SC-WACCM · SC-WACCM Physics • Based on CESM1(WACCM) • Ozone and CO 2 speciﬁed from prior...

SC-WACCM!and !

Problems with Specifying the Ozone Hole

R. Neely III1, K. Smith2, D. Marsh1,L. Polvani2 1NCAR, 2Columbia

Thanks to: Mike Mills, Francis Vitt and Sean Santos

Motivation

To design a stratosphere-resolving model that can be used for studies of middle atmosphere dynamics without the expense of running interactive chemistry.

SC-WACCM Physics• Based on CESM1(WACCM) • Ozone and CO2 specified from prior fully-interactive WACCM simulations • Excludes comprehensive chemistry - solves only for H2O, CH4, N2O, CFC-11 and CFC-12 • Radiative transfer:

• CAM-RT below ~65 km • Short-wave heating rates prescribed above >65 km from same ‘fully-interactive’

simulations • Non-LTE cooling calculated from model temperature and prescribed CO2 >65km

• No auroral physics • Parameterized non-orographic gravity waves as in WACCM • TMS turned on

X - 34 SMITH ET AL.: SPECIFIED-CHEMISTRY WACCM

Pre

ss

ure

(h

Pa

)

66 Levels

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Alt

itu

de

(k

m)

(a) WACCM/SC−WACCM Vertical Resolution

0

20

40

60

80

100

120

140

Pre

ss

ure

(h

Pa

)

26 Levels

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Alt

itu

de

(k

m)

(b) CCSM4 Vertical Resolution

0

20

40

60

80

100

120

140

NO, atomic oxygen, CO2 and

chemical and shortwaveheating rates prescribed.

Figure 1. Hybrid model levels for (a) WACCM/SC-WACCM and (b) CCSM4. In SC-WACCM,

below the merge region, shortwave and longwave heating rates are calculated as in CCSM4 while

above the merge region longwave heating rates are calculated as in WACCM and chemical and

shortwave heating rates (QRS) are prescribed from a prior integration of WACCM. The region

from 63 to 70 km, over which heating and cooling rates are merged between WACCM/SC-

WACCM and CCSM4, is shaded gray in panel (a). SC-WACCM, monthly mean, zonal mean

ozone from a companion WACCM integration is prescribed everywhere and, like QRS, monthly

mean, zonal mean NO, atmoic oxygen (O), CO2

prescribed only above the merge region.

D R A F T May 14, 2014, 11:33am D R A F T


Pre

ss

ure

(h

Pa

)

66 Levels

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Alt

itu

de

(k

m)

(a) WACCM/SC−WACCM Vertical Resolution

0

20

40

60

80

100

120

140

Pre

ss

ure

(h

Pa

)

26 Levels

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Alt

itu

de

(k

m)

(b) CCSM4 Vertical Resolution

0

20

40

60

80

100

120

140

NO, atomic oxygen, CO2 and

chemical and shortwaveheating rates prescribed.

Figure 1. Hybrid model levels for (a) WACCM/SC-WACCM and (b) CCSM4. In SC-WACCM,

below the merge region, shortwave and longwave heating rates are calculated as in CCSM4 while

above the merge region longwave heating rates are calculated as in WACCM and chemical and

shortwave heating rates (QRS) are prescribed from a prior integration of WACCM. The region

from 63 to 70 km, over which heating and cooling rates are merged between WACCM/SC-

WACCM and CCSM4, is shaded gray in panel (a). SC-WACCM, monthly mean, zonal mean

ozone from a companion WACCM integration is prescribed everywhere and, like QRS, monthly

mean, zonal mean NO, atmoic oxygen (O), CO2

prescribed only above the merge region.


SC-WACCM Resolution!• 1.9˚ latitude x 2.5˚ longitude • Same 66 levels as WACCM (fully-resolved stratosphere and mesosphere):

• model top at 5.1x10-6 hPa (~140 km) • 18 pressure levels between the surface and 100 hPa are identical to CCSM4 • Stratosphere: 17 levels in WACCM between 100 and 3 hPa (versus 8 in CCSM4) • 9 levels above 100 km

SC-WACCM Performance

SC-WACCM is half the computational cost of WACCM

Pre-Industrial WACCM and SC-WACCM Simulations

• WACCM & SC-WACCM!• 200 years, coupled 1850 pre-industrial control simulation!• daily and monthly output (SC-WACCM available on glade and soon on the ESG)!

!• CCSM4!

• 500 years, coupled 1850 pre-industrial control simulation with monthly output!• 54 years of daily output

Zonal Mean Differences in Wind and Temperature

SMITH ET AL.: SPECIFIED-CHEMISTRY WACCM X - 39

Figure 6. Di↵erence (WACCM minus SC-WACCM) in zonal mean temperature (a),(b) and

zonal wind for (c),(d) for December-January-February (DJF) (a),(c) and June-July-August (JJA)

(b), (d). Red (blue) contours are positive (negative) values. Contour intervals are 1 K and 1 m

s�1.


ΔT

ΔU

DJF JJA

Surface !climate


SAT

(K)

Latitude

(a) SAT (DJF)

−90 −60 −30 0 30 60 90200

220

240

260

280

300

320

SAT

(K)

Latitude

(b) SAT (JJA)

−90 −60 −30 0 30 60 90200

250

300

350

SLP

(hPa

)

Latitude

(c) SLP (DJF)

−90 −60 −30 0 30 60 90980

990

1000

1010

1020

1030

SLP

(hPa

)

Latitude

(d) SLP (JJA)

−90 −60 −30 0 30 60 90980

990

1000

1010

1020

1030PR

ECIP

(mm

day

−1)

Latitude

(e) PRECIP (DJF)

−90 −60 −30 0 30 60 900

2

4

6

8

PREC

IP (m

m d

ay−1

)

Latitude

(f) PRECIP (JJA)

−90 −60 −30 0 30 60 900

2

4

6

8

10

WACCMSC−WACCMCCSM4


Figure 13. Zonal mean surface air temperature (a), (b), sea-level pressure (c), (d) and

precipitation (e), (f) for December-January-February (DJF) (a), (c), (e) and June-July-August

(JJA) (b), (d), (f). Black, red and blue curves are WACCM, SC-WACCM and CCSM4.


DJF$ JJA#

Problems with Specifying an Ozone Hole

Sensitivity of climate to dynamically-consistent zonal asymmetries

in ozone

N. P. Gillett,1 J. F. Scinocca,1 D. A. Plummer,1 and M. C. Reader1

Received 9 January 2009; revised 17 March 2009; accepted 8 April 2009; published 22 May 2009.

[1] Previous investigations into the effect of zonalasymmetries in ozone on climate have compared simulationswith prescribed 3-D ozone, in which the ozone is notnecessarily consistent with the model dynamics, tosimulations with prescribed zonal mean ozone. We assessthe impact of zonal asymmetries in ozone by comparing acontrol simulation of a coupled chemistry version of theCanadian Middle Atmosphere Model (CMAM) in which theozone andmodel dynamics are consistent, with a simulation inwhich only the zonal mean of the ozone is passed to theradiative transfer scheme. These simulations reveal a robuststratospheric zonal-mean temperature and geopotential heightresponse to zonal asymmetries in ozone that is consistent withthat identified in previous studies and of a magnitudecomparable to observed trends. These results suggest that theinclusion of zonal asymmetries in ozone may be essential forthe accurate simulation of future stratospheric temperaturetrends. Citation: Gillett, N. P., J. F. Scinocca, D. A. Plummer, andM. C. Reader (2009), Sensitivity of climate to dynamically-consistent zonal asymmetries in ozone, Geophys. Res. Lett., 36,L10809, doi:10.1029/2009GL037246.

