NASA Technical Reports Server (NTRS) 20140011349: Interannual Variations of MLS Carbon Monoxide Induced by Solar Cycle

0.33 MB·English

by  NASA Technical Reports Server (NTRS)

JournalofAtmosphericandSolar-TerrestrialPhysics102(2013)99–104
ContentslistsavailableatSciVerseScienceDirect
Journal of Atmospheric and Solar-Terrestrial Physics
journal homepage: www.elsevier.com/locate/jastp
Interannual variations of MLS carbon monoxide induced by solar cycle
Jae N. Leea,n, Dong L. Wub, Alexander Ruzmaikinc
aJointCenterforEarthSystemsTechnology,UniversityofMaryland,BaltimoreCounty,Baltimore,MD,USA
bNASAGoddardSpaceFlightCenter,Greenbelt,MD,USA
cJetPropulsionLaboratory,CaliforniaInstituteofTechnology,Pasadena,CA,USA
a r t i c l e i n f o a b s t r a c t
Articlehistory: Morethaneightyears(2004–2012)ofcarbonmonoxide(CO)measurementsfromtheAuraMicrowave
Received18October2012 LimbSounder(MLS)areanalyzed.ThemesosphericCO,largelyproducedbythecarbondioxide(CO )
2
Receivedinrevisedform photolysis in the lower thermosphere, is sensitive to the solar irradiance variability. The long-term
29April2013 variationofobservedmesosphericMLSCOconcentrationsathighlatitudesislikelydrivenbythesolar-
Accepted17May2013
cycle modulated UV forcing. Despite of different CO abundances in the southern and northern
Availableonline25May2013
hemisphericwinter,thesolar-cycledependenceappearstobesimilar.Thissolarsignalisfurthercarried
Keywords: down to the lower altitudes by the dynamical descent in the winter polar vortex. Aura MLS CO is
Carbonmonoxide comparedwiththeSolarRadiationandClimateExperiment(SORCE)totalsolarirradiance(TSI)andalso
Totalsolarirradiance
withthespectralirradianceinthefarultraviolet(FUV)regionfromtheSORCESolar-StellarIrradiance
Middleatmosphere ComparisonExperiment(SOLSTICE).Significantpositivecorrelation(upto0.6)isfoundbetweenCOand
MicrowaveLimbSounder
FUV/TSI in a large part of the upper atmosphere. The distribution of this positive correlation in the
mesosphereisconsistentwiththeexpectationofCOchangesinducedbythesolarirradiancevariations.
&2013ElsevierLtd.Allrightsreserved.
1. Introduction referencestherein).Thephotolysisintensityvarieswiththesolar
irradiance,especiallywiththecontributionsfromLymanαinthe
A long photochemical lifetime and strong vertical and hori- uppermiddleatmosphere(mesosphereandlowerthermosphere)
zontal gradients make carbon monoxide (CO) a good tracer for and those from the Schumann–Runge bands in the mesosphere
studying dynamics of the middle atmosphere (Allen et al., 2000; andstratosphere.
Leeetal.,2011).AuraMicrowaveLimbSounder(MLS)COshowsa TheCO photodissociationrateevaluatedwithMLSandAtmo-
2
CO volume mixing ratio (VMR) exceeding 20ppmv (parts per spheric Chemistry Experiment-Fourier Transform Spectrometer
million by volume) in the winter polar upper mesosphere. Gen- (ACE-FTS) measurements show that it maximizes near 145nm
erallyspeaking,theCOmixingratiodecreaseswithheightinthe in the thermosphere (Minschwaner et al., 2010). The dominant
lower stratosphere, reaching a minimum around 30hPa, then contribution to photo-dissociation occurs at wavelengths greater
increases with height up to the thermosphere due to its high than 175nm where the O Schumann–Runge bands control the
2
altitude source. While some CO is produced by the oxidation of atmosphericopacity.Solarultravioletirradiancesweretakenfrom
methaneinthestratosphere,mostofthestratosphericandmeso- theSolarUltravioletSpectralIrradianceMonitor(SUSIM)measure-
sphericCOiscreatedbyphotolysisofcarbondioxide(CO )inthe mentintheirphoto-dissociationfrequencycalculation.
