Improvement of the GEOS‐5 AGCM upon updating the air‐searoughness parameterization

C. I. Garfinkel,1 A. M. Molod,2,3 L. D. Oman,4 and I.‐S. Song5

Received 8 July 2011; revised 24 August 2011; accepted 25 August 2011; published 24 September 2011.

[1] The impact of an air‐sea roughness parameterizationover the ocean that more closely matches recent observa-tions of air‐sea exchange is examined in the NASA GoddardEarth Observing System, version 5 (GEOS‐5) atmosphericgeneral circulation model. Surface wind biases in theGEOS‐5 AGCM are decreased by up to 1.2m/s. The newparameterization also has implications aloft as improve-ments extend into the stratosphere. Many other GCMs (bothfor operational weather forecasting and climate) use a simi-lar class of parameterization for their air‐sea roughnessscheme. We therefore expect that results from GEOS‐5 arerelevant to other models as well. Citation: Garfinkel, C. I.,A. M. Molod, L. D. Oman, and I.‐S. Song (2011), Improvementof the GEOS‐5 AGCM upon updating the air‐sea roughness param-eterization, Geophys. Res. Lett., 38, L18702, doi:10.1029/2011GL048802.

1. Introduction

[2] The interaction between the ocean surface and thelowest levels of the atmosphere is a crucial component ofany atmospheric GCM. The exchange of momentum,moisture, and sensible heat between the ocean and atmo-sphere occurs on spatial and temporal scales far finer thanany GCM can directly simulate. Many models therefore relyon Monin‐Obhukov Similarity Theory (MOST) to specifyair‐sea exchange as a function of bulk winds, temperature,and humidity. Early attempts at quantifying the exchangecoefficients underlying MOST were conducted under con-ditions far removed from that actually experienced in theocean [e.g., Charnock, 1955; Large and Pond, 1981].Nevertheless, generations of atmospheric models have reliedon these earlier measurements for tuning their air‐searoughness scheme. For example, GEOS‐5 currently imple-ments Large and Pond [1981] for moderate and strongwinds and Kondo [1975] for weak winds [Helfand andSchubert, 1995]. See Table 1 for a description of theschemes in a range of models.[3] More recent in‐situ observations have improved our

understanding of air‐sea exchange over deep ocean waters,

especially over high wind regions like the Southern Ocean.In particular, recent field campaigns have measured turbu-lent exchange over the Southern Ocean, over the GulfStream, and over the North Atlantic in high wind speeds[e.g., Edson, 2008, also manuscript in preparation, 2011;Yelland et al., 1998; Edson et al., 2007; Banner et al., 1999].These field campaigns have found that the Charnok param-eter appears to increase with wind speed beyond 10 m/s, sothat a parameterization based on Charnock [1955] or Largeand Pond [1981] underestimates the drag on surface winds[Fairall et al., 2003, section 3c]. Recent observations of air‐sea exchange imply that the current air‐sea roughnessscheme in GEOS‐5 produces too little drag on surface windsin the range of wind speeds common in the Southern Ocean.[4] Accurate climatologies of surface winds over ocean

regions were not available when the current Large and Pond[1981]‐based parameterization in the GEOS‐5 model wascreated, but satellite‐based climatologies of surface windsare now available [e.g., Chou et al., 2003]. These satellitebased climatologies suggest that surface winds in theGEOS‐5 model are too strong over the Southern Ocean andoff the coast of Asia in the North Pacific (Figures 1a–1c andFigure S1 in the auxiliary material).1 As surface winds overthe Southern Ocean drive present and future oceanic uptakeof CO2 [Downes et al., 2011; Matebr and Hirst, 1999], it isimportant to accurately simulate surface climate in thisregion. The GEOS‐5 model is not alone in its poor repre-sentation of Southern Ocean surface wind; Barnes andHartmann [2010] find that the latitude of the SouthernHemisphere jet maxima varies by over 5° in the CoupledModel Intercomparison Project (CMIP3) ensemble, and thatsuch a bias has implications for the response of a GCM todoubled CO2. Regional models also have difficulty captur-ing mesoscale turbulent surface fluxes [Renfrew et al.,2009].[5] This paper discusses efforts to reduce this bias in

GEOS‐5 by updating the air‐sea roughness parameterizationfrom Helfand and Schubert [1995]. Section 2 describeschanges made to the model and Section 3 presents results.As other atmospheric GCMs base their air‐sea roughnessparameterization for momentum exchange on similarly olddata, we expect that the reduction in model bias shown heremight be common to other GCMs as well.

