U.S. CLIVAR

Page 8

variability. Multi-decadal regime shifts inthe North Pacific (Mantua et al. 1997) leadto extended periods of relatively strongeror weaker ENs, depending upon whetherthese events develop out of a backgroundwarm or cool north Pacific regime. Thegeneral warming trend of the past centuryhas also resulted in an implied greateroverall impact of recent EN events(Mendelssohn et al. 2005). Understandingthe interactions between El Niño cyclesand other climate variability, and predict-ing their combined future impact onmarine ecosystems and fishery popula-tions, would be an important activity forCLIVAR to consider. ■

REFERENCESBograd, S.J., and Lynn, R.J., 2001. Physical-biologi-cal coupling in the California Current during the1997-99 El Niño – La Niña cycle, Geophys. Res.Lett., 28, 275-278.

Chavez, F.P., Collins, C.A., Huyer, A., and Mackas,D.L., 2002. El Niño along the west coast of NorthAmerica, Prog. Oceanogr., 54(1-4), 1-5.

Chelton, D.B., Bernal, P.A., and McGowan, J.A.,1982. Large-scale interannual and physical interac-tion in the California Current, J. Mar. Res., 30, 1095-1125.

Emery, W.J., and Hamilton, K., 1985. Atmosphericforcing of interannual variability in the northeastPacific Ocean: Connections with El Niño, J.Geophys. Res., 90, 857-868.

Kahru, M., and Mitchell, B.G., 2000. Influence of the1997-98 El Niño on the surface chlorophyll in theCalifornia Current, Geophys. Res. Lett., 27, 2937-2940.

Mantua, N.J., Hare, S.R, Zhang, Y., Wallace, J.M.,and Francis, R.C., 1997. A Pacific interdecadal cli-mate oscillation with impacts on salmon production.Bull. Am. Meteorol. Soc., 78, 1069-1079.

Mendelssohn, R., Schwing, F.B., and Bograd, S.J.,2003. The spatial structure of subsurface temperaturevariability in the California Current System, 1950-1993, J. Geophys. Res., 108,doi:10.1029/2002JC001568.

Mendelssohn, R., Bograd, S.J., Schwing, F.B., andPalacios, D.M., 2005. Teaching old indices newtricks: A state-space decomposition of El Niño relat-ed climate indices, Geophys. Res. Lett., 32,doi:10.1029/2005GL022350.

Mysak, L.A., 1986. El Niño, interannual variabilityand fisheries in the Northeast Pacific Ocean, Can. J.Fish. Aquat. Sci., 43, 464-497.

Palacios, D.P., Bograd, S.J., Mendelssohn, R., andSchwing, F.B., 2004. Long-term and seasonal trendsin stratification in the California Current, 1950-1993,J. Geophys. Res., 109, doi:10.1029/2004JC002380.

Roemmich, D., and McGowan, J.A., 1995. Climaticwarming and the decline of zooplankton in theCalifornia Current, Science, 267, 1324-1326.

Sette, O.E., and Isaacs, J.D., 1960. The changingPacific Ocean in 1957 and 1958, CalCOFI Reports,7, 13-127.

Wooster, W.S., and Fluharty, D.L., 1985. Eds., ElNiño North. Washington Sea Grant Program,University of Washington, Seattle, 312 pp.

1 Introduction

CLIMODE (CLIvar MOde WaterDynamic Experiment) isfocused on a region of hugeocean to atmosphere annual-

mean heat loss (>200 Wm-2) which occurs

over the separated Gulf Stream in theNorth Atlantic. The region of most intensewintertime ocean heat loss corresponds toan area with relatively warm surfacewaters that are carried there by the GulfStream, Fig.1. Late winter SST’s fall toapproximately 18˚C as water parcels moveeast under this cooling. The associatedbuoyancy loss from the ocean is believedto trigger ocean convection on the north-ern rim of the subtropical gyre to formwhat is known as Eighteen Degree Water(EDW) – Worthington (1959; 1976) – theNorth Atlantic Subtropical Mode Water.The wedge of weakly stratified waterspanning temperatures between about17˚C and 19˚C characteristic of modewater are clearly evident in the GulfStream section shown in Fig.2.