2
upper mesosphere and thermosphere and transported down to Laboratory measurements of the absorption cross sections
the lower mesosphere and stratosphere (Solomon et al., 1985; of CO have been made in the region 117 -200nm (Yoshino
2
Allenetal.,1999): et al., 1996; Parkinson et al., 2003). These measurements show
that the absorption cross sections of CO is maximum near 135–
CO þhv-COþO ð1Þ 2
2 145nm. CO photolysis is a function of solar UV spectral irradi-
2
ance, CO cross section, and atmospheric opacity, so that the
As seen from the CO absorption cross-section, the photo- 2
dissociation takes place m2 ost effectively at the solar Lyman α relevantwavelengthsdependonaltitude.
(121.6nm) and Schumann–Runge band (175nm–200nm) radia- AmajorchemicallossofCOabove∼100kmisduetoathree-
bodyrecombinationwithatomicoxygen:
tion,whichiseffectiveonlyathighaltitudes(Solomonetal.,1985;
Brasseur and Solomon, 2005; Minschwaner et al., 2010; and COþOþM-CO2þM ð2Þ
However,thephotochemicaltimeoftheabovereactionismuch
nCorrespondingauthor.Tel.:[phone];fax:[phone]. slower than transport at these altitudes. In the upper stratosphere
E-mailaddress:[email protected](J.N.Lee). and mesosphere, CO is also produced by methane oxidation (not
1364-6826/$-seefrontmatter&2013ElsevierLtd.Allrightsreserved.
http://dx.doi.org/10.1016/j.jastp.2013.05.012
100 J.N.Leeetal./JournalofAtmosphericandSolar-TerrestrialPhysics102(2013)99–104
shown)andrapidlydestroyedbytheOHoxidation,i.e., (UV)from115to320nmwitharesolutionof0.1nm,anabsolute
accuracyofbetterthan5%,andarelativeaccuracyof0.5%peryear
COþOH-CO þH ð3Þ
2 (McClintocketal.,2005a,2005b;Snowetal.,2005).TheSOLSTICE
This reaction occurs mostly during the day, since upper- measurements are made using a pair of identical spectrometers,
atmospheric OHisproducedfromphotolysisreactions (e.g., with SOLSTICE A and SOLSTICE B. Each spectrometer independently
H O). makes ultraviolet spectral measurements in two intervals: far
2
ChemicallossofthethermosphericCOisnegligibleduetolow ultraviolet (FUV) in 115–180nm and mid ultraviolet (MUV)
amountofOH.ItgivesCOalonglifetimetobetransporteddown in180–320nm.
to the mesosphere. In the polar night, mesospheric CO is con- Spectral measurement of solar radiation from SORCE Spectral
serveddue tolackof OH. Therefore,COVMR has a small diurnal IrradianceMonitor(SIM)showsthatsolarirradiancechangesover
variationandcanbeusedasatracerofatmosphericdynamicsin 11-year cycle are wavelength dependent (Harder et al., 2009).
bothdayandnight,particularlyinthewintertimepolardynamics. These findings are incorporated into the Sun-climate inter con-
The CO lifetime is long enough to maintain the anomalous nection studies (Haigh et al., 2010; Merkel et al., 2011; Li et al.,
distribution ofCOinsidethevortexthroughoutthedescent from 2012; Swartz et al., 2012) since different regions of the Earth's
the mesosphere to the lower stratosphere over 30 days in the atmosphereresponddifferentlytoeachpartofthesolarirradiance
polar stratosphere and mesosphere during winter (Minschwaner spectrum.
