlier Common to all these community products has been a phases of PAGES 2k will be a culture of collegiality, tremendous collaborative effort involving hundreds of transparency, and reciprocity. Phase 3 seeks to stimulate scientists from all regions of the globe, aiming to improve community based projects and to facilitate collaboration our understanding of mechanisms of climate variation on of researchers from different regions and career stages, interannual to bicentennial time scales by contributing drawing on the breadth and depth of the global PAGES 2k expert knowledge, data and metadata. As an example of community. A key vision of PAGES 2k is to support end- the impressive progress in data collection and synthesis, access, which are key conditions for the inclusion of screened temperature-sensitive proxy data (PAGES 2k futureto-end PAGESworkflow 2k transparency,projects. The initiativeopen data seeks and knowledge to further Consortium,Figure 4, shows in press). the current availability of quality- develop collaborations with other research communities and engage with stakeholders at the project and network level through interaction with related institutions Our vision for Phase 3 The goals of Phase 3 (2017-2019), which was launched EarthCube. or initiatives such as Future Earth, WCRP, IPCC and Call for participation 1.in May 2017 at the PAGES Open Science Meeting in There are many ways to participate in, and to be part Zaragoza,climate Spain, variability are to: and change on interannual to of, the PAGES 2k community. You may contribute to the centennialFurther understand time scales the (Theme: mechanisms “Climate driving Variability, regional ongoing database and knowledge-base efforts with your data and expertise. You may initiate a new 2k project, or 2. Reduce uncertainties in the interpretation of participate in an emerging one, by contributing towards observationsModes and Mechanisms”); imprinted in paleoclimatic archives project coordination, data-analysis, interpretation or by environmental sensors (Theme: “Methods and writing. In the spirit of Phases 1 and 2, PAGES 2k projects Uncertainties”) are expected to be inclusive and open to any researcher 3. Identify and analyse the extent of agreement between who wishes to contribute. Members from related reconstructions and climate model simulations communities including CLIVAR scientists are warmly (Theme: “Proxy and Model Understanding”) welcomed to be part of or even initiate 2k projects.
73 CLIVAR Exchanges No. 72, June 2017 Past Global Changes Magazine, Volume 25, No. 1 If you would like to participate in Phase 3 of the PAGES 2k Network or simply to receive updates, please visit http:// [email protected] www.pastglobalchanges.org/ini/wg/2k-network/ Steven Phipps, University of Tasmania, Hobart, Australia, intro to join our mailing list or contact a coordinating Scott St. George, University of Minnesota, Minneapolis, committee member. A call for new PAGES 2k projects will USA, [email protected] soon be issued via the mailing list.
References Bern, Switzerland, [email protected] Lucien von Gunten, PAGES International Project Office, McKay, N.P. Kaufman, D.S., Thirumalai, K., and PAGES 2k Abram, N.J., McGregor, H.V., Tierney, J.E., Evans, M.N.,
Consortium, 2016: Early Onset of Industrial-Era Warming across the Oceans and Continents. Nature, 536, 411–18.
Ljungqvist, F.C., Krusic, P.J., Sundqvist, H.S., Zorita, E., Nature,Brattström, 532, G. 94–98. and Frank, D., 2016: Northern Hemisphere Hydroclimate Variability over the Past Twelve Centuries. PAGES 2k Consortium, 2013: Continental-scale temperature variability during the past two millennia. Nature Geoscience, 6, 339-346.
PAGES 2k Consortium: A global multiproxy database for temperature reconstructions of the Common Era.
PAGES2k-PMIP3Scientific Data, in press.group, 2015: Continental-Scale Temperature Variability in PMIP3 Simulations and PAGES 2k Regional Temperature Reconstructions over the Past Millennium. Climate of the Past, 11, 1673–99.