The region of EDW formation is par-ticularly relevant to wider CLIVAR goalsbecause, first, the annual mean ocean toatmosphere heat flux over the EDW for-mation region might be crucial for themaintenance of the Atlantic Storm track(Hoskins and Valdes, 1990). Second,EDW and the associated Gulf Streamrecirculation and thermal structure is a keyregion where oceanic timescales can pos-sibly imprint themselves on the atmos-

phere. Seasonal to interannual timescalesare introduced by the thermal inertia of theocean mixed layer/EDW layer system,whose evolution through the annual cycleis strongly connected to the re-emergenceof SST anomalies from winter to winter(Alexander and Deser, 1995; deCoëtlogon and Frankignoul, 2003). Onlonger timescales the intensity and path ofthe Gulf Stream affects air-sea exchangeand mode water formation through inter-annual variations in low-frequency flowas well as lateral eddy heat fluxes –Marshall et al (2001), Czaja and Marshall(2001), Dong and Kelly (2004). Howexactly such oceanic influences on climatework is a subject of great importance, con-troversy and subtlety. Finally, CLIMODEshould also be seen as making an impor-tant contribution to tying down the basinscale air-sea heat budget and, by implica-tion, quantifying the meridional transportof heat in the Atlantic basin.

CLIMODE is motivated by the factthat there is presently a major disconnectbetween the best available estimates ofEDW formation rates based on air-seafluxes and what we (think we) know aboutlikely dissipation rates. Either our air-seaflux estimates are grossly in error and/orthere is ‘missing physics’ involved in thebasic mechanism of mode water forma-tion, which is not represented in our mod-els. CLIMODE is designed to get to thebottom of this conundrum. A prime candi-date for the missing physics is lateral, dia-batic exchange through the mixed layer by

CLIMODE: a mode waterdynamics experiment in

support of CLIVARJohn Marshall

Massachusetts Institute of Technologyfor the CLIMODE group:

W. Dewar (FSU), J. Edson (U Conn), R. Ferrari (MIT), D. Fratantoni (WHOI), M. Gregg (UW), T. Joyce (WHOI), K. Kelly (UW),

R. Lumpkin (AOML), J. Marshall (MIT), R. Samelson (OSU), E. Skyllingstad (OSU), B. Sloyan (WHOI), F. Straneo (WHOI), L. Talley (Scripps), J. Toole (WHOI)

and R. Weller (WHOI). See http://www.climode.org/.

Page 9

VARIATIONSmesoscale eddy processes which, weargue below, play an order one balance inthe buoyancy budget.

Our working hypothesis inCLIMODE is that the onset of late winterconvection, when combined with GulfStream heat transport, intensifies themeridional slopes of near surface isopyc-nals, resulting in an explosion of baroclin-ic instability in the ocean. The northwardheat flux so generated is envisioned as bal-ancing much of the heat loss due to air-seainteraction as sketched in Fig.3 (right). Inocean climate models which do notresolve the eddies, this process mustappear as some sort of eddy advective/dif-fusive transport directed laterally throughthe mixed layer. But it is not yet at all clearhow to parameterize this process.

As the ocean surface is approached,eddy fluxes must develop a diapycnalcomponent because density is maintainedvertically homogeneous by strong surfaceboundary layer mixing whilst, as sketchedin Fig.3 (left), largescale eddying motionsare constrained to be horizontal by theupper boundary. We call this transitionlayer between the mixed layer and the adi-abatic interior, in which isopycnals areintermittently in contact with the turbulentmixed layer, the ‘surface diabatic zone’.We believe that this zone is likely to play akey role in mode water formation and dis-sipation. It is a key focus in CLIMODE.

The evidence that lateral eddy fluxesplay an important role in the dynamics ofthe upper ocean has only recently come tothe attention of the modeling community.It was in recognition of the importance ofnear-surface mixing in climate models thatthe Climate Process Team (CPT) EMILIE(EddyMIxed-Layer Interactions–see theEmilie web site maintained by Raf Ferrari:http://cpt-emilie.org/) was set up to fosterour understanding of the effect of transienteddy motions in the upper ocean and todevelop parameterizations of these effectsfor IPCC-class climate models.CLIMODE’s focus on the role of the sur-face diabatic zone in the cycle of mode-water formation provides a specific con-text in which the general issues of upper-ocean mixing can be addressed.

In this short article we briefly reviewthe science questions that motivateCLIMODE and the observational andmodeling plan designed to tackle it. In sec-

Figure 1: Wintertime net heat flux (colors in W/m2– COADS), selected SST outcrops (black lines) anddynamic height field (dotted lines, provided by the ECCO data assimilation scheme using the MITocean model). The black cross marks Bermuda.

tion 2 we return to the central conundrumof reconciling EDW formation and dissi-pation rates that is at the heart of ourexperiment. In section 3, we summarizethe observational and modeling elementsthat have been brought together to tacklethe problem. Contact information is inSection 4.