et al., 2010; Lee et al., 2011). It has a relatively shorter lifetime The Aura MLS data, which cover the solar minimum period
(about10days)atthelowandmid-latitudesinthemidtoupper betweenthecycles23and24,containusefulinsightsonthesolar-
stratosphere, which helps to remove CO quicklyonce it is mixed driven chemistry and dynamics of the middle atmosphere. The
outofthevortex. 27-day variations in MLS stratospheric ozone and temperature
Inadditiontothephotochemicalsources,ionizationbyenergetic were investigated in comparison with SORCE UV radiation by
particles, primarily by electrons with energies of ∼1–6MeV in the Ruzmaikin et al. (2007). In this study we focus on interannual
auroral regions, is another source of mesospheric CO, as suggested variationsofthemesosphericCOobservedbyMLSandanalyzethe
byMarsh(2011).Statisticallysignificanttracercompositionchanges data in comparisonwith TSI measured by SORCE TIM. We find a
with large solar photon events (SPEs) are shown with the Whole goodcorrelation betweentheMLSmesosphericCOandthesolar
Atmosphere Community Climate Model (WACCM3) and various irradiance variations, suggesting the long-term MLS CO variation
satellite instrument measurements (Jackman et al., 2008, 2009, inthemesosphereislikelytracingthe11-yearsolarcycle.
2011),butsignificantCOchangeswithSPEwerenotreportedyet.
The unique CO lifetime and photolysis lead to a large vertical
gradientwithVMRvaluesfromppmvinthemesospheretoppbv 2. DataandMethods
(parts per billion per volume) in the low stratosphere. The
descending air in the wintertime polar vortex brings CO further Weusetheversion2.2oftheMLSdataforthedailyCO,gridded
down to the lower stratosphere, producing strong hemispheric separatelyfordayandnightat29pressurelevelsfromtheupper
asymmetry in its concentration in the middle atmosphere troposphere (316hPa) to the mesopause (∼0.005hPa). The daily
(Solomonetal.,1985). COfieldsaremappedontoa41(latitude)(cid:2)81(longitude)gridfor
The seasonal climatologyof zonal mean COabundance shows daytime (ascending) and nighttime (descending) orbits. Because
maxima during the winter in the high-latitude region (Filipiak Aura MLS sampling does not cover the regions poleward of 821
et al., 2005; Jin et al., 2009), which is a manifestation of the latitude, we use the observations in latitude bins between 821N
summer-to-winter meridional circulation at these altitudes and and 821S during the winter months (December for Northern
the descending air motion near the winter pole. Dynamical Hemisphere(NH);JulyforSouthernHemisphere(SH).Thetypical
perturbations from planetary and gravity wave activities also single-profile precision of MLSV2.2 COvaries from0.02ppmvat
contributetolargeCOvariabilityattheextra-tropicsinthemiddle 100hPa,0.2ppmvat1hPa,to11ppmvat0.002hPa,withvertical
atmosphere (Allen et al., 1999). CO variances in middle atmo- resolutionof4,3,and9km,respectively(Pumphreyetal.,2007).
spheretropicsarelargelyaffectedbyQBOandsolarinfluencesare MLShasagoodverticalandhorizontalresolution(9km(cid:2)200km)
relativelyweakinthatregion. andsingleprofileprecisionof11ppmvintheuppermesosphere.
Understanding mesospheric CO variability requires accurate As shown in the zonal mean CO in section 3, there is a large
measurements of solar spectral irradiance in order to quantify meridional gradient in its volume mixing ratio. This gradient,
theCOphotolysisinputmodulatedbythesolarcycle.RecentSolar evidentinMLSradiancemeasurementsaswell,isnotaffectedby
Radiation and Climate Experiment (SORCE) observations of solar theuniversalinitialguessusedintheretrievalalgorithm.Random
irradiancefromthespace(Rottman,2005;Pilewskieetal.,2005) errors in MLS CO must be greatly reduced in the monthly mean
enableustoinvestigatetheresponseofthephotolysisdrivenCO, data. However, there is a significant day to day variability in the
whichcanbesensitivetothesolar11-yearand27-dayvariability, monthly zonal mean CO which is included in the standard
andtotheSPEs. deviationasshowninFig.1.
