Large-Scale Circulation: Three Big Ideas

Home , Eddy (fluid dynamics)

Large-Scale Circulation: Three Big Ideas 1) Meridional Temperature and Momentum Transport (Mean Meridional Circulation) • Stationary and Transient Eddies 2) Large-Scale Tropics • ENSO 3) Climate Change • Weaker Overturning Circulation

1) Meridional Temperature and Momentum Transport

Atmosphere “Slower” Atmosphere “Faster”

Temperature and Potential Temperature Structure of the Atmosphere Zonal Wind Sturcture Structure of the Atmosphere ITCZ Intertropical Convergence Zone

• Latitude of Tropical Precipitation Maximum • Not Necessarily Latitude of Maximum Rising Motion

Seasonal Cycle of Rainfall

(from IRI)

Indian Monsoon

Australian Monsoon

ESS55 Prof. Jin-Yi Yu

1) Meridional Temperature and Momentum Transport • Stationary Eddies and Transient Eddies

Atmosphere “Slower” Atmosphere “Faster” Canadian High Siberian High Icelandic Low Aleutian Low Azores Bermuda High Pacific High Pacific High Azores Bermuda High Monsoonal Low GlobalGlobal DistributionDistribution ofof DesertsDeserts

Global Deserts Global

(from Global Physical Climatology) ESS55 Prof. Jin-Yi Yu Monsoons How ManyMany MonsoonsMonsoons Worldwide?Worldwide?

North America Monsoon Asian Monsoon

Australian Monsoon

South America Monsoon Africa Monsoon ESS55 (figure from Weather & Climate) Prof. Jin-Yi Yu

Transient and Stationary Eddy Flux of Westerly Momentum Transient and Stationary Eddy Flux of Temperature Large-Scale Circulation: Three Big Ideas 1) Meridional Temperature and Momentum Transport (Mean Meridional Circulation) • Stationary and Transient Eddies 2) Large-Scale Tropics • ENSO 3) Climate Change • Weaker Overturning Circulation

September-October-November Sea Surface Temperature Climatology

Warm Pool Cold Tongue

September-October-November Equatorial Temperature Climatology

Warm Pool Cold Tongue

Thermocline

Air-Sea Heat Flux

Heat Flux INTO the Ocean Heat Flux OUT OF the Ocean December 1982-2001 SST Climatology

December 1997 Total SST

December 1997 SST Anomaly

Warm Upper Layer Thermocline

Cold Lower Layer El Nino vs. Normal Thermocline Depth

Thermocline is Deeper than Thermocline is Shallower Normal than Normal Positive Depth Anomaly Negative Depth Anomaly December 1997 ( El Nino) Thermocline Depth Anomalies) Normal Conditions in the Tropical Pacific Warm (El Nino) Conditions in the Tropical Pacific Cold (La Nina) Conditions in the Tropical Pacific Walker Circulation Overturning Circulation

Evolution of the 1997-98 ENSO Event

Warm Sub-Surface Temperature (Deeper Thermocline) Anomaly is the Precursor of the Coming Warm Event – Why We Can Predict

Mature Warm Event

The Precursor of the Coming Cold Event

Teleconnections SST  Rainfall

Dec 1982

Nov 1988

410 ppm How We Understand Human Influence on Climate

Projections of Future Human Influence on Climate

Global Changes with Warming

NATURE GEOSCI ENCE DOI: 10.10 38/ NGEO868 Increase in Water Vapor Changes inREVIEW Precipitation AR TICLE

a 50 b 50

40 40

)

(

n o

30 a 30

t i

w 20 20

n i

P o

C 10 10

0 7.1% K–1 0 1. 2 % K–1

–2 02468 –2 02468 Temperature change (K) Temperature change (K) Expected From Clausius-Clapeyron Smaller than Expected -> Decrease in Overturning c 200 d 0.020 Circulation

0.015

)

– )

a 100

(

0.010

g s

x 0 0.005

o S

z 0.000

4 M

–100 0

N I

N 0.005 –27.7 hPa K–1 0.003 N m–2 K–1

–200 0.010 –2 02468 –2 02468 Temperature change (K) Temperature change (K)