2 Reconciling EDW formationand dissipation rates

EDW is formed very close to or with-in the Gulf Stream where surface heat lossis large. Based on air-sea flux integrationsusing Walin’s (1982) framework, Speerand

Tziperman (1992) estimated a forma-tion rate of 15 to 20 Sv of EDW - see Fig.2

(right). The fundamental problem we areaddressing is why this rate is so much larg-er than the order 5 Sv inferred from sea-sonal changes based on profiling floats(e.g. Kwon and Riser, 2005) and impliedby thermocline diapycnal mixing rates.

2.1 The Walin framework

Walin considered the volume budgetof an isopycnal layer outcropping at thesea surface, integrated across the oceanfrom one coast to the other, as sketched inFig.3 (middle). He showed that even in atime-dependent, eddying ocean, A, thediapycnal volume flux across σ, could beexpressed precisely in terms of the diffu-sive fluxes, ‘D’, acting across the surfaceof the control volume, and air-sea fluxes

U.S. CLIVARVARIATIONS

Page 11

CLIVAR/OOPC/GOOS/ARGOWorkshop on the South Pacific11-14 October 2005Concepcion, ChileAttendance: LimitedContact: CLIVAR Office(icpo@soc.soton.ac.uk)

Tropical Atlantic Variability Workshop17-19 October 2005Venice, ItalyAttendance: InvitedContact: Paola Malanotte-Rizzoli(Rizzoli@mit.edu)

CRCES Workshop on DecadalVariability17-20 October 2005West Virginia, USAAttendance: OpenContact: http://www.crces.org

NOAA Climate Diagnostics andPrediction Workshop / Climate TestBed Meeting24-28 October 2005State College, PennsylvaniaAttendance: OpenContact: www.cdc.ncep.noaa.gov

AGU Fall Meeting5-8 December 2005San Francisco, CAAttendance: OpenContact: http://www.agu.org/meetings/

13th Ocean Sciences Meeting, a jointmeeting of ASLO, TOS and AGU20-24 February 2006Honolulu, HIAttendance: OpenContact: http://www.agu.org/meetings/

Seasonal to Interannual ClimateVariability: its Prediction and Impacton Society - NATO Advanced StudyInstitute (ASI)23 May – 3 June 2005Gallipoli, ItalyAttendance: OpenContact: Alberto Troccoli

AMS joint conference on Atmosphericand Ocean Fluid Dynamics, MiddleAtmospheres and Climate Variability13-17 June 2005Boston, MAAttendance: OpenContact: www.ametsoc.org

Pan WCRP Monsoon Workshop15-17 June 2005Irvine, CAAttendance: InvitedContact: icpo@soc.soton.ac.uk

International GEWEX Workshop20-24 June 2005Orange County, CAAttendance: OpenContact: www.gewex.org/5thconf.html

10th Annual CCSM Meeting21-23 June 2005Breckenridge, COAttendance: LimitedContact: www.ccsm.ucar.edu

Modes of Variability in the SouthernOcean Region27-28 June 2005Cambridge, United KingdomAttendance: InvitedContact: www.clivar.org

“The Ocean Carbon System: RecentAdvances and Future Opportunities”An Ocean Carbon and ClimateChange Workshop1-4 August 2005Woods Hole, MAAttendance: OpenContact: www.ioc.unesco.org/ioccp

The International Association ofMeteorology and Atmospheric Sciences(IAMAS) Biennial Scientific Assembly2-11 August 2005Beijing, ChinaAttendance: OpenContact:http://web.lasg.ac.cn/IAMAS2005

PAGES 2nd Open Science Meeting10-12 August 2005Beijing, ChinaAttendance: OpenContact: http://www.pages2005.org

U.S. CLIVAR Summit15-18 August 2005Keystone, COAttendance: InvitedContact: www.usclivar.org

Joint assembly of the InternationalAssociation of Geodesy, InternationalAssociation for Physical Sciences of theOceans and the InternationalAssociation for BiologicalOceanography22-26 August 2005Cairns, AustraliaAttendance: OpenContact: info@dynamicplanet2005.com

Calendar of CLIVAR and CLIVAR-related meetingsFurther details are available on the U.S. CLIVAR and International CLIVAR web sites: www.usclivar.org and www.clivar.org

U.S. CLIVAR

data by monitoring the seasonal cycle oflow PV waters. From the autumn to springvolume difference, the implied annualEDW production rate is 7.3 (float) or 3.5(climatology) Sv, much less than thatimplied by air-sea fluxes.