1. Introduction

[2] To date almost all coupled atmosphere-ocean climatemodels [e.g., Meehl et al., 2007], and many model simu-lations of the middle atmosphere [e.g., Ramaswamy et al.2001] have used specified zonal average ozone distribu-tions. Son et al. [2008] demonstrated that the SouthernHemisphere tropospheric circulation response to ozonerecovery is larger in a set of coupled chemistry models thanin a set of climate models in which zonal mean ozone isspecified. One possible reason for this difference is that thecoupled chemistry models included zonally asymmetricchanges in the ozone distribution. Several recent studieshave highlighted the importance of zonal asymmetries inozone for the simulation of stratospheric and troposphericconditions in the Northern Hemisphere [Kirchner andPeters, 2003; Gabriel et al., 2007], and in the SouthernHemisphere [Crook et al., 2008]. Zonal asymmetries inozone are particularly large in the Southern Hemisphereduring the breakup of the vortex, when the region ofmaximum ozone depletion is often displaced from the pole.Crook et al. [2008] demonstrated that accounting for zonalasymmetries in ozone in a simulation using observed ozonefrom a year with particularly strong zonal asymmetryresulted in stratospheric cooling comparable to that due toozone depletion itself. However, these studies all specified

fixed three-dimensional distributions of ozone: In suchsimulations the position of the ozone minimum is notconstrained by the dynamics, and may not be collocatedwith the dynamical polar vortex. Further, ozone-dynamicsfeedbacks [Nathan and Cordero, 2007] are not resolved,since the ozone distribution is specified.[3] Most studies on the influence of the zonal asymmetry of

ozone have focused on the stratosphere. However, large zonalasymmetries in ozone are found in the mesosphere and thermo-sphere associated with the diurnal cycle. Since ozone concen-trations are much higher at night than in the day at these levels,using a zonal mean of ozone has the potential to introduce a biasin the radiative heating rates at these levels. For this reasonPaul et al. [1998] restrict their ozone climatology to levels below0.3 hPa. However the widely-used [Li and Shine, 1995] zonal-mean ozone climatology extends to 0.0011 hPa, but usesdaytime ozone values from near-infrared airglowmeasurementsfrom the Solar Mesosphere Explorer [Li and Shine, 1995].However, in some cases climate models are run with prescribedozone from coupled chemistry models, particularly for futuresimulations, and in these cases the use of zonal mean ozonecould introduce a bias in the radiative heating rates.[4] One way in which the influence of zonal asymmetries

in ozone may be examined in a more realistic context is bycomparing a simulation of a coupled chemistry model, with asecond simulation in which the zonal mean of the calculatedozone is prescribed. Reddmann et al. [1999] carried out sucha comparison using the 3-D Karlsruhe simulation model ofthe middle atmosphere (KASIMA), accounting for the diur-nal cycle in ozone above 50 km by prescribing the daytimemean ozone above this level in the second simulation, andwith conditions at the 10-km lower boundary prescribed fromEuropean Centre for Medium Range Weather Forecasting(ECMWF) analyses. Their simulations start in July, meaningthat they are not able to realistically simulate the Antarcticwinter vortex or Antarctic ozone depletion. They simulate asingle northern winter, and use trace gases representative ofapproximately 1992. They examine temperature differencesbetween the two simulations in December and find only smalldifferences between the two simulations, which leads them toconclude that ‘for undisturbed ozone conditions (no polarozone hole) a realistic ozone climatology is sufficient formodel simulations’. In this study, we also assess the role ofthree dimensional ozone variations in a fully-coupled con-text, but using 40-yr simulations with a realistic annual cycleand stratospheric chlorine levels representative of approxi-mately 1990.

2. Model and Experiments

[5] We use the Canadian Middle Atmosphere Model withcoupled chemistry (CMAM) [Scinocca et al., 2008; de

GEOPHYSICAL RESEARCH LETTERS, VOL. 36, L10809, doi:10.1029/2009GL037246, 2009

1Canadian Centre for Climate Modelling and Analysis, EnvironmentCanada, Victoria, British Columbia, Canada.

Published in 2009 by the American Geophysical Union.

L10809 1 of 5

Effect of zonal asymmetries in stratospheric ozone on simulated

Southern Hemisphere climate trends

D. W. Waugh,1 L. Oman,1 P. A. Newman,2 R. S. Stolarski,2 S. Pawson,3 J. E. Nielsen,3

and J. Perlwitz4

Received 6 August 2009; revised 21 August 2009; accepted 27 August 2009; published 22 September 2009.

[1] Stratospheric ozone is represented in most climatemodels by prescribing zonal-mean fields. We examine theimpact of this on Southern Hemisphere (SH) trends using achemistry climate model (CCM): multi-decadal simulationswith interactive stratospheric chemistry are compared withparallel simulations using the samemodel in which the zonal-mean ozone is prescribed. Prescribing zonal-mean ozoneresults in a warmer Antarctic stratosphere when there is alarge ozone hole, with much smaller differences at othertimes. As a consequence, Antarctic temperature trendsfor 1960 to 2000 and 2000 to 2050 in the CCM areunderestimated when zonal-mean ozone is prescribed. Theimpacts of stratospheric changes on the troposphericcirculation (i.e., summertime trends in the SH annularmode) are also underestimated. This shows that SH trendsrelated to ozone depletion and recovery are underestimatedwhen interactions between stratospheric ozone and climateare approximated by an imposed zonal-mean ozone field.Citation: Waugh, D.W., L. Oman, P. A. Newman, R. S. Stolarski,S. Pawson, J. E. Nielsen, and J. Perlwitz (2009), Effect of zonalasymmetries in stratospheric ozone on simulated SouthernHemisphere climate trends, Geophys. Res. Lett., 36, L18701,doi:10.1029/2009GL040419.

1. Introduction

[2] It is now well established that the ozone hole hasplayed a major role in changes in the summer troposphericcirculation of the southern hemisphere (SH) over the last twodecades, and that the expected recovery of Antarctic ozonewill likely also be a major factor in SH climate change overthe next fifty years [e.g., Thompson and Solomon, 2002;Marshall, 2003; Gillett and Thompson, 2003; Perlwitz et al.,2008; Son et al., 2008, 2009]. As a result it is importantto include the impact of ozone depletion and recovery insimulations (predictions) of changes in SH climate.[3] However, Perlwitz et al. [2008] and Son et al. [2008]

suggest that the impact of changes in stratospheric ozone onthe tropospheric climate may not be fully captured in theWorld Climate Research Programme’s Coupled Model Inter-comparison Project phase 3 (CMIP3) multi-model dataset

used in the Fourth Assessment of the IntergovernmentalPanel on Climate Change [Intergovernmental Panel onClimate Change, 2007]. They showed that, even in theCMIP3 models that prescribed ozone recovery, the tropo-spheric response is weaker than that in the coupled chemistrymodels (CCMs) in the SPARC Chemistry-Climate ModelValidation Activity (CCMVal), which calculate stratosphericozone interactively.[4] There are several possible reasons for the difference in

the tropospheric response in CMIP3 and CCMVal models,including lack of interactive chemistry, incorrect specifica-tion of ozone, inadequate representation of the stratosphericcirculation in the CMIP3 models, or the lack of a dynamicocean in the CCMVal models. Here we focus on the impor-tance of interactive stratospheric chemistry and the impact ofprescribing monthly-mean zonal-mean ozone (as is done inthe CMIP3 models). Sassi et al. [2005], Crook et al. [2008],and Gillett et al. [2009] have shown that the Antarctic vortexis weaker and warmer in simulations without zonal asymme-tries in O3. This suggests that the use of prescribed zonal-mean ozone in the CMIP3 models may be the cause of thedifference from CCMVal models. However, the above stud-ies considered only conditions with high levels of ozonedepleting substances (ODSs) and a large ozone hole, andthey did not examine the impact on long-term trends in thestratosphere or troposphere.[5] In this study we examine the impact of zonal asymme-

tries in ozone on simulated trends by comparing simulationsfor 1955 to 2055 from the NASA’s Goddard Earth ObservingSystem Chemistry-Climate Model (GEOS CCM) [Pawsonet al., 2008], which has full, interactive stratospheric chem-istry, with parallel simulations using the same CCM exceptthat the monthly-mean zonal-mean stratospheric ozone fromthe first simulation is prescribed.

2. Model and Simulations

[6] The GEOS CCM includes representations of atmo-spheric dynamics, radiation, and stratospheric chemistry andtheir coupling through transport and radiative processes.Pawson et al. [2008] show that the climate structure andozone in GEOS CCM agree quite well with observations.Additional evaluations of GEOS CCM [Eyring et al., 2006;Perlwitz et al., 2008; Oman et al., 2008; Waugh and Eyring,2008] reveal good comparisons with observations.[7] In this study we compare GEOS CCM simulations

with identical greenhouse gas (GHG), ODSs, and SSTs butdifferent ozone fields in the radiation scheme. In the ‘‘con-trol’’ (CTL) simulations the O3 field is three-dimensionaland determined interactively within the CCM, whereas in the‘‘zonal meanO3’’ (ZM) simulations themonthly-mean zonal-


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

2Atmospheric Chemistry and Dynamics Branch, NASA Goddard SpaceFlight Center, Greenbelt, Maryland, USA.