Figure 1 | Changes in global and tropical Pacific mean climate from complex climate models as a function of the global mean t emperature change. Red diamonds are values derived from the CMIP3 multi-model database 8, black asterisks are values derived from perturbed physics ensembles with the HadCM3 model73. a, Percentage change in global mean column integrated water vapour. b, Percentage global precipitation change. c, Change in the mean sea-level-pressure (MSLP) gradient in hPa across the tropical Pacific basin24. d, Changes in the mean wind stress averaged in the NINO4 region in the central Pacific. Positive values indicate a reduction in the strength of the easterly trade winds. is caused by a dynamical adjustment to the reduced equatorial Because a reduction in the strength of the equatorial Pacif c trade winds25 and by the tendency of surface waters to warm at a faster winds is not necessarily accompanied by a reduction in the magni- rate than the deep ocean25,30,33. Hence the Bjerknes feedback does tude of the east–west gradient of SST — as would be expected from not operate on climate change timescales24 and other processes the typical relationship seen on interannual timescales — the ter m may inf uence the distribution of SST anomalies30,35–37. Despite the ‘El Niño-like’ climate change35 is of limited use when describing relatively robust observation of a reduction in the mean sea-level mean tropical Pacif c climate change in obser vations and models. pressure gradient across the equatorial Pacif c24,26–28,39, trends in It also creates confusion in many parts of the world wher e rainfall the observed east–west SST gradient are ambiguous25,28,29. changes expected as a result of global warming are dif er ent to those Averaged results from the dif erent Coupled Model normally associ ated with El Niño25,34. Intercomparison Project (CMIP3) models, forced by increas- T e changes outlined above are physically consistent, and ing green house gas concentrations over the next 100 years, show describe our understanding of expected variations in the mean that SSTs rise fast er along the equator than in the of -equatorial climate of the tropical Pacif c region under enhanced levels of regions32,36,38 (Fig. 2). T e CGCMs show that this is because the atmospheric greenhouse gases. Nevertheless, there are further weaker Walker circulation leads to a slowing of the horizontal complications. CGCMs have common spatial biases such as cold ocean circulation and reduced heat-f ux divergence throughout the tongues that extend too far west and unrealistic patterns of tropical equatorial Pacif c25,30, that is, less heat is transported away from the precipitation (e.g. ‘double InterTropical Convergence Zones’39). In equator. In the west , cloud-cover feedbacks and evaporation balance some climate models the existing biases in the cold tongue region the additional dynamical heating as well as the greenhouse-gas- of the eastern equatorial Pacif c are comparable in magnitude to related radiative heating. In the east , increased cooling by vertical the projected anthropogenic climate change signal. Moreover, heat transport within the ocean balances the additional warming there are aspects of the present global circulation that are either over the cold tongue. T e increased cooling tendency arises from not simulated well by CGCMs, or are not represented at all. For increased near-surface thermal st ratif cation, despite a reduction in example, organized intraseasonal variability in the form of the vertical velocity associated with the weakened trades31. Madden–Julian Oscillation (MJO)40,41 is absent in some CGCMs,

NATURE GEOSCIENCE | VOL 3 | JUNE 2010 | www.nature.com/ naturegeoscience 393

ngeo_868_JUN10.indd 393 19/5/10 11:25:44

REVIEW ARTICLE NATURE GEOSCI ENCE DOI: 10.10 38/ NGEO868

a Normal conditions

Walker circulation

Warm pool Cold tongue

Thermocline

REVIEW ARTICLE NATURE GEOSCI ENCE DOI: 10.10 38/ NGEO868 Upwelling a Normal conditions b El Nino~ conditions El Nino~ conditions (SST anomalies)

Walker circulation

Warm pool Cold tongue

Thermocline Thermocline Thermocline

Upwelling Upwelling Upwelling b El Nino~ conditions El Nino~ conditions (SST anomalies) c Climate change Climate change (SST anomalies)