2. Change to Scheme

[6] We first describe the air‐sea roughness scheme inGEOS‐5 before discussing the changes made to increase thesurface friction. GEOS‐5 contains 72 vertical levels, withapproximately 8 in the boundary layer [Rienecker et al.,

1Department of Earth and Planetary Science, Johns HopkinsUniversity, Baltimore, Maryland, USA.

2Global Modeling and Assimilation Office, NASA Goddard SpaceFlight Center, Greenbelt, Maryland, USA.

3Earth System Science Interdisciplinary Center, University ofMaryland, College Park, Maryland, USA.

4Atmospheric Chemistry and Dynamics Branch, NASA GoddardSpace Flight Center, Greenbelt, Maryland, USA.

5Next Generation Model Development Project, Korea MeteorologicalAdministration, Seoul, South Korea.

Copyright 2011 by the American Geophysical Union.0094‐8276/11/2011GL048802

1Auxiliary materials are available in the HTML. doi:10.1029/2011GL048802.

GEOPHYSICAL RESEARCH LETTERS, VOL. 38, L18702, doi:10.1029/2011GL048802, 2011

L18702 1 of 6

2008]. Like many atmospheric GCMs, GEOS‐5 uses MOSTto describe momentum, heat, and moisture flux coefficientsin terms of bulk quantities (e.g., zonal wind, specifichumidity, and temperature) in the model. The wind stressvector at the surface can be expressed as

�x; �y� � ¼ �avsCDD u; v½ �; �j j ¼ �au*

2; ð1Þ

where ra is the air density, CD the transfer coefficient formomentum, vs the surface wind speed, D[u, v] is the dif-ference between the ocean and atmosphere surface windvector, and u* the friction velocity. MOST computes u*

(and CD) as a function of bulk parameters via the followingequations:

CD ¼ �2 YMO �z0ð Þ½ ��2;

u* ¼ C1=2D v

z0 ¼ A1

u*þ A2 þ A3u*þ A4u*

2 þ A5u*3

ð2Þ

where z0 is the roughness length, � is the Von‐Karmanconstant, A1 through A5 are tunable parameters used tomatch the air‐sea roughness scheme to observations, andYMO(z z0) is controlled by stability of the air column above.

Table 1. Momentum Exchange Over Oceans in Different GCMsa

Modeling Group Source Description

CAM2 Kiehl et al. [1998] based on Large et al. [1994]CAM5 Neale et al. [2010, p. 181] Large et al. [1994]GEOS‐5 (old) Helfand and Schubert [1995] Large and Pond [1981]AM2.0 GFDL Model Development Team [2004] Beljaars [1995], aCharnok = 0.018WRF(MM5) Skamarock et al. [2008, p. 72] based on Charnock [1955]uncoupled ECMWF European Centre for Medium‐Range Weather Forecasts [2010, p. 38] based on Charnock [1955], aCharnok = 0.018

aThe Community Atmosphere Model (CAM) has not upgraded its scheme since version 2.0 (Kiehl et al. [1998] versus Neale et al. [2010[,section 4.11.2]).

Figure 1. (a) Surface wind speed in the control run, (b) surface wind speed in observations, (c) control minus the observa-tions, and (d) the new run minus the control. For Figures 1a and 1b, the contour interval is 2m/s and the color scale is on thetop left. For Figures 1c and 1d, the contour interval is 0.7 m/s. For Figure 1c, the color scale is on the left. For Figure 1d,regions with anomalies whose statistical significance exceeds 95% are in color. The zero contour is omitted and negativecontours are dashed.