3Global Modeling and Assimilation Office, NASA Goddard SpaceFlight Center, Greenbelt, Maryland, USA.

4Cooperative Institute for Research in Environmental Sciences,University of Colorado, NOAA, Boulder, Colorado, USA.

Copyright 2009 by the American Geophysical Union.0094-8276/09/2009GL040419

L18701 1 of 6

Sensitivity of Southern Hemisphere climate to zonal asymmetry

in ozone

Julia A. Crook,1 Nathan P. Gillett,1 and Sarah P. E. Keeley1,2

Received 16 November 2007; revised 14 February 2008; accepted 27 February 2008; published 3 April 2008.

[1] Climate model simulations of past and future climateinvariably contain prescribed zonal mean stratosphericozone. While the effects of zonal asymmetry in ozonehave been examined in the Northern Hemisphere, muchgreater zonal asymmetry occurs in the Southern Hemisphereduring the break up of the Antarctic ozone hole. Weprescribe a realistic three-dimensional distribution of ozonein a high vertical resolution atmospheric model andcompare results with a simulation containing zonal meanozone. Prescribing the three dimensional ozone distributionresults in a cooling of the stratosphere and uppertroposphere comparable to that caused by ozone depletionitself. Our results suggest that changes in the zonalasymmetry of ozone have had important impacts onSouthern Hemisphere climate, and will continue to do soin the future. Citation: Crook, J. A., N. P. Gillett, and S. P. E.Keeley (2008), Sensitivity of Southern Hemisphere climate tozonal asymmetry in ozone, Geophys. Res. Lett., 35, L07806,doi:10.1029/2007GL032698.

1. Introduction

[2] The seasonal variation in ozone concentration overthe Antarctic is large, with the greatest depletion occurringin the austral spring when the sun returns to Antarctica[Solomon et al., 2005]. The resulting ozone hole is generallynot centered over the pole, but is displaced towards theAtlantic sector [Grytsai et al., 2007]. Observations[Ramaswamy et al., 2001; Randel and Wu, 1999a; Thompsonand Solomon, 2002; Thompson et al., 2005] and climatemodel simulations of the response to stratospheric ozonedepletion [Baldwin and Dameris, 2007; Forster and Shine,1997; Gillett and Thompson, 2003; Gillett et al., 2003;Hegerl et al., 2007;Keeley et al., 2007; Polvani and Kushner,2002; Ramaswamy et al., 2001; Shindell and Schmidt, 2004;Sexton, 2001; van Lipzig et al., 2006] indicate that the ozonehole has played a dominant role in forcing stratosphericcooling trends in the Antarctic stratosphere, and that it hasalso strengthened the westerly winds over the SouthernOcean, cooled the Antarctic interior, and warmed the Ant-arctic Peninsula, particularly during spring and summer.However, results of these simulations cannot be entirelyrealistic because zonal mean ozone concentrations are inva-riably specified in models, both for climate simulations[Gillett and Thompson, 2003; Hegerl et al., 2007; Randallet al., 2007], and for simulations of stratospheric temperature

trends [Ramaswamy et al., 2001]. Studies with mechanisticmodels indicate that zonal asymmetries in ozone may beimportant in modulating wave driving of the stratosphere[Nathan and Cordero, 2007]. Modeling studies of the North-ern Hemisphere suggest that changes in the zonal asymmetryof ozone may have caused significant 30-year stratospherictemperature trends there and altered the circulation over theNorth Atlantic and European region [Gabriel et al., 2007;Kirchner and Peters, 2003], but departures from zonalsymmetry of the ozone distribution are much larger in theSouthern Hemisphere during the break up of the ozone hole.Based on data from the European Centre for Medium RangeWeather Forecasting 40-yr Reanalysis (ERA-40) [Uppala etal., 2005], the monthly mean amplitude over the 1990s ofzonal wave number 1 in ozone mass mixing ratio is muchlarger in the Southern Hemisphere than the Northern Hemi-sphere from August through November at 50 hPa. TheNorthern Hemisphere amplitude peaks in February at50 hPa, with 0.9 mg kg!1 at 65!N, whereas the SouthernHemisphere amplitude peaks in October at 50 hPa, with1.7 mg kg!1 at 65!S.

2. Data and Methods

[3] Here we study the effects of zonally asymmetricozone on the Southern Hemisphere climate using a highvertical resolution version of the Hadley Centre slab model,denoted HadSM3-L64. This consists of a 50-m-deep mixed-layer ocean and a dynamic and thermodynamic sea-icemodel coupled to a 64-level atmospheric model extendingup to 0.01 hPa [Gillett et al., 2003].[4] In the absence of a suitable three-dimensional gridded

observational ozone dataset, we used ozone concentrationsfrom ERA-40. ERA-40 provides global gridded ozone dataof adequate quality for a sensitivity study such as this andhas been used for a similar study of the Northern Hemi-sphere [Gabriel et al., 2007]. Ozone is assimilated fromTOMS and SBUV data when available and the reanalysismodel contains a chemistry parameterization and tracertransport equation [Dethof and Holm, 2004]. Despite somedeficiencies, the ERA-40 ozone describes well the largescale structures of stratospheric ozone such as the formationand break up of the ozone hole [Dethof and Holm, 2004].[5] Annually repeating ozone was prescribed in the

model from the 12 month period of July 2000 to June2001. This period was chosen because October 2000 showsparticularly large zonal asymmetry in its ozone distribution(Figure 1a). The greatest zonal asymmetry is concentratedin September–November from 100 hPa to 10 hPa. In othermonths the zonal asymmetry is minimal. The ozone distri-bution in July 2000 and June 2001 was very similar,therefore wrapping the ozone data at this point is sensible.


1Climatic Research Unit, School of Environmental Sciences, Universityof East Anglia, Norwich, UK.

2Department of Meteorology, University of Reading, Reading, UK.


L07806 1 of 5

Effect of zonal asymmetries in stratospheric ozone on simulated

Southern Hemisphere climate trends

D. W. Waugh,1 L. Oman,1 P. A. Newman,2 R. S. Stolarski,2 S. Pawson,3 J. E. Nielsen,3

and J. Perlwitz4

Received 6 August 2009; revised 21 August 2009; accepted 27 August 2009; published 22 September 2009.

[1] Stratospheric ozone is represented in most climatemodels by prescribing zonal-mean fields. We examine theimpact of this on Southern Hemisphere (SH) trends using achemistry climate model (CCM): multi-decadal simulationswith interactive stratospheric chemistry are compared withparallel simulations using the samemodel in which the zonal-mean ozone is prescribed. Prescribing zonal-mean ozoneresults in a warmer Antarctic stratosphere when there is alarge ozone hole, with much smaller differences at othertimes. As a consequence, Antarctic temperature trendsfor 1960 to 2000 and 2000 to 2050 in the CCM areunderestimated when zonal-mean ozone is prescribed. Theimpacts of stratospheric changes on the troposphericcirculation (i.e., summertime trends in the SH annularmode) are also underestimated. This shows that SH trendsrelated to ozone depletion and recovery are underestimatedwhen interactions between stratospheric ozone and climateare approximated by an imposed zonal-mean ozone field.Citation: Waugh, D.W., L. Oman, P. A. Newman, R. S. Stolarski,S. Pawson, J. E. Nielsen, and J. Perlwitz (2009), Effect of zonalasymmetries in stratospheric ozone on simulated SouthernHemisphere climate trends, Geophys. Res. Lett., 36, L18701,doi:10.1029/2009GL040419.