Thermocline ThermoclineThermocline Thermocline

Upwelling Upwelling Upwelling Upwelling c Climate change Climate change (SST anomalies) Figure 2 | Idealized schematic showing atmospheric and oceanic conditions of the tropical Pacific region and their interactions during normal conditions, El Niño conditions, and in a warmer world. a, Mean climate conditions in thNote tropica lEl Pac ifiNinoc, indicati-ngLike SSTs, su rface wind stress and associated Walker circulation, the mean position of c onvection, and the mean upwelling and position of the thermocline. b, Typical conditions during an El Niño event. SSTs are anomalously warm in the east; convection moves into the central Pacific; the trade winds weaken in the east and the Walker circulation is disrupt ed; the thermocline flatt ens and the upwelling is reduced. c, The likely mean conditions under climate change derived from observations, theory and CGCMs. The trade winds weaken; the thermocline flattens and shoals; the upwelling is reduced although the mean vertical temperature gradient is increased; and SSTs (shown as anomalies with respect to the mean tropical-wide warming) increase more on the equator than of . Diagrams with absolute SST fields are shown on the left, diagrams with SST anomalies are shown on the right. For the climate change fields, anomalies are expressed with respect to the basin average temperature change so that blue colours indicate a warming smaller than the basin mean, Thermocline not a cooling. Thermocline and where it is present, the characteristics do not fully resemble hypothesis is continually tested as we improve our observations, Upwelling observations. If the spatial resolution of most CGUpCweMllings continues understanding and representation of the climate system. to increase as slowly as it has done in the past, simulation of tropical cyclones and typhoons as an emergent property is still some The response of ENSO t o external forcing Figure 2 | Idealized schematic showing atmospheric and owceayan iocf c.o nWdihtieotnhse or f the ster opihceanl Poamciefinca r eagfioenc ta ntdh eth edier tianitlesr aocft iomnesa dnu ringT neo rmseanl sitivity of ENSO to climate change can be studied by conditions, El Niño conditions, and in a warmer world. a, Mcheaanng celism iant ec lcimonadtieti ownes dino t hneo ttr okpniocawl .P Bacuitfi wc, ein cdaicna htinygp oSSthTse,s sizuerf athcea tw iitn d sitnrevsess taignadt ing the past history of ENSO using proxy-based climate associated Walker circulation, the mean position of convection,is no andt ne theces smeanarily upeswseellingntial andto bpositione able tofo thesim thermocline.ulate all the bs,e T aysppiceacl tcso ndirteiocnosn dstururincgti ons42–44. CGCM studies of the mid-Holocene (6,000 years an El Niño event. SSTs are anomalously warm in the east; coofn vcelicmtioant em woitvehs ainbtsoo tluthe ec efndtrealitl Pya bciefifcor; teh ew tera cdaen w eixntdrsa wcte uakseenfu inl itnhfeor ea-st aagndo t)h ea Wnda lktehre Last Glacial Maximum (21,000 years ago) provide circulation is disrupt ed; the thermocline flatt ens and the upmwaellingtion isf rroeducm Ced.G Cc, MThse alibkeoluy tm tehaen cdoentdaitliso nosf ucnldimera ctleim cahtea nchgaen. gTe dies riveind fsigromht into ENSO dynamics45–48, as do studies that examine the observations, theory and CGCMs. The trade winds weaken; the thermocline flattens and shoals; the upwelling is reduced although the mean vertical temperature gradient is increased; and SSTs (shown as ano39m4al ies with respect to the mean tropical-wide warming) increase more on the equaNAtorTURE GEOSCIENCE | VOL 3 | JUNE 2010 | www.nature.com/naturegeoscience than of . Diagrams with absolute SST fields are shown on the left, diagrams with SST anomalies are shown on the right. For the climate change fields, © 2010 Macmillan Publishers Limited. All rights reserved anomalies are expressed with respect to the basin average temperature change so that blue colours indicate a warming smaller than the basin mean, not a cooling. ngeo_868_JUN10.indd 394 19/5/10 11:25:44 and where it is present, the characteristics do not fully resemble hypothesis is continually tested as we improve our observations, observations. If the spatial resolution of most CGCMs continues understanding and representation of the climate system. to increase as slowly as it has done in the past, simulation of tropical cyclones and typhoons as an emergent property is still some The response of ENSO t o external forcing way of . Whether these phenomena af ect the details of mean T e sensitivity of ENSO to climate change can be studied by changes in climate we do not know. But we can hypothesize that it investigating the past history of ENSO using proxy-based climate is not necessarily essential to be able to simulate all these aspects reconstructions42–44. CGCM studies of the mid-Holocene (6,000 years of climate with absolute f delity before we can extract useful infor- ago) and the Last Glacial Maximum (21,000 years ago) provide mation from CGCMs about the details of climate change. T is insight into ENSO dynamics45–48, as do studies that examine the