GARFINKEL ET AL.: UPDATED AIR‐SEA ROUGHNESS SCHEME L18702L18702

2 of 6

After an initial guess is made at CD (in practice CD assumingneutral stability), equation (2) is solved iteratively until anew value for CD has been reached consistent with theactual stability.[7] Previously, the A1 through A5 coefficients were chosen

to interpolate between the reciprocal relation of Kondo[1975] for weak winds and the piecewise linear relation ofLarge and Pond [1981] for moderate to large winds. Thekey change described by this paper is that the values for theA1 through A5 coefficients are changed so that the roughnesslength is increased for a given friction velocity. Neither theformulation for YMO(z z0) nor the coefficients at low windspeeds is changed. For very strong winds (e.g., hurricanes),roughness length no longer increases with wind speed (A.Molod and G. Partyka, The Impact on GEOS‐5 HurricaneForecasts of Limiting Ocean Roughness, submitted toJournal of Climate, 2011). See Table 2 for the coefficientsused. Runs with the old polynomial for z0 are referred to asCONTROL, and runs with the new polynomial for z0 arereferred to as NEW.[8] Figure 2a shows CDN10m as a function of 10m wind

speed for the old and new coefficients and in observations.The drag coefficient has been increased beyond the averagesuggested by the most recent observations but within theuncertainty. We chose the highest drag coefficient justifiedby the observations to achieve the maximum impact onthe GEOS‐5 wind bias. Any further increase would dis-tance us from the range of observational uncertainty. Notethat the drag coefficient in Community Atmosphere Model

(CAM/CCSM) (dashed red line) appears to be too small.CAM has not upgraded its scheme since version 2.0 (Kiehlet al. [1998] versus Neale et al. [2010, section 4.11.2]).Figure 2b compares modeled output roughness length andfriction velocity for the NEW and CONTROL runs. Asexpected, surface roughness dramatically increases withthe new coefficients. Figure 2b also includes curves ofz0 = aCharnoku*

2/g [Charnock, 1955] but with differentvalues of the Charnok parameter aCharnok. Older measure-ments suggest values of aCharnok∼0.011 to aCharnok∼0.018[see Fairall et al., 2003, section 3c]. Newer observations[Edson, 2008, also manuscript in preparation, 2011] wouldimply a higher aCharnok.[9] This change has been implemented in GEOS‐5.

Several GEOS‐5 atmosphere‐only simulations with the oldand new coefficients were performed to examine the impactof the increased drag:[10] 1. 2 × 2.5 degree 30 year runs with interactive

stratospheric chemistry,[11] 2. 2 × 2.5 degree 12 year run without interactive

stratospheric chemistry,[12] 3. 1 × 1.25 degree 25 year run without interactive

stratospheric chemistry,[13] 4. a series of 1/4 degree 5‐day forecasts.[14] All simulations showed similar impact of the new

roughness parameterization, and we will focus here onresults from the 30 year run with stratospheric chemistry.CONTROL and NEW differ only in the air‐sea roughnessscheme; all other models settings are fixed. A Student‐T

Table 2. Coefficients for MOST Scheme Equation Relating u* to z0a

A1 A2 A3 A4 A5

CONTROLu* < 0.0632456 0.2030325E – 5 0 0 0 00.0632456 < u* < 0.381844 −0.402451E‐08 0.239597E‐04 0.117484E‐03 0.191918E‐03 0.395649E‐040.381844 < u* −0.237910E‐04 0.228221E‐03 −0.860810E‐03 0.176543E‐02 0.784260E‐04

NEWu* < 0.0632456 0.2030325E – 5 0 0 0 00.0632456 < u* −1.102451E‐08 0.1593E‐04 0.1E‐03 2.918E‐03 0.695649E‐04

az0 =A1

u* + A2 + A3u* + A4u*2 + A5u*

3.

Figure 2. (a) Neutral drag coefficient for momentum exchange at the ocean surface (CDn10m) as a function of wind speed at10m in observations [Banner et al., 1999; Yelland et al., 1998] and in models. COARE3.0, COARE4.0, ECMWF wavemodel (i.e., not the uncoupled atmospheric model as in Table 1), and binned data are based on Edson [2008, also personalcommunication, 2011]. Error bars for binned data denote 1 standard deviation. Model results are from CAM2.0‐CAM5[Kiehl et al., 1998], and the original and new curves from GEOS‐5. (b) Relationship between friction velocity (u*) androughness length(z0) over all ocean gridpoints averaged over one day of GEOS‐5 model output. Isolines of z0 = aCharnoku*

2/g[Charnock, 1955] but with different values of the Charnok parameter aCharnok are included for comparison.

GARFINKEL ET AL.: UPDATED AIR‐SEA ROUGHNESS SCHEME L18702L18702

3 of 6

two‐tailed test is used to assess statistical significance. Eachyear is taken as one degree of freedom. Surface winds andsurface stress from Version 2 of the Goddard Satellite‐Based Surface Turbulent Fluxes (GSSTF) Data [Chou et al.,2003] are used to validate the model. We now address theimpact of this change in the air‐sea roughness parameteri-zation on bulk quantities in the model.