3. The elements of CLIMODEOf the two likely ameliorating influ-

ences, which can remedy the apparentimbalance between EDW production anddissipation, i.e. (1) lateraleddy fluxes in the mixedlayer, (which have onlybeen subject to rather coarseestimation), and (2) theinaccuracy of the estimationof transformation air-seaflux using climatologicalair-sea flux and SST data,we suspect the former ismore important, butCLIMODE is designed toaddress both processes.

CLIMODE has beenconstructed around a two-year period of field meas-urements (2006, 2007) with

particular emphasis on the late-winter/early-spring periods, times whenEDW ‘formation’ is highest. Observationswill be collected at high spatial resolutionover the top 500 m of the ocean to capturethe processes associated with mode waterformation in the context of the meander-ing front. Simultaneously, we will meas-ure the evolving marine boundary layerabove and document the air-sea fluxes thatdrive the two fluids. On longer timescales, the subsequent capping and initial

injection of the modewater into the subtropi-cal thermocline willalso be observed, aswell as its eventual dis-persal.

A variety of meas-urements and modelingactivities will be car-ried out underCLIMODE. Fig.4 (bot-tom) and Table 1 pro-vide an overview. Forthe two-year observa-tion period, moorings(one surface, two sub-surface) will be main-

tained in the EDW transformation regionsurrounded by an array of profiling floats.Continuous remote sensing of the oceansurface properties (SST, winds, sea levelanomalies) will also be carried out, in con-junction with an array of surface driftingbuoys. Extensive discussion of the obser-vational component of CLIMODE can befound at: http://www.climode.org/cruis-es.html.

In parallel with the observational pro-gram, modeling and theoretical studieswill be carried out. The modeling compo-nent of CLIMODE is directed at testingthe hypotheses that underlie the programdiscussed above, and, at the same time,will encourage transfer of understandingto the large-scale models used in climateresearch. A combination of regional andprocess ocean models will be used toaddress the phenomenology of EDW for-mation and dissipation. We are fortunate inhaving a strong common interest with theCPT-EMILIE (http://cpt-emilie.org) inupper-ocean mixing. In conjunction withthat program we plan to explore the wholerange of scales with a hierarchy of numer-ical models of increasing complexity.

Figure 3: (Left) Schematic diagram showing the interaction of a mixed layer (low PV) and the stratified interior (high PV) in a strong frontal region withoutcropping isopycnal surfaces, σ, undergoing buoyancy loss, B. Eddies forming along the front play a central role in controlling horizontal fluxesthrough the mixed layer and two-way quasi-adiabatic exchange between the mixed layer and the interior. (Middle) Application of the formalism due toWalin (1982): lateral diapycnal volume flux, A, whose divergence drives subduction, is related to ‘diffusive’ fluxes, D, acting across the boundary of theshaded control volume (which includes small-scale and diapycnal eddy fluxes) and air-sea buoyancy fluxes acting across the upper surface, F = ∂B/∂σ.(Right) Air-sea buoyancy loss triggering convection and EDW formation may be largely balanced by lateral diabatic eddy fluxes associated withmesoscale variability seen in Fig.4 (top). The sense of the eddy-induced flow in the ocean is also marked.

Continued from Page 10

Continued on Page 14

Our working hypothesisin CLIMODE is that the

onset of late winter convection, when

combined with GulfStream heat transport,

intensifies the meridional slopes of

near surface isopycnals,resulting in an

explosion of baroclinicinstability in the ocean.

Page 12

U.S. CLIVAR

Figure 4: (top) Wintertime SST fromthe AMSR-E microwave sensor, cour-tesy of Remote Sensing Systems; con-tour interval 1 degree. Positions of thesurface (blue) and subsurface (white)moorings are indicated. Note thewarm core of the Gulf Stream and theirregular opening of the EDW ventila-tion window (classically betweenabout 17.5 and 18.5˚C). Bias errors ofup to 0.5˚C may be present in thesenewly available data. (bottom)Schematic of CLIMODE fieldwork.Shown are nominal beginning andending locations for the spar drifts, theSeaSoar and XCTD survey patterns, asubset of microstructure samplingsites, and two hydrographic sectionlines. Positions of the surface and sub-surface moorings and two of thesound sources are also indicated.