1. Introduction

[2] It is now well established that the ozone hole hasplayed a major role in changes in the summer troposphericcirculation of the southern hemisphere (SH) over the last twodecades, and that the expected recovery of Antarctic ozonewill likely also be a major factor in SH climate change overthe next fifty years [e.g., Thompson and Solomon, 2002;Marshall, 2003; Gillett and Thompson, 2003; Perlwitz et al.,2008; Son et al., 2008, 2009]. As a result it is importantto include the impact of ozone depletion and recovery insimulations (predictions) of changes in SH climate.[3] However, Perlwitz et al. [2008] and Son et al. [2008]

suggest that the impact of changes in stratospheric ozone onthe tropospheric climate may not be fully captured in theWorld Climate Research Programme’s Coupled Model Inter-comparison Project phase 3 (CMIP3) multi-model dataset

used in the Fourth Assessment of the IntergovernmentalPanel on Climate Change [Intergovernmental Panel onClimate Change, 2007]. They showed that, even in theCMIP3 models that prescribed ozone recovery, the tropo-spheric response is weaker than that in the coupled chemistrymodels (CCMs) in the SPARC Chemistry-Climate ModelValidation Activity (CCMVal), which calculate stratosphericozone interactively.[4] There are several possible reasons for the difference in

the tropospheric response in CMIP3 and CCMVal models,including lack of interactive chemistry, incorrect specifica-tion of ozone, inadequate representation of the stratosphericcirculation in the CMIP3 models, or the lack of a dynamicocean in the CCMVal models. Here we focus on the impor-tance of interactive stratospheric chemistry and the impact ofprescribing monthly-mean zonal-mean ozone (as is done inthe CMIP3 models). Sassi et al. [2005], Crook et al. [2008],and Gillett et al. [2009] have shown that the Antarctic vortexis weaker and warmer in simulations without zonal asymme-tries in O3. This suggests that the use of prescribed zonal-mean ozone in the CMIP3 models may be the cause of thedifference from CCMVal models. However, the above stud-ies considered only conditions with high levels of ozonedepleting substances (ODSs) and a large ozone hole, andthey did not examine the impact on long-term trends in thestratosphere or troposphere.[5] In this study we examine the impact of zonal asymme-

tries in ozone on simulated trends by comparing simulationsfor 1955 to 2055 from the NASA’s Goddard Earth ObservingSystem Chemistry-Climate Model (GEOS CCM) [Pawsonet al., 2008], which has full, interactive stratospheric chem-istry, with parallel simulations using the same CCM exceptthat the monthly-mean zonal-mean stratospheric ozone fromthe first simulation is prescribed.

2. Model and Simulations

[6] The GEOS CCM includes representations of atmo-spheric dynamics, radiation, and stratospheric chemistry andtheir coupling through transport and radiative processes.Pawson et al. [2008] show that the climate structure andozone in GEOS CCM agree quite well with observations.Additional evaluations of GEOS CCM [Eyring et al., 2006;Perlwitz et al., 2008; Oman et al., 2008; Waugh and Eyring,2008] reveal good comparisons with observations.[7] In this study we compare GEOS CCM simulations

with identical greenhouse gas (GHG), ODSs, and SSTs butdifferent ozone fields in the radiation scheme. In the ‘‘con-trol’’ (CTL) simulations the O3 field is three-dimensionaland determined interactively within the CCM, whereas in the‘‘zonal meanO3’’ (ZM) simulations themonthly-mean zonal-


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

2Atmospheric Chemistry and Dynamics Branch, NASA Goddard SpaceFlight Center, Greenbelt, Maryland, USA.

3Global Modeling and Assimilation Office, NASA Goddard SpaceFlight Center, Greenbelt, Maryland, USA.

4Cooperative Institute for Research in Environmental Sciences,University of Colorado, NOAA, Boulder, Colorado, USA.


L18701 1 of 6

Sensitivity of Southern Hemisphere climate to zonal asymmetry

in ozone

Julia A. Crook,1 Nathan P. Gillett,1 and Sarah P. E. Keeley1,2

Received 16 November 2007; revised 14 February 2008; accepted 27 February 2008; published 3 April 2008.

[1] Climate model simulations of past and future climateinvariably contain prescribed zonal mean stratosphericozone. While the effects of zonal asymmetry in ozonehave been examined in the Northern Hemisphere, muchgreater zonal asymmetry occurs in the Southern Hemisphereduring the break up of the Antarctic ozone hole. Weprescribe a realistic three-dimensional distribution of ozonein a high vertical resolution atmospheric model andcompare results with a simulation containing zonal meanozone. Prescribing the three dimensional ozone distributionresults in a cooling of the stratosphere and uppertroposphere comparable to that caused by ozone depletionitself. Our results suggest that changes in the zonalasymmetry of ozone have had important impacts onSouthern Hemisphere climate, and will continue to do soin the future. Citation: Crook, J. A., N. P. Gillett, and S. P. E.Keeley (2008), Sensitivity of Southern Hemisphere climate tozonal asymmetry in ozone, Geophys. Res. Lett., 35, L07806,doi:10.1029/2007GL032698.

1. Introduction

[2] The seasonal variation in ozone concentration overthe Antarctic is large, with the greatest depletion occurringin the austral spring when the sun returns to Antarctica[Solomon et al., 2005]. The resulting ozone hole is generallynot centered over the pole, but is displaced towards theAtlantic sector [Grytsai et al., 2007]. Observations[Ramaswamy et al., 2001; Randel and Wu, 1999a; Thompsonand Solomon, 2002; Thompson et al., 2005] and climatemodel simulations of the response to stratospheric ozonedepletion [Baldwin and Dameris, 2007; Forster and Shine,1997; Gillett and Thompson, 2003; Gillett et al., 2003;Hegerl et al., 2007;Keeley et al., 2007; Polvani and Kushner,2002; Ramaswamy et al., 2001; Shindell and Schmidt, 2004;Sexton, 2001; van Lipzig et al., 2006] indicate that the ozonehole has played a dominant role in forcing stratosphericcooling trends in the Antarctic stratosphere, and that it hasalso strengthened the westerly winds over the SouthernOcean, cooled the Antarctic interior, and warmed the Ant-arctic Peninsula, particularly during spring and summer.However, results of these simulations cannot be entirelyrealistic because zonal mean ozone concentrations are inva-riably specified in models, both for climate simulations[Gillett and Thompson, 2003; Hegerl et al., 2007; Randallet al., 2007], and for simulations of stratospheric temperature

trends [Ramaswamy et al., 2001]. Studies with mechanisticmodels indicate that zonal asymmetries in ozone may beimportant in modulating wave driving of the stratosphere[Nathan and Cordero, 2007]. Modeling studies of the North-ern Hemisphere suggest that changes in the zonal asymmetryof ozone may have caused significant 30-year stratospherictemperature trends there and altered the circulation over theNorth Atlantic and European region [Gabriel et al., 2007;Kirchner and Peters, 2003], but departures from zonalsymmetry of the ozone distribution are much larger in theSouthern Hemisphere during the break up of the ozone hole.Based on data from the European Centre for Medium RangeWeather Forecasting 40-yr Reanalysis (ERA-40) [Uppala etal., 2005], the monthly mean amplitude over the 1990s ofzonal wave number 1 in ozone mass mixing ratio is muchlarger in the Southern Hemisphere than the Northern Hemi-sphere from August through November at 50 hPa. TheNorthern Hemisphere amplitude peaks in February at50 hPa, with 0.9 mg kg!1 at 65!N, whereas the SouthernHemisphere amplitude peaks in October at 50 hPa, with1.7 mg kg!1 at 65!S.

2. Data and Methods

[3] Here we study the effects of zonally asymmetricozone on the Southern Hemisphere climate using a highvertical resolution version of the Hadley Centre slab model,denoted HadSM3-L64. This consists of a 50-m-deep mixed-layer ocean and a dynamic and thermodynamic sea-icemodel coupled to a 64-level atmospheric model extendingup to 0.01 hPa [Gillett et al., 2003].[4] In the absence of a suitable three-dimensional gridded

observational ozone dataset, we used ozone concentrationsfrom ERA-40. ERA-40 provides global gridded ozone dataof adequate quality for a sensitivity study such as this andhas been used for a similar study of the Northern Hemi-sphere [Gabriel et al., 2007]. Ozone is assimilated fromTOMS and SBUV data when available and the reanalysismodel contains a chemistry parameterization and tracertransport equation [Dethof and Holm, 2004]. Despite somedeficiencies, the ERA-40 ozone describes well the largescale structures of stratospheric ozone such as the formationand break up of the ozone hole [Dethof and Holm, 2004].[5] Annually repeating ozone was prescribed in the

model from the 12 month period of July 2000 to June2001. This period was chosen because October 2000 showsparticularly large zonal asymmetry in its ozone distribution(Figure 1a). The greatest zonal asymmetry is concentratedin September–November from 100 hPa to 10 hPa. In othermonths the zonal asymmetry is minimal. The ozone distri-bution in July 2000 and June 2001 was very similar,therefore wrapping the ozone data at this point is sensible.


1Climatic Research Unit, School of Environmental Sciences, Universityof East Anglia, Norwich, UK.

2Department of Meteorology, University of Reading, Reading, UK.


L07806 1 of 5

Sensitivity of climate to dynamically-consistent zonal asymmetries

in ozone

N. P. Gillett,1 J. F. Scinocca,1 D. A. Plummer,1 and M. C. Reader1

Received 9 January 2009; revised 17 March 2009; accepted 8 April 2009; published 22 May 2009.