The SORCE Total Irradiance Monitor(TIM) observations reveal The TSI data from the TIM is used in this study as a proxy of
that the averaged total solar irradiance (TSI) is ∼1360.8W/m2 solarirradiancevariationoversolarcycle.TheSORCETIMTSIdata
during the 2008 solar minimum, ∼5W/m2 lower than the pre- have 3 times higher accuracy (0.035%) than prior TSI measure-
viously reported values (Kopp and Lean, 2011). The lower TSI mentsandwithalongtermstabilityof0.001%peryear(Koppand
values are due to the advanced design of the TIM in which the Lawrence, 2005; Kopp et al., 2005). The FUV data from SORCE
scattered light, a primary cause of the higher irradiance values SOLSTICEisalsousedtoshowthesolarcycleinCOconcentration.
measured by the earlier generation of solar radiometers, is pre- The SORCE SOLSTICE instrument has spectral coverage in FUV
cluded by the forward replacement of the defining aperture (115–180nm) and MUV (180–320nm), but we only used far
relative to the view-limiting aperture. TIM observations also ultraviolet (FUV) integrated over 120nm–160nm band to avoid
indicatethattheTSIintotheEarth'satmospherechangesbyabout the discontinuities in the FUV and MUV spectrum near 180nm
∼0.1%fromsolarminimumtosolarmaximum. region.
The SORCE Solar-Stellar Irradiance Comparison Experiment To study the CO dependence on the solar-cycle forcing, we
(SOLSTICE) measures the solar spectral irradiance in ultraviolet search for correlations between the MLS CO and SORCE TSI/FUV
J.N.Leeetal./JournalofAtmosphericandSolar-TerrestrialPhysics102(2013)99–104 101
the thermosphere and mesosphere (Andrews et al.,1987; Garcia
andBoville,1994).
3. Resultsanddiscussions
AuraMLSmonthlyCOVMRshows astronginterannualvaria-
tionduringtheperiodof2004–2012thatcorrelatesinphasewith
theTSIdata(Fig.1).IntheNH,theCOat821N,0.005hPavariesby
about16%,showinglessthan22ppmvinthe2006–2009assolar
irradiance reaches the minimum, and more than 26ppmvas the
TSIincreasestowardssolarmaximumafter2009.Similarly,inthe
SH,COchangessignificantlywithsolarcycle.Thesolarirradiance
variation influence on the photolysis and transport from the low
thermosphere likely plays the key role in the interannual varia-
Fig.1. MonthlymeanMLSCOVMRforDecemberat821N(black)andJulyat821S bilityofwintertimeCOdistribution.
(green)at0.005hPa.RedcurveisthemonthlySORCETSIwithstandarddeviation. The observed CO sensitivity to TSI during this period of
Dotted red curve is the same TSI values but shifted down by 1.5W/m2 for
observationcanbeusedtoinfertheCOmodulationbythesolar-
illustratingthecorrelationwiththeCOchangeintheSH.InbothCOandTSI,the
standarddeviationiscalculatedfromdailyzonalmeanvalues. cycle in other periods of time, for example for one year since
October 1991 when ISAMS (Improved Stratospheric and Meso-
spheric Sounder) was operatingon board the Upper Atmosphere
using monthly averaged data during December 2004–July 2012. Research Satellite(UARS). TheISAMSinstrument providednearly
Wealsoperformthecorrelationstudywiththedailydataforthe global coverage with good vertical resolution to examine the CO
sameperiodtoanalyzethesolarsignalandothervariationsinCO distributionfrom10hPato0.03hPa.Becausethesolaractivitywas
atshorterperiods. Duetothestrongseasonaldependence of the atitsmaximumduring1990–1991,ISAMStookthemeasurement
mesospheric CO composition, we extract the solar-cycle signal slightlyafterthesolarmaximum(in1992).