Page 13

U.S. CLIVAR

4. Contact information andtimetable

Terry Joyce (WHOI) and JohnMarshall (MIT) have overall responsibilityfor the CLIMODE program. Terry Joyce isoverseeing the observational element;John Marshall is coordinating the theoryand modeling activities and the interactionof CLIMODE with the CLIVAR Ocean-Mixing CPT. The seagoing element ofCLIMODE begins in the November of2005, when moorings and floats will bedeployed. The intensive winter observa-tional periods follow in February 2006,2007. Floats will track dispersal of modewaters in subsequent years autonomously.

Many more details and latest informa-tion can be found from the CLIMODEwebsite (http://www.climode.org/).

5. AcknowledgementsWe thank the Physical Oceanography pro-gram of NSF, and particularly its directorEric Itsweire, for their support ofCLIMODE. We would also like to

acknowledge the support and advice of theAtlantic and US CLIVAR committees.

6. ReferencesAlexander, M. A., and C. Deser, 1995. A mechanismfor the recurrence of wintertime midlatitude SSTanomalies. J. Phys. Oceanogr., Vol. 25, 122—137.

Czaja, A. and J. Marshall, 2001. Observations ofAtmosphere-ocean coupling in the North AtlanticQJRMS, 127, 1893—1916.

de Coëtlogon and Claude Frankignoul, 2003. ThePersistence of Winter Sea Surface Temperature in theNorth Atlantic. J. Climate, 16, 1364—1377.

Dong, S. andK.A. Kelly, 2004. The heat budget in theGulf Streamregion: the importance of heat storageand advection. J. Phys. Oceanogr., 34, 1214—1231.

Garrett, C., Speer, K., Tragou, E., 1995. The relation-ship between water mass transformation and circula-tion with application to Phillips’ Red Sea model. J.Phys. Oceanogr. 25, 1696—1705.

Hoskins, B. J., and P. J. Valdes, 1990. On the exis-tence of storm-tracks. J. Atmos. Sci., 47, 1854—1864.

Kwon, Y.-O., and S.C. Riser, 2005. Eighteen degreewater of the North Atlantic observed using profiingfloats. Submitted to J. Phys. Oceanogr.

Large, W.G. and A.J. G. Nurser, 2001. Ocean surfacewater mass transformation. In Ocean circulation andClimate: Observing and modeling the global ocean,edited by G. Siedler, J. Church and J. Gould, pp.317—335, International Geophysics series, volume77, Academic Press.

Marshall, J., Kushnir, Y., Battisti, D., Chang, P.,Czaja, A., Dickson, R., McCartney, M., Saravanan,R., Visbeck, M. (2001) North Atlantic ClimateVariability: phenomena, impacts and mechanisms.Inter. Jour. Climatology, 21, 1863-1898

Speer, K., and E. Tziperman, 1992. Rates of watermass formation in the North Atlantic, J. Phys.Oceanogr., 22, 93-104.

Walin, G., 1982. On the relation between sea-surfaceheat flow and thermal circulation in the ocean. Tellus,34, 187—185.

Worthington, L.V., 1959. The 18˚ water in theSargasso Sea. Deep-Sea Res., 5, 297 305.

Worthington, L.V., 1976. On the North Atlantic circu-lation. Johns Hopkins Oceanographic studies, 6.

Continued from Page 12

Page 14

CLIMODE

Observations

Models

Air-sea fluxesF

Edson,Weller, Kelly

Direct air-sea fluxes

Moored atmosphereboundary layer observations

Remote sensing of SST,winds, sea level anomalies

Regional atmospheric modelSamelson, Skyllingstad

Eddies and mixingD

Joyce, Gregg,Toole,Lumpkin

Ocean µstructure profiles

Fine-scale Gulf Streamfrontal surveys

Lagrangian observationsof surface, upper ocean

velocity T/S

Process/regional oceanmode1

Dewer, Ferrari, Marshall

Subduction, dispersalA

Fratantoni, Sloyan,Straneo,Talley

Lagrangian & Eulerian observations of stratification

and bolus flux

EDW volumeobservations

Process regional oceanmodel

Dewer, Ferrari, Marshall

Table 1. Overview of measurement and modeling activities within CLIMODE