[1] Previous investigations into the effect of zonalasymmetries in ozone on climate have compared simulationswith prescribed 3-D ozone, in which the ozone is notnecessarily consistent with the model dynamics, tosimulations with prescribed zonal mean ozone. We assessthe impact of zonal asymmetries in ozone by comparing acontrol simulation of a coupled chemistry version of theCanadian Middle Atmosphere Model (CMAM) in which theozone andmodel dynamics are consistent, with a simulation inwhich only the zonal mean of the ozone is passed to theradiative transfer scheme. These simulations reveal a robuststratospheric zonal-mean temperature and geopotential heightresponse to zonal asymmetries in ozone that is consistent withthat identified in previous studies and of a magnitudecomparable to observed trends. These results suggest that theinclusion of zonal asymmetries in ozone may be essential forthe accurate simulation of future stratospheric temperaturetrends. Citation: Gillett, N. P., J. F. Scinocca, D. A. Plummer, andM. C. Reader (2009), Sensitivity of climate to dynamically-consistent zonal asymmetries in ozone, Geophys. Res. Lett., 36,L10809, doi:10.1029/2009GL037246.

1. Introduction

[2] To date almost all coupled atmosphere-ocean climatemodels [e.g., Meehl et al., 2007], and many model simu-lations of the middle atmosphere [e.g., Ramaswamy et al.2001] have used specified zonal average ozone distribu-tions. Son et al. [2008] demonstrated that the SouthernHemisphere tropospheric circulation response to ozonerecovery is larger in a set of coupled chemistry models thanin a set of climate models in which zonal mean ozone isspecified. One possible reason for this difference is that thecoupled chemistry models included zonally asymmetricchanges in the ozone distribution. Several recent studieshave highlighted the importance of zonal asymmetries inozone for the simulation of stratospheric and troposphericconditions in the Northern Hemisphere [Kirchner andPeters, 2003; Gabriel et al., 2007], and in the SouthernHemisphere [Crook et al., 2008]. Zonal asymmetries inozone are particularly large in the Southern Hemisphereduring the breakup of the vortex, when the region ofmaximum ozone depletion is often displaced from the pole.Crook et al. [2008] demonstrated that accounting for zonalasymmetries in ozone in a simulation using observed ozonefrom a year with particularly strong zonal asymmetryresulted in stratospheric cooling comparable to that due toozone depletion itself. However, these studies all specified

fixed three-dimensional distributions of ozone: In suchsimulations the position of the ozone minimum is notconstrained by the dynamics, and may not be collocatedwith the dynamical polar vortex. Further, ozone-dynamicsfeedbacks [Nathan and Cordero, 2007] are not resolved,since the ozone distribution is specified.[3] Most studies on the influence of the zonal asymmetry of

ozone have focused on the stratosphere. However, large zonalasymmetries in ozone are found in the mesosphere and thermo-sphere associated with the diurnal cycle. Since ozone concen-trations are much higher at night than in the day at these levels,using a zonal mean of ozone has the potential to introduce a biasin the radiative heating rates at these levels. For this reasonPaul et al. [1998] restrict their ozone climatology to levels below0.3 hPa. However the widely-used [Li and Shine, 1995] zonal-mean ozone climatology extends to 0.0011 hPa, but usesdaytime ozone values from near-infrared airglowmeasurementsfrom the Solar Mesosphere Explorer [Li and Shine, 1995].However, in some cases climate models are run with prescribedozone from coupled chemistry models, particularly for futuresimulations, and in these cases the use of zonal mean ozonecould introduce a bias in the radiative heating rates.[4] One way in which the influence of zonal asymmetries

in ozone may be examined in a more realistic context is bycomparing a simulation of a coupled chemistry model, with asecond simulation in which the zonal mean of the calculatedozone is prescribed. Reddmann et al. [1999] carried out sucha comparison using the 3-D Karlsruhe simulation model ofthe middle atmosphere (KASIMA), accounting for the diur-nal cycle in ozone above 50 km by prescribing the daytimemean ozone above this level in the second simulation, andwith conditions at the 10-km lower boundary prescribed fromEuropean Centre for Medium Range Weather Forecasting(ECMWF) analyses. Their simulations start in July, meaningthat they are not able to realistically simulate the Antarcticwinter vortex or Antarctic ozone depletion. They simulate asingle northern winter, and use trace gases representative ofapproximately 1992. They examine temperature differencesbetween the two simulations in December and find only smalldifferences between the two simulations, which leads them toconclude that ‘for undisturbed ozone conditions (no polarozone hole) a realistic ozone climatology is sufficient formodel simulations’. In this study, we also assess the role ofthree dimensional ozone variations in a fully-coupled con-text, but using 40-yr simulations with a realistic annual cycleand stratospheric chlorine levels representative of approxi-mately 1990.

2. Model and Experiments

[5] We use the Canadian Middle Atmosphere Model withcoupled chemistry (CMAM) [Scinocca et al., 2008; de


1Canadian Centre for Climate Modelling and Analysis, EnvironmentCanada, Victoria, British Columbia, Canada.

Published in 2009 by the American Geophysical Union.

L10809 1 of 5

Effect of Specifying Monthly Mean Ozone

Daily Mean

Monthly Mean

Percent !Difference

0

0.5

1

1.5

2

2.5

3

Ozon

e (pp

m)

Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar−30

−25

−20

−15

−10

−5

0

5

10

Month

Per

cent

Cha

nge (

%)

a)

b)52hPa, 90S-70S Mean over 10yrs

Red minus Black

20th Century Historical Simulations• WACCM: 6 (3 New) members from 1955 to 2005

(started from differing atmospheric ICs) • SC-WACCM: Uses ensemble mean values from prior WACCM runs

for prescribed values • 2 x1955 to 2005 ensembles:

1) Zonal Mean, Monthly Mean O3

2) Zonal Mean, Daily Mean O3

1850 (200 yrs) 1850-1955

1 Member Each

1 Member

1955-2005

3 Members EachDaily Mean O3

Monthly Mean O3

J J A S O N D J F M A M

0

0

0

0

15

15

0

15

−15

15

30

30

Month

Pres

sure

(hPa

)

b) Ozone: WACCM minus SC-WACCM(daily)

10

30

100

300

700 −300

−200

−100

0

100

200

300


−15

1

−15

0

0

0

15

−15

−15

−30

30

30

−45

0

−60

45

45

−75

60 60

15

15 75

90

30

30

105−15

−15

−120

45

60

−15075

−180

90

135

120

105

195

−225

−45

210

165−225

135

−30

−240

Month

Pres

sure

(hPa

)

a) Ozone: WACCM minus SC-WACCM(monthly)

10

30

100

300

700

ΔO3 (ppb)

ΔSWH (K/day)


0.01

0

0

0

0

0.020.04

−0.01

0.06

−0.03−0.05

0

0

0 0.01

−0.07

0.08

0.01

0.1

Month

Pres

sure

(hPa

)

10

30

100

300

700

−0.09

c) SW Heating Rate: WACCM minus SC-WACCM(monthly)


0

0.01

0

00

0

−0.01

−0.01

0.010

0

0−0.02

0

−0.01

0.02

Month

Pres

sure

(hPa

)

10

30

100

300

700

−0.1

−0.05

0

0.05

0.1

0.15

−0.15

d) SW Heating Rate: WACCM minus SC-WACCM(daily)

ΔT(K)

e) Temperature: WACCM minus SC-WACCM(monthly) f) Temperature: WACCM minus SC-WACCM(daily)

300


−1

0

0

0

0

1

0

−1

−2

0

1

0

−2

−2

Month

Pres

sure

(hPa

)

10

100

700 −4

−3

−2

−1

0

1

2

3

4

30

300


−1

−1

0

0 0

0

1

−2 0

−3

−3

−1

Month

Pres

sure

(hPa

)

10

100

700

30

Impact of a Monthly Mean Specified Ozone Hole

Ozone!