M.,McGregor, Sicre, H.V.,M., Phipps,Evans, M.N., S.J., Goosse,Selvaraj, H., K., Leduc, Thirumalai, G., Martrat, K., B., Addison, J.A., Graham M.P., Oppo, D.W., Seidenkrantz, Cooling Trend for the Pre-Industrial Common Era. Nature Geoscience,Filipsson, H.L., 8, 671–77. and Ersek, V., 2015: Robust Global Ocean
*PAGES2k Coordinators: Nerilie Abram, The Australian National University, Canberra, Australia, [email protected]
Germany, [email protected] Oliver Bothe, Helmholtz-Zentrum Geesthacht, Geesthacht,
Sweden, [email protected] Hans Linderholm, University of Gothenburg, Göteborg,
Barcelona, Spain, & University of Cambridge, Cambridge, UK,Belen [email protected] Martrat, Spanish Council for Scientific Research,
Australia, [email protected] Helen McGregor, University of Wollongong, Wollongong, Raphael Neukom, University of Bern, Switzerland, [email protected]
Past Global Changes Magazine, Volume 25, No. 1 CLIVAR Exchanges No. 72, June 2017 74 Volume 25, No. 1 No. 72 June, 2017
Editorial Yochanan Kushnir, Christophe Cassou, Scott St George...... 1
An overview of decadal-scale sea surface temperature variability in the observational record Clara Deser, Adam Phillips...... 2
Global impacts of the Atlantic Multidecadal Variability during the boreal winter Yohan Ruprich-Robert, Rym Msadek...... 7
Arctic sea ice seasonal-to-decadal variability and long-term change Dirk Notz...... 14
Decadal climate variability and the global energy balance Richard P. Allan...... 20
Toward predicting volcanically-forced decadal climate variability Davide Zanchettin, Francesco S.R. Pausata, Myriam Khodri, Claudia Timmreck, Hans Graf, Johann H. Jungclaus, Alan Robock, Angelo Rubino, Vikki Thompson...... 25
Towards the prediction of multi-year to decadal climate variability in the Southern Hemisphere Scott Power, Ramiro Saurral, Christine Chung, Rob Colman, Viatcheslav Kharin, George Boer, Joelle Gergis, Benjamin Henley, Shayne McGregor, Julie Arblaster, Neil Holbrook, Giovanni Liguori...... 32
Initialization Shock in CCSM4 Decadal Prediction Experiments Haiyan Teng, Gerald A. Meehl, Grant Branstator, Stephen Yeager, Alicia Karspeck...... 41
Internal and forced decadal variability: lessons from the past millennium Hugues Goosse, François Klein, Didier Swingedouw, Pablo Ortega...... 47
Abrupt Northward Shift of SPCZ position in the late-1920s Indicates Coordinated Atlantic and Pacific ITCZ Change Braddock K. Linsley, Robert B. Dunbar, Donna Lee, Neil Tangri, Emilie Dassié...... 52
Summer North Atlantic Oscillation (SNAO) variability on decadal to palaeoclimate time scales Hans W. Linderholm, Chris K. Folland...... 57
A last millennium perspective on North Atlantic variability: exploiting synergies between models and proxy data Pablo Ortega, Jon Robson, Paola Moffa-Sanchez, David Thornalley, Didier Swingedouw...... 61
Reconciling disparate views on decadal climate variability from proxies and models Toby R. Ault...... 68
The third phase of the PAGES 2k Network PAGES2k Coordinators...... 71
The CLIVAR Exchanges is published by the International CLIVAR Project Office ISSN No: 1026-0471 The PAGES Magazine is published by the PAGES International Project Office ISSN No.: 2411-605X
Editors: Nico Caltabiano (ICPO) and Lucien von Gunten (PAGES IPO) Guest editors: Yochanan Kushnir (Lamont-Doherty Earth Observatory, USA), Christophe Cassou (CNRS-CERFACS, France) and Scott St George (University of Minnesota, USA)
Layout: Harish J. Borse, ICMPO at IITM, Pune, India This issue’s DOI: 10.22498/pages.25.1 Hardcopy circulation: 2300
Note on Copyright This publication is distributed under a CC-BY licence. Agreement should be obtained from the authors for the use of figures. The PAGES International Office and its publications are supported by the Swiss Academy of Sciences and the US National Science Foundation WCRP is sponsored by the World Meteorological Organization,the International Council for Science and the Intergovernmental Oceanographic Commission of UNESCO. Contact: Contact: PAGES International Project Office (IPO) Executive Director, ICPO Falkenplatz 16; 3012 Bern First Institute of Oceanography, SOA, Switzerland 6 Xianxialing Road, Laoshan District, Qingdao 266061,China [email protected] [email protected] http://www.pastglobalchanges.org http://www.clivar.org
Please recycle this newsletter by passing on to a colleague or library or disposing in a recognised recycle point