394 NATURE GEOSCIENCE | VOL 3 | JUNE 2010 | www.nature.com/ naturegeoscience

ngeo_868_JUN10.indd 394 19/5/10 11:25:44 NATURE GEOSCI ENCE DOI: 10.10 38/ NGEO868 REVIEW ARTICLE past millennium49,50. However, there is no direct palaeo-analogue

of the rapid greenhouse-gas-induced climate change that we are 0.4 Increased variability )

currently exper iencing. .

d .

s 0.2 (

Detecting externally forced changes in the characteristics of

d .

ENSO using obser vational and climate change simulations is dif- s

n i f cult because of the large intrinsic variations in ENSO behaviour, 0

CMIP3 e g

which can occur on multidecadal and centennial timescales, even n a

52–54 h –0 .2 in the absence of exter nal changes . T is problem can be par- C tially overcome in CGCMs by performing multiple runs with the Decreased variability same model and measuring forced changes against natural vari- –0 .4 C C C C E F G G G IN IP M M M P U U C G N S C G F F I S I I R C K K ability from long, unforced control experiments. However, in the S C I H O D D S M L R R M M M M R R A A L L S – – O O I– M M O – – –E C C C C C O O 3 3 – – M LS C C H M M G – – real world this is not possible, and naturally occurring variability . C M 5 – M M 3 3 C H H 1( M / g 3 4 .2 .2 M a a T k3 M 1 2 2 .0 ( ( d d 4 3 . .0 .0 .1 h m 2 C G could be masking changes driven by global warming. 7 0 P ir e . M ) I- e d 3 EM O s) r .2 3 M e 1 ENSO processes and feedbacks may be af ected by green house- s) gas-induced changes in mean climate or by direct changes to some of those physical feedbacks and this could, in turn, lead to changes Figure 3 | Projected changes in the amplitude of ENSO variability, in the characteristic amplitude or frequency of ENSO events. As as a response to global warming, from the CMIP3 models8,9. The illustrated in Fig. 3, some CGCMs show an increase in the ampli- measure is derived from the interannual standard deviation (s.d.) tude of ENSO variability in the future, others show a decrease, and of a mean sea-level-pressure index, which is related to the strength some show no statistically signif cant changes. Figure 3 is based of the Southern Oscillation variations. Positive changes indicate a on just one of many studies that have come to the same conclu- strengthening of ENSO, and negative changes indicate a weakening. 9,10,55–60 sions . Based on the assessment of the curCMIP5rent generatio n of Statistical signiﬁcance is assessed by the size of the blue bars, and the CGCMs, there is no consistent picture of changes in ENSO ampli- bars indicated in bold colours are from those CMIP3 CGCMs that are tude or frequency in the future. However, by assessing individual judged to have the best simulation of present-day ENSO characteristics 16 feedback processes separately in CGCMs, we can shed some light and feedbacks. on how ENSO might be af ected by climate change: gradient amplif es El Niño events during their growth phase. As Mean upwelling and advection. Both the mean upwelling of cold there is little change in the mean zonal SST gradient in CGCMs water in the easter n equatorial Pacif c and the mean subsurface (Fig. 2c), it is unlikely that this feedback would change signif - advection act to strengthen the climatological temperature gradi- cantly under climate change. However, it might be important if ents in the horizontal and the ver tical. If a positive ther mal anomaly the relative frequency of occurrence of dif erent types of ENSO occurs in the east Pacif c, then these processes damp that anom- modes changes31. T e zonal advective feedback is more promi- aly. Mean upwelling and mean advection in CGCMs are reduced nent in central Pacif c El Niño, or ‘Modoki’, variability in which under climate change due to the general weakening of the equa- SST anomalies occur principally in the central Pacif c without the torial trade