Monthly Daily

WACCM minus SC; 90S-70S Mean; Dots Cover Insignificant Areas


0

0

0

0

15

15

0

15

−15

15

30

30

Month

Pres

sure

(hPa

)


10

30

100

300

700 −300

−200

−100

0

100

200

300


−15

1

−15

0

0

0

15

−15

−15

−30

30

30

−45

0

−60

45

45

−75

60 60

15

15 75

90

30

30

105−15

−15

−120

45

60

−15075

−180

90

135

120

105

195

−225

−45

210

165−225

135

−30

−240

Month

Pres

sure

(hPa

)


10

30

100

300

700

ΔO3 (ppb)

ΔSWH (K/day)


0.01

0

0

0

0

0.020.04

−0.01

0.06

−0.03−0.05

0

0

0 0.01

−0.07

0.08

0.01

0.1

Month

Pres

sure

(hPa

)

10

30

100

300

700

−0.09



0

0.01

0

00

0

−0.01

−0.01

0.010

0

0−0.02

0

−0.01

0.02

Month

Pres

sure

(hPa

)

10

30

100

300

700

−0.1

−0.05

0

0.05

0.1

0.15

−0.15


ΔT(K)


300


−1

0

0

0

0

1

0

−1

−2

0

1

0

−2

−2

Month

Pres

sure

(hPa

)

10

100

700 −4

−3

−2

−1

0

1

2

3

4

30

300


−1

−1

0

0 0

0

1

−2 0

−3

−3

−1

Month

Pres

sure

(hPa

)

10

100

700

30


Ozone!

Monthly Daily

Short!Wave !

Heating!




0

0

0

0

15

15

0

15

−15

15

30

30

Month

Pres

sure

(hPa

)


10

30

100

300

700 −300

−200

−100

0

100

200

300


−15

1

−15

0

0

0

15

−15

−15

−30

30

30

−45

0

−60

45

45

−75

60 60

15

15 75

90

30

30

105−15

−15

−120

45

60

−15075

−180

90

135

120

105

195

−225

−45

210

165−225

135

−30

−240

Month

Pres

sure

(hPa

)


10

30

100

300

700

ΔO3 (ppb)

ΔSWH (K/day)


0.01

0

0

0

0

0.020.04

−0.01

0.06

−0.03−0.05

0

0

0 0.01

−0.07

0.08

0.01

0.1

Month

Pres

sure

(hPa

)

10

30

100

300

700

−0.09



0

0.01

0

00

0

−0.01

−0.01

0.010

0

0−0.02

0

−0.01

0.02

Month

Pres

sure

(hPa

)

10

30

100

300

700

−0.1

−0.05

0

0.05

0.1

0.15

−0.15


ΔT(K)


300


−1

0

0

0

0

1

0

−1

−2

0

1

0

−2

−2

Month

Pres

sure

(hPa

)

10

100

700 −4

−3

−2

−1

0

1

2

3

4

30

300


−1

−1

0

0 0

0

1

−2 0

−3

−3

−1

Month

Pres

sure

(hPa

)

10

100

700

30

Ozone!

Monthly Daily

Short!Wave !

Heating!

Temperature!



Ozone!

Monthly Daily

Short!Wave !

Heating!

Temperature!


0

0

0

0

15

15

0

15

−15

15

30

30

Month

Pres

sure

(hPa

)


10

30

100

300

700 −300

−200

−100

0

100

200

300


−15

1

−15

0

0

0

15

−15

−15

−30

30

30

−45

0

−60

45

45

−75

60 60

15

15 75

90

30

30

105−15

−15

−120

45

60

−15075

−180

90

135

120

105

195

−225

−45

210

165−225

135

−30

−240

Month

Pres

sure

(hPa

)


10

30

100

300

700

ΔO3 (ppb)

ΔSWH (K/day)


0.01

0

0

0

0

0.020.04

−0.01

0.06

−0.03−0.05

0

0

0 0.01

−0.07

0.08

0.01

0.1

Month

Pres

sure

(hPa

)

10

30

100

300

700

−0.09



0

0.01

0

00

0

−0.01

−0.01

0.010

0

0−0.02

0

−0.01

0.02

Month

Pres

sure

(hPa

)

10

30

100

300

700

−0.1

−0.05

0

0.05

0.1

0.15

−0.15


ΔT(K)


300


−1

0

0

0

0

1

0

−1

−2

0

1

0

−2

−2

Month

Pres

sure

(hPa

)

10

100

700 −4

−3

−2

−1

0

1

2

3

4

30

300


−1

−1

0

0 0

0

1

−2 0

−3

−3

−1

Month

Pres

sure

(hPa

)

10

100

700

30



Ozone!

Monthly Daily

Short!Wave !

Heating!

Temperature!


0

0

0

0

15

15

0

15

−15

15

30

30

Month

Pres

sure

(hPa

)


10

30

100

300

700 −300

−200

−100

0

100

200

300


−15

1

−15

0

0

0

15

−15

−15

−30

30

30

−45

0

−60

45

45

−75

60 60

15

15 75

90

30

30

105−15

−15

−120

45

60

−15075

−180

90

135

120

105

195

−225

−45

210

165−225

135

−30

−240

Month

Pres

sure

(hPa

)


10

30

100

300

700

ΔO3 (ppb)

ΔSWH (K/day)


0.01

0

0

0

0

0.020.04

−0.01

0.06

−0.03−0.05

0

0

0 0.01

−0.07

0.08

0.01

0.1

Month

Pres

sure

(hPa

)

10

30

100

300

700

−0.09



0

0.01

0

00

0

−0.01

−0.01

0.010

0

0−0.02

0

−0.01

0.02

Month

Pres

sure

(hPa

)

10

30

100

300

700

−0.1

−0.05

0

0.05

0.1

0.15

−0.15


ΔT(K)


300


−1

0

0

0

0

1

0

−1

−2

0

1

0

−2

−2

Month

Pres

sure

(hPa

)

10

100

700 −4

−3

−2

−1

0

1

2

3

4

30

300


−1

−1

0

0 0

0

1

−2 0

−3

−3

−1

Month

Pres

sure

(hPa

)

10

100

700

30


-3 -1.5

Changes in the DJF Zonal Mean Winds!(1995-2005 mean minus 1960-1969 mean)

c) SC-WACCM(daily)

b) SC-WACCM(monthly)

0

0

−5

5

−90 −80 −70 −60 −50 −40 −30

10

100

1000 −5

0

5

10

15

13−15

5

5

5

5

5

510

10

10

10

10

1015

15

15

15

15

15

20

20

20

20

20

2525

25

25

30

30

−10−5

a) WACCMDifference (m

/s)Pres

sure

(mb)

Latitude

Difference (m/s)Pr

essu

re (m

b)

Latitude

Difference (m/s)Pr

essu

re (m

b)

Latitude

00

5

5

15

10−15

−10

−5

−5

5

55

5 510

10

10

10

10

1015

15

15

15

20

20

20

20

20

2525

25

25

30

30

−90 −80 −70 −60 −50 −40 −30

10

100

1000 −5

0

5

10

15

0

12−15

−10

−5

−5

5

5

5

5

5

510

10

10

10

10

10

15

15

15

15

15

1520

20

20

20

20

2525

25

25

30

30

−90 −80 −70 −60 −50 −40 −30

10

100

1000 −5

0

5

10

15

Peak = 13 m/s

Peak = 10 m/s

Peak = 12 m/s

WACCM

SC-Monthly

SC-Daily

Impact on Surface Climate Trends

−90 −80 −70 −60 −50 −40 −30−0.5

0

0.5

1

−90 −80 −70 −60 −50 −40 −30 −10

−5

0

5

10

15a) DJF Zonal Mean Zonal Wind at 867hPa

WACCMSC-WACCM(monthly)SC-WACCM(daily)

Tren

d (m

/s/d

ecad

e) Climotology (m

/s)

Latitude

Climotology (m

m)

−90 −80 −70 −60 −50 −40 −30 0

50

100

150

200

250

−90 −80 −70 −60 −50 −40 −30−5

0

5

10

Latitude

Tren

d (m

m/d

ecad

e)b) DJF Zonal Mean Precipitation

WACCMSC-WACCM(monthly)SC-WACCM(daily)

DJF Zonal Mean!Precipitation

DJF Zonal Mean!Wind at 867hPa

Summary• SC-WACCM’s climatology in the troposphere and

stratosphere are indistinguishable from WACCM.

• 1/2 Cost Of WACCM (with Chemistry)

• Temporal smoothing of the specified ozone forcing file leads to significant changes in southern hemispheric trends from 1955 to 2005.

Back Up Slides and Extra Info

SC-WACCM (1850) Ozone


Month

Latit

ude

WACCM Total Column Ozone

J F M A M J J A S O N D−90

−60

−30

0

30

60

90

215

245

275

305

335

365

395

425

455

485

Figure 3. Climatological monthly and zonal mean WACCM pre-industrial total column ozone

in Dobson Units (DU).