3. Results

[15] We now discuss how the change in friction influencesthe momentum budget in the model. The exchange coef-ficient for momentum increases over most oceanic regions,with the strongest increase over the Southern Ocean(Figure 3b). Biases in surface wind are reduced across theocean regions in response to the altered surface roughness

coefficients (Figures 1c and 1d). Winds over the SouthernOcean decrease by over 1m/s, but winds are reduced overmost ocean covered regions. Near surface geopotential heightanomaly biases are also reduced (not shown), consistent withthe wind speed improvement.[16] Figures 3c–3f show zonal surface stress on the ocean.

Changes in surface stress are smaller than changes in eitherCD or wind speed, as might be expected from equation (1).Namely, the decrease in wind speed and increase in CD

largely balance each other, so that their product is nearlyconstant. Nevertheless, the changes are significant in theSouthern Hemisphere, whereby surface stress on theSouthern Ocean is increased while surface stress furtherequatorward is decreased. The change is particularly strongin the Indian Ocean/Australia region. Biases in the controlrun are partially ameliorated. Runs in which the atmosphere

Figure 3. (a, b) Cm, drag coefficient for momentum exchange at the surface (Cm =CD*surface wind speed) in the control runand in the new run minus the control. Contour interval is 5 · 10−3 kgm−2 s−2 for Figure 3a and 10−3 kgm−2 s−2 for Figure 3b.Eastward surface stress at the surface in the (c) control run, (d) observations, (e) control‐observations, and (f ) new‐control.Contour interval is 5 · 10−2 Nm−2 for Figures 3c and 3d and 10−2 Nm−2 for Figures 3e and 3f. Regions with anomalies whosestatistical significance exceeds 95% are in color in Figures 3b and 3f. The zero contour is omitted and negative contours aredashed.

GARFINKEL ET AL.: UPDATED AIR‐SEA ROUGHNESS SCHEME L18702L18702

4 of 6

is coupled to a full ocean are planned in order to understandthe potential impact on the ocean circulation.[17] These changes in surface stress imply anomalous

eddy momentum flux convergence aloft, as verticallyaveraged @u′v′

@y must balance surface friction for a steady statesurface jet [Held, 1975; Vallis, 2006, section 12.1]. Figure 4shows that poleward momentum flux is increased through-out the upper troposphere, as implied by the dipole of sur-face stress. Eddies are fluxing more momentum poleward inorder to counteract the weakening of the surface jet. E. A.Barnes and C. I. Garfinkel (Barotropic eddy‐jet coexistenceand the response to surface friction, submitted to Journal ofAtmospheric Sciences, 2011) are investigating this changein momentum flux in much more detail. Associated withthis change in momentum flux are statistically significantimprovements in extratropical forecasting skill (not shown).

4. Conclusions

[18] The old air‐sea roughness scheme in GEOS‐5 isbased on 30‐year old observational data, but newer datasuggests seas are rougher. Associated with the old param-eterization are overly strong surface winds. By incorporatingmore recent observations of air‐sea exchange into themodel’s air‐sea roughness scheme, we have improved thesurface climate in the GEOS‐5 AGCM. Preliminary resultsindicate that the improvement is present at resolutions up to1/4 degree.[19] Modifying the air‐sea roughness parameterization

leads to statistically significant changes in cloud distribu-tion, heat flux, stratospheric ozone, and planetary wavedriving of the stratosphere. Presentation of these changes, adiscussion of the surface moisture and sensible heat budgets,will be reported in detail in a future paper. The microphysicsscheme in all runs considered does not include interactiveaerosols; preliminary results indicate that including inter-active aerosols along with this change in surface roughness

leads to large changes in sea salt aerosol concentration andsubsequent cloud formation.[20] Other atmospheric GCMs appear to use a similar

scheme to parametrize the exchange of momentum, heat,and moisture with the ocean. We expect that biases in theseother models might be reduced if these models were retunedto more closely match available observations.

[21] Acknowledgments. This work was supported by the NASAgrant NNX06AE70G. The authors thank J. Edson for making available datafrom Edson [2008] and his manuscript in preparation, Larry Takacs forperforming the model simulation in forecast mode, Andrew Eichmann forperforming the 1 degree model simulations, and our anonymous reviewersfor their comments.[22] The Editor thanks the two anonymous reviewers for their assis-

tance in evaluating this paper.