separately for the boreal winter (December) and austral winter ThezonalmeanISAMSCOVMRduringthedayofJanuary1992
(July)seasons. was ∼3.2ppmv near 601N at 0.1hPa (∼65km) (López-Valverde
We limit the time of our analysis within the winter season etal.,1996).ThisamountisslightlyhigherthantheCOmeasured
whenthesolarsignalinCOissignificant.Thesesignificantsignals byAuraMLSatthesameheightandlatitudeduringDecemberof
last only a few months. The correlation is shown for December 2004and2011(∼1.8ppmvand2.4ppmv)andJanuaryof2005and
when the most statistical correlation is shown in the polar NH 2011 (∼1.4ppmv and 1.8ppmv), respectively, since Aura MLS
beforethemiddleatmosphericCOsignalisaffectedbyNHstrato- observationsdonotreachthesolarmaximumsofar.Solaractivity
spheric sudden warmings, which occurred in January and/or was at maximum in 2000 and 2001 with double peaks, and at
February of 2006, 2009, and 2010 in the mesosphere. Data for minimumin2009duringsolarcycle23.WiththesecaveatstheCO
onemonth(DecemberandJuly,respectively)fromCOandTSI/FUV value from the ISAMS present a reasonable agreement with the
observations over ∼8 years have enough samples to show the valuemeasuredbyMLS.
statistically significant correlations. The sensitivity of CO to the The large vertical gradient in the MLS CO VMR from meso-
FUV solar-cycle variability is estimated in terms of ppmv per % sphere to the low stratosphere (Fig. 2) makes it a good tracer to
change of FUV, by regressing the CO VMR on the unit percent studythedownwardtransportofCOinducedbythesolarforcings
changeoftheFUVirradiance. in the lower thermosphere. The chemical balance between the
Weevaluatethestatisticalsignificanceoftheregressedcoeffi- source and the sink leads to the strong vertical gradient of CO
cients using a Student's t statistic applied to the correlation mixing ratios with relatively abundant CO in the upper meso-
coefficientrwithnn–2degreesoffreedom: sphere,butwithextremelysmallVMRsinthelowerstratosphere.
rffiffiffiffiffiffiffiffiffiffiffi Horizontal CO gradients reflect the transport in the winter polar
n*−2
t¼r ð4Þ vortex. The descent in the polar region brings high CO from the
1−r2
mesosphere tothe lower stratosphere. The long COlifetime over
TheeffectivesamplesizennisestimatedusingtheQuenouille's 30 days in the polar stratosphere and mesosphere during winter
procedureasdescribedbyAngellandKorshover(1981). helpstomaintainthisverticalandhorizontalgradients.
The CO response to solar forcing is not homogeneous both The solar cycle signal extends down to the mesosphere
horizontally and vertically, especially to the 27-day forcing, and upper stratosphere, as revealed in the correlation coefficient
because the CO anomalies are transported from one region to between daily mean MLS CO VMR and daily SORCE TSI for
another region. To better understand these dynamically–coupled December(Fig.2(c))andJuly(Fig.2(d)).Thisextendedsolarsignal
atmospheric responses, we look for the lagged correlation isalsoconsistentwiththecorrelationpatterncalculatedwithdaily
betweenTSI and MLS CO at different altitudes during the boreal SORCESOLSTICEFUVirradiance(Fig.2(e)–(f)).Thisextendedsolar
winter when the descent is strongest in the polar middle and signalcomeslikelyfromtheindirectatmosphericresponsestothe
upperatmosphere. solar-cycle forcing, as a result of the dynamical transport. How-