Latitude

Pres

sure

(hPa

)

(a) WACCM QRS (ANN)

8

0.5

2

8

64

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Latitude

Pres

sure

(hPa

)

(b) �QRS (K day−1; ANN)

0.25

−1.75−1−0.25

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Latitude

Pres

sure

(hPa

)

(c) �QRS (%; ANN)

−55−15

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Figure 4. Annual and zonal mean (a) total short-wave heating rate (QRS) for WACCM in K

day�1, (b) � QRS TOT (WACCM minus SC-WACCM) in K day�1 and (c) � QRS in % . Red

(blue) contours are positive (negative) values. Contour interval is 2�2, 2�1, 1, 2, 22, ... Kday�1

in (a), 0.25 K day�1 in (b) and ...,-25, -15, -5, 5, 15, 25,... in (c). Gray shading indicates regions

that are significantly di↵erent at the 95% level.


65#km#

Annual Short-Wave Heating Rate Differences

WACCM


Latitude

Pres

sure

(hPa

)

(a) WACCM QRS (ANN)

8

0.5

2

8

64

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Latitude

Pres

sure

(hPa

)


0.25

−1.75−1−0.25

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Latitude

Pres

sure

(hPa

)

(c) �QRS (%; ANN)

−55−15

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000








WACCMWACCM minus

SC-WACCM!

Ozone has a diurnal cycle in WACCM but not SC-WACCM

Instantaneous)zonal)profile)of)ozone)(ppmv))for)a)day)in)January)at)the)equator,)at)60km,)

and)at)12)midnight)0E.)Solid)in)WACCM)ozone)and)dashed)is)SCHWACCM)ozone)(Sassi%and%Garcia,%2005).)

ozone%that%WACCM%sees%

ozone%that%%SC/WACCM%sees%


Latitude

Pres

sure

(hPa

)

(a) WACCM QRS (ANN)

8

0.5

2

8

64

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Latitude

Pres

sure

(hPa

)


0.25

−1.75−1−0.25

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Latitude

Pres

sure

(hPa

)

(c) �QRS (%; ANN)

−55−15

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000








WACCMWACCM minus

SC-WACCM!WACCM minus

SC-WACCM (%)!


Latitude

Pres

sure

(hPa

)

(a) WACCM QRS (DJF)

8

0.5

0.12

5

2

8

32

128

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Latitude

Pres

sure

(hPa

)

(b) �QRS (K day−1; DJF)

0.25

0.5

1.25

3.25

−2−1.5−0.75

−0.25

−2.5

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Latitude

Pres

sure

(hPa

)

(c) �QRS (%; DJF)

−65−5

−5

−90 −60 −30 0 30 60 90

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

Figure 5. As in Figure 2 but for December-January-February (DJF).


Interpolation of monthly QRS onto model time-step causes seasonal biases

Surface !climate


SAT

(K)

Latitude

(a) SAT (DJF)

−90 −60 −30 0 30 60 90200

220

240

260

280

300

320

SAT

(K)

Latitude

(b) SAT (JJA)

−90 −60 −30 0 30 60 90200

250

300

350

SLP

(hPa

)

Latitude

(c) SLP (DJF)

−90 −60 −30 0 30 60 90980

990

1000

1010

1020

1030

SLP

(hPa

)

Latitude

(d) SLP (JJA)

−90 −60 −30 0 30 60 90980

990

1000

1010

1020

1030PR

ECIP

(mm

day

−1)

Latitude

(e) PRECIP (DJF)

−90 −60 −30 0 30 60 900

2

4

6

8

PREC

IP (m

m d

ay−1

)

Latitude

(f) PRECIP (JJA)

−90 −60 −30 0 30 60 900

2

4

6

8

10



Figure 13. Zonal mean surface air temperature (a), (b), sea-level pressure (c), (d) and

precipitation (e), (f) for December-January-February (DJF) (a), (c), (e) and June-July-August

(JJA) (b), (d), (f). Black, red and blue curves are WACCM, SC-WACCM and CCSM4.


DJF$ JJA#

Surface Climate


SAT

(K)

Month

(a) Global Mean SAT

J F M A M J J A S O N D284

285

286

287

288

289

Stan

dard

Dev

iatio

n (K

)

Month

(d) Standard Deviation of Global Mean SAT

J F M A M J J A S O N D

0.5

1

1.5

2

SAT

(K)

Month

(b) NH Polar Cap SAT


250

260

270

280

290

Stan

dard

Dev

iatio

n (K

)

Month

(e) Standard Deviation of NH Polar Cap SAT


0.5

1

1.5

2

SAT

(K)

Month

(c) SH Polar Cap SAT


245

250

255

260

265

Stan

dard

Dev

iatio

n (K

)

Month

(f) Standard Deviation of SH Polar Cap SAT


0.5

1

1.5

2



Figure 12. Surface air temperature (SAT) in WACCM, SC-WACCM and CCSM4 as a function

of month. (a) Global mean, (b) Northern Hemisphere (NH) polar cap average and (c) Southern

Hemisphere polar cap average. Months when CCSM4 is not significantly di↵erent from WACCM

at the 95% level are highlighted with gray shading. (d), (e) and (f) as (a), (b) and (c) but for

the standard deviation of SAT.


Surface Climate


Model SAT (K) P (mmday�1) SLP (hPa) SIE(10

6km2)

Global WACCM 286.8 (0.2) 2.83 (0.03) 1011.3 (0.05) –

SC-WACCM 286.7 (0.2) 2.83 (0.03) 1011.4 (0.05) –

CCSM4 286.5 (0.2) 2.93 (0.02) 1011.2 (0.04) –

21

�-90

�N WACCM 281.1 (0.3) 2.00 (0.04) 1016.9 (0.4) 14.0 (0.6)

SC-WACCM 281.0 (0.3) 1.99 (0.05) 1016.9 (0.4) 14.0 (0.5)

CCSM4 281.0 (0.3) 2.13 (0.04) 1014.9 (0.5) 13.3 (0.5)

21

�S-21

�N WACCM 298.2 (0.4) 4.10 (0.08) 1010.6 (0.4) –

SC-WACCM 298.2 (0.5) 4.10 (0.08) 1010.7 (0.4) –

CCSM4 298.2 (0.4) 4.20 (0.07) 1011.9 (0.3) –

21

�-90

�S WACCM 279.9 (0.2) 2.25 (0.04) 1006.6 (0.4) 16.4 (0.9)

SC-WACCM 279.8 (0.2) 2.25 (0.04) 1006.5 (0.4) 16.5 (0.7)

CCSM4 279.0 (0.2) 2.33 (0.04) 1006.7 (0.4) 20.4 (1.1)

Table 1. WACCM, SC-WACCM and CCSM4 annual mean surface air temperature (SAT ),

precipitation (P ), sea-level pressure (SLP ), and sea ice extent (SIE) for preindustrial conditions.

Climatological means are calculated over 200 years for WACCM, 195 years for SC-WACCM and

501 years for CCSM4. The 2� uncertainties in the means are listed in parentheses.


Zonal Mean Temperature ComparisonX - 40 SMITH ET AL.: SPECIFIED-CHEMISTRY WACCM

Pres

sure

(hPa

)

Latitude

(a) DJF TZM (WACCM)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000

Pres

sure

(hPa

)

Latitude

(b) JJA TZM (WACCM)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000

Pres

sure

(hPa

)

Latitude

(c) DJF TZM (SC−WACCM)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000Pr

essu

re (h

Pa)

Latitude

(d) JJA TZM (SC−WACCM)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000

Pres

sure

(hPa

)

Latitude

(e) DJF TZM (CCSM4)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000

Pres

sure

(hPa

)

Latitude

(f) JJA TZM (CCSM4)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000 150

160

170

180

190

200

210

220

230

240

250

260

270

280

290

300

Figure 7. Zonal mean temperature for (a),(c),(e) December-January-February (DJF) and

(b),(d),(f) June-July-August (JJA). Panels (a) and (b) are for WACCM, (c) and (d) are for

SC-WACCM and (e) and (f) are for CCSM4. Shading interval is 10 K.