winds25. T is would lead to a tendency for en hanced warm anomalies in the east. ENSO activity. Atmospher ic damping. T e atmospheric damping of SST anoma- T ermocline feedback. Changes to the eastern equatorial Pacif c lies is generally partitioned into components associated with sen- thermocline depth can af ect the character of El Niño. A f atten- sible and latent heat f uxes, and surface short wave (SW) and long ing of the equatorial thermocline on interannual timescales leads wave (LW) f uxes. In general we expect that SST anomaly damp- to a positive SST anomaly in the east Pacif c. As the climatologi- ing through surface f uxes will increase because of increased cli- cal thermocline shoals in CGCMs under greenhouse warming, matological SSTs15,17. T is increase would therefore tend to reduce the SST response to an anomaly in thermocline depth should ENSO variability. Surface f ux damping might also change because increase15. In CGCM projections, changes in the mean depth of mean cloud changes brought about by weakening of Walker cir- of the thermocline in the east Pacif c are af ected by two com- culation and/or changes in cloud properties. Cloud feedbacks and pensating processes; thermocline shoaling or rising up tends to their link to the two large-scale circulation regimes that operate in reduce the depth in the east, but a reduction of the equatorial the east Pacif c (subsidence and convective61) remain a large uncer- thermocline slope tends to deepen it24,25. T ese changes could tainty in CGCMs17,62, probably driving a large fraction of the ENSO be expected to enhance the amplitude of ENSO events under er rors in the control climate conditions of present-day CGCMs17. climate change. Atmospheric variability. Westerly wind variability in the west SST/wind stress (Ekman) feedback. A weakening of the wind stress Pacif c, of en associated with coherent intraseasonal variability during El Niño events on interannual timescales leads to positive and the MJO, has been shown to be important in triggering and SST anomalies as less cold water is pumped upwards from below the amplifying El Niño events63–66. T ermocline anomalies excited surface of the ocean. T ose positive SST anomalies further weaken in the west can propagate to the east, where they are amplif ed. the wind stress. T is ef ect could increase under climate change Climate change simulations in several CGCMs project a future because of the reduced mixed-layer depth that arises as a result of enhancement of the intraseasonal variability in the equatorial the reduced mean trade wind strength, and increased ther mal strat- Pacif c in response to greenhouse gas increase, and this is an if cation15,33. Wind stress anomalies could become more ef ective important factor for potential intensi f cation of the El Niño activ- at exciting SST anomalies; in addition, the wind stress response to ity38. However, it should be noted that the simulation of the MJO SST anomalies can become stronger in regions wher e SST increases and related activity is perhaps one of the major weaknesses of cur- are largest15, that is, on the equator. Both ef ects would tend to rent CGCMs, but is an area in which there is considerable poten- amplify ENSO. tial for improvement.

Surface zonal advective feedback. T is is a positive feedback in Other processes and feedbacks. Other processes have been shown the ENSO cycle. T e anomalous zonal advection of the mean SST to play a role in deter mining the precise character istics of ENSO

NATURE GEOSCIENCE | VOL 3 | JUNE 2010 | www.nature.com/ naturegeoscience 395

ngeo_868_JUN10.indd 395 19/5/10 11:25:45 El Niño Features and Processes

Walker Circulation

external atmospheric damping noise SST response to thermocline anomalies surface zonal current

SST response to wind stress anomalies zonal advective feedback

upwellingupwelling