ReferencesBanner, M. L., W. Chen, E. J. Walsh, J. B. Jensen, S. Lee, and C. Fandry

(1999), The Southern Ocean Waves Experiment. Part I: Overview andmean results, J. Phys. Oceanogr., 29, 2130–2145, doi:10.1175/1520-0485(1999)029<2130:TSOWEP>2.0.CO;2.

Barnes, E. A., and D. L. Hartmann (2010), Testing a theory for the effect oflatitude on the persistence of eddy‐driven jets using CMIP3 simulations,Geophys. Res. Lett., 37, L15801, doi:10.1029/2010GL044144.

Beljaars, A. C. M. (1995), The parametrization of surface fluxes in large‐scale models under free convection, Q. J. R. Meteorol. Soc, 121,255–270, doi:10.1002/qj.49712152203.

Charnock, H. (1955), Wind stress on a water surface, Q. J. R. Meteorol.Soc., 81, 639–640, doi:10.1002/qj.49708135027.

Chou, S.‐H., E. Nelkin, J. Ardizzone, R. M. Atlas, and C.‐L. Shie (2003),Surface turbulent heat and momentum fluxes over global oceans basedon the Goddard Satellite Retrievals, Version 2 (GSSTF2), J. Clim., 16,3256–3273, doi:10.1175/1520-0442(2003)016<3256:STHAMF>2.0.CO;2.

Downes, S. M., A. S. Budnick, J. L. Sarmiento, and R. Farneti (2011),Impacts of wind stress on the Antarctic Circumpolar Current frontsand associated subduction, Geophys. Res. Lett. , 38 , L11605,doi:10.1029/2011GL047668.

Edson, J. B. (2008), Review of air‐sea transfer processes, paper presented atthe ECMWFWorkshop onOcean‐Atmosphere Interactions, Eur. Cent. forMedium‐Range Weather Forecasts, Reading, U. K., 10–12 Nov.

Figure 4. Momentum flux (u′v′) latitude‐height cross section. Contour intervals are (a, b) 12.5m2 s−2, (c) 2m2 s−2, and(d) 1m2 s−2 for. The zero contour is omitted and negative contours are dashed. Regions with anomalies whose statisticalsignificance exceeds 95% (95%) are in light(dark) blue in Figure 4d. Momentum fluxes are validated against ERA‐40[Uppala et al., 2005].

GARFINKEL ET AL.: UPDATED AIR‐SEA ROUGHNESS SCHEME L18702L18702

5 of 6

Edson, J., et al. (2007), The coupled boundary layers and air sea transferexperiment in low winds, Bull. Am. Meteorol. Soc., 88, 341–356,doi:10.1175/BAMS-88-3-341.

European Centre for Medium‐Range Weather Forecasts (2010), IFSdocumentation—Cy36r1 Operational implementation 26 January 2010.Part IV: Physical processes, ECMWF Tech. Note Cy36r1, Eur. Cent.for Medium‐Range Weather Forecasts, Reading, U. K. [Available athttp://www.ecmwf.int/research/ifsdocs/CY36r1/PHYSICS/IFSPart4.pdf.]

Fairall, C. W., E. F. Bradley, J. E. Hare, A. A. Grachev, and J. B. Edson(2003), Bulk parameterization of air sea fluxes: Updates and verificationfor the COARE algorithm, J. Clim., 16, 571–591, doi:10.1175/1520-0442(2003)016<0571:BPOASF>2.0.CO;2.

GFDL Model Development Team (2004), The new GFDL global atmo-sphere and land model AM2–LM2: Evaluation with prescribed sst simu-lations, J. Clim., 17, 4641–4673, doi:10.1175/JCLI-3223.1.

Held, I. M. (1975), Momentum transport by quasi‐geostrophic eddies,J. Atmos. Sci., 32, 1494–1496, doi:10.1175/1520-0469(1975)032<1494:MTBQGE>2.0.CO;2.

Helfand, H. M., and S. D. Schubert (1995), Climatology of the simulatedGreat Plains low‐level jet and its contribution to the continental moisturebudget of the United States, J. Clim., 8, 784–806, doi:10.1175/1520-0442(1995)008<0784:COTSGP>2.0.CO;2.