Zonal Mean Wind ComparisonSMITH ET AL.: SPECIFIED-CHEMISTRY WACCM X - 41

Pres

sure

(hPa

)

Latitude

(a) DJF UZM (WACCM)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000

Pres

sure

(hPa

)

Latitude

(b) JJA UZM (WACCM)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000

Pres

sure

(hPa

)

Latitude

(c) DJF UZM (SC−WACCM)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000Pr

essu

re (h

Pa)

Latitude

(d) JJA UZM (SC−WACCM)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000

Pres

sure

(hPa

)

Latitude

(e) DJF UZM (CCSM4)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000

Pres

sure

(hPa

)

Latitude

(f) JJA UZM (CCSM4)

−90 −60 −30 0 30 60 90

0.01

0.1

1

10

100

1000−100

−80

−60

−40

−20

0

20

40

60

80

100

Figure 8. Zonal mean zonal wind for (a),(c),(e) December-January-February (DJF) and

(b),(d),(f) June-July-August (JJA). Panels (a) and (b) are for WACCM, (c) and (d) are for

SC-WACCM and (e) and (f) are for CCSM4. Shading interval is 10 m s�1.


The residual circulation is also well represented in SC-WACCM

The residual circulation is well represented in SC-WACCM

The tropical water vapor tape recorder

Plots&show&the&devia.on&in&water&vapor&mixing&ra.o&(ppmv)&from&the&.me8mean&average&profile&averaged&over&10°N810°S.&

Sudden stratospheric warming (SSW) frequency

Sudden stratospheric warming (SSW) frequency

Winter'Frequencies:''' 'WACCM'' '0.5'SSWs'yr61'' ' ' ' ' ' ' 'SC6WACCM '0.4'SSWs'yr61'' ' ' ' ' ' ' 'CCSM4 ' '0.08'SSWs'yr61'

Polar vorticesU"at"60N,"10hPa" U"at"60S,"10hPa"

RMSE"of"U"at"60N,"10hPa" RMSE"of"U"at"60S,"10hPa"

NH" SH"

DJFM Heat Flux and 10hPa Temperatures


Temperature (K)

Prob

abili

ty D

ensi

ty

(b) DJFM 10 hPa Temperature

180 200 220 240 2600

0.01

0.02

0.03

0.04

0.05

Temperature (K)

Prob

abili

ty D

ensi

ty

(d) DJFM 10 hPa Temperature Anomalies

−30 −20 −10 0 10 20 30 400

0.02

0.04

0.06

0.08

Heat Flux (m K s−1)

Prob

abili

ty D

ensi

ty

(a) DJFM 10hPa Heat Flux

−200 −100 0 100 200 300 400 5000

0.002

0.004

0.006

0.008

0.01

0.012

Heat Flux (m K s−1)

Prob

abili

ty D

ensi

ty

(c) DJFM 10hPa Heat Flux Anomalies

−200 0 200 4000

0.002

0.004

0.006

0.008

0.01

0.012



Figure 14. Probability density distributions of 10 hPa December-January-Febraury-March

(DJFM) total (a) and anomalous (c) zonal mean eddy heat flux averaged from 45��75�N, total

(b) and anomalous (d) polar cap averaged temperature. Black, red and blue curves are WACCM,

SC-WACCM and CCSM4. Probability density distributions are computed using a kernel density

estimator, which performs a non-parametric, smoothed fit to the data.


NAO


NAO Index

Prob

abili

ty D

ensi

ty

DJFM NAO Index

−6 −4 −2 0 2 40

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45WACCMSC−WACCMCCSM4

Figure 16. Probability density distributions of the North Atlantic Oscillation (NAO) index.

Black, red and blue curves are WACCM, SC-WACCM and CCSM4. The NAO index is the time

series of the leading EOF of monthly sea-level pressure anomalies for the North Atlantic region,

20��90�N and 90�W-30�E. Probability density distributions are computed using a kernel density

estimator, which performs a non-parametric, smoothed fit to the data.


NINO 3.4


10 5 3 2 1.5Period (years)

0 0.01 0.02 0.03 0.04 0.05 0.06 0.070

50

100

150

Pow

er S

pect

ral D

ensi

ty

Frequency (cyc mo−1)


Nino 3.4 Index

Pro

babi

lity

Den

sity

DJF Nino 3.4 Index

−5 0 50

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45WACCMSC−WACCMCCSM4

Figure 17. (a) Power spectrum and (b) probability density distrubtions of the NINO 3.4 index

for WACCM (black), SC-WACCM (red) and CCSM4 (blue). Thin gray, pink and blue lines in

(a) show 50-year spectra for WACCM, SC-WACCM and CCSM4, respectively. The NINO 3.4

index is the time series of sea surface temperature anomaly averaged over the tropical Pacific

region, 5�S-5�N and 170�W-120�W.


0

0

0

0

0.25

0.25 0.75

0.5

−0.5

−0.25

1.25

−1

−3

−2

−1

0

1

2

3

0

0

0

0.25

−0.25

0.25

−0.75

0.5

−0.25

−1.25

−1.75

0.5

1

−2.25

1.5

−3

−2

−1

0

1

2

3

c) SC-WACCM(daily)

b) SC-WACCM(monthly)

0

0

−5

5

−90 −80 −70 −60 −50 −40 −30

10

100

1000 −5

0

5

10

15

13−15

5

5

5

5

5

510

10

10

10

10

1015

15

15

15

15

15

20

20

20

20

20

2525

25

25

30

30

−10−5

a) WACCMDifference (m

/s)Pres

sure

(mb)

Latitude

Change in Difference (m/s)

Pres

sure

(mb)

Latitude

Change in Difference (m/s)

Pres

sure

(mb)

Latitude

00

5

15

−15

−10

−5

−5

5

55

5 510

10

10

10

10

1015

15

15

15

20

20

20

20

20

2525

25

25

30

30

−90 −80 −70 −60 −50 −40 −30

10

100

1000

0

−15

−10

−5

−5

5

5

5

5

5

510

10

10

10

10

10

15

15

15

15

15

1520

20

20

20

20

2525

25

25

30

30

−90 −80 −70 −60 −50 −40 −30

10

100

1000

Change from WACCM

Difference❴Changes in the Zonal Mean Winds!

(1995-2005 mean minus 1960-1969 mean)

Jet Changes and Temperature

ΔT(K)

ΔT(K)

d) DJF Temperature: WACCM minus SC-WACCM(monthly) f) DJF Temperature: WACCM minus SC-WACCM(daily)

−2.4

−2.1 −1.8−1.5

−1.2 −0.9−0.6

−0.3

0

0

0

0.3

0.60.9

0.3

0

0.3

−0.3

0.6

Latitude

Pres

sure

(hPa

)

−90 −80 −70 −60 −50 −40 −30

10

30

100

300

7001000 −3

−2.5

−2

−1.5

−1

−0.5

0

0.5

1

−2.55

1.05

−1.8

−1.5

−0.9−0.6

−0.30

00

0

0.3

0.6

0.3

0

0.3

−0.3

0.3

Latitude

Pres

sure

(hPa

)

−90 −80 −70 −60 −50 −40 −30

10

30

100

300

7001000 −3

−2.5

−2

−1.5

−1

−0.5

0

0.5

1

−1.95

1.050.9

c) DJF Zonal Wind: SC-WACCM(daily)

Difference (m/s)Pr

essu

re (m

b)

Latitude0

12−15

−10

−5

−5

5

5

5

5

5

510

10

10

10

10

10

15

15

15

15

15

1520

20

20

20

20

2525

25

25

30

30

−90 −80 −70 −60 −50 −40 −30

10

100

1000 −5

0

5

10

15b) DJF Zonal Wind: SC-WACCM(monthly)

Pres

sure

(mb)

Latitude

00

5

5

15

10−15

−10

−5

−5

5

5

5

5 510

10

10

10

10

1015

15

15

15

20

20

20

20

20

2525

25

25

30

30

−90 −80 −70 −60 −50 −40 −30

10

100

1000 −5

0

5

10

15

0

0

−5

5

−90 −80 −70 −60 −50 −40 −30

10

100

1000 −5

0

5

10

15

13−15

5

5

5

5

5

510

10

10

10

10

1015

15

15

15

15

15

20

20

20

20

20

2525

25

25

30

30

−10−5

a) DJF Zonal Wind: WACCM

Difference (m/s)Pr

essu

re (m

b)

Latitude

Difference (m/s)

30

300

700

30

300

700

30

300

700

SC-WACCM · SC-WACCM Physics • Based on CESM1(WACCM) • Ozone and CO 2 speciﬁed from prior...

Documents

Transcript of SC-WACCM · SC-WACCM Physics • Based on CESM1(WACCM) • Ozone and CO 2 speciﬁed from prior...