Kiehl, J. T., J. J. Hack, G. B. Bonan, B. A. Boville, D. L. Williamson, andP. J. Rasch (1998), The National Center for Atmospheric ResearchCommunity Climate Model: CCM3, J. Clim., 11, 1131–1150, doi:10.1175/1520-0442(1998)011<1131:TNCFAR>2.0.CO;2.

Kondo, J. (1975), Air‐sea bulk transfer coefficients in diabatic conditions,Boundary Layer Meteorol., 9, 91–112.

Large, W. G., and S. Pond (1981), Open ocean momentum flux measure-ments in moderate to strong winds, J. Phys. Oceanogr., 11, 324–336,doi:10.1175/1520-0485(1981)011<0324:OOMFMI>2.0.CO;2.

Large, W. G., J. C. McWilliams, and S. C. Doney (1994), Oceanic verticalmixing: A review and a model with a nonlocal boundary layer parame-terization, Rev. Geophys., 32, 363–403, doi:10.1029/94RG01872.

Matebr, R. J., and A. C. Hirst (1999), Climate change feedback on thefuture oceanic CO2 uptake, Tellus, Ser. B, 51, 722, doi:10.1034/j.1600-0889.1999.t01-1-00012.x.

Neale, L., et al. (2010), Description of the NCAR Community AtmosphereModel (CAM 5.0), NCAR Tech. Note, Natl. Cent. for Atmos. Res.,

Boulder, Colo. [Available at http://www.cesm.ucar.edu/models/cesm1.0/cam/docs/description/cam5_desc.pdf.]

Renfrew, I. A., G. N. Petersen, D. A. J. Sproson, G. W. K. Moore,H. Adiwidjaja, S. Zhang, and R. North (2009), A comparison of aircraft‐based surface‐layer observations over Denmark Strait and the IrmingerSea with meteorological analyses and QuikSCAT winds, Q. J. R.Meteorol. Soc., 135, 2046–2066, doi:10.1002/qj.444.

Rienecker, M. M., et al. (2008), The GEOS‐5 Data Assimilation System—Documentation of versions 5.0.1, 5.1.0, and 5.2.0, in Technical ReportSeries on Global Modeling and Data Assimilation, vol. 27, edited byM. J. Suarez, NASA Tech. Memo. NASA/TM‐2008‐104606, NASAGoddard Space Flight Cent., Greenbelt, Md. [Available at http://gmao.gsfc.nasa.gov/pubs/docs/Rienecker369.pdf.]

Skamarock, W. C., J. B. Klemp, J. Dudhia, D. O. Gill, D. M. Barker,M. G. Duda, X.‐Y. Huang, W. Wang, and J. G. Powers (2008), Adescription of the advanced research WRF version 3, NCAR Tech. Note475, Natl. Cent. for Atmos. Res., Boulder, Colo. [Available at http://www.mmm.ucar.edu/wrf/users/docs/arw_v3.pdf.]

Uppala, S. M. (2005), The ERA‐40 reanalysis, Q. J. Roy. Meteor. Soc.,131, 2961–3012.

Vallis, G. K. (2006), Atmospheric and Oceanic Fluid Dynamics:Fundamentals and Large‐Scale Circulation, Cambridge Univ. Press,Cambridge, U. K.

Yelland, M. J., B. I. Moat, P. K. Taylor, R. W. Pascal, J. Hutchings, andV. C. Cornell (1998), Wind stress measurements from the open oceancorrected for airflow distortion by the ship, J. Phys. Oceanogr., 28,1511–1526, doi:10.1175/1520-0485(1998)028<1511:WSMFTO>2.0.CO;2.

C. I. Garfinkel, Department of Earth and Planetary Science, JohnsHopkins University, Baltimore, MD 21209, USA. (cig4@jhu.edu)A. M. Molod, Global Modeling and Assimilation Office, NASA Goddard

Space Flight Center, Code 610.0, Greenbelt, MD 20771, USA.L. D. Oman, Atmospheric Chemistry and Dynamics Branch, NASA

Goddard Space Flight Center, Code 613.3, Greenbelt, MD 20771, USA.I.‐S. Song, Next Generation Model Development Project, Korea

Meteorological Administration, Shindaebang‐dong, Dongjak‐gu, Seoul156‐710, South Korea.

GARFINKEL ET AL.: UPDATED AIR‐SEA ROUGHNESS SCHEME L18702L18702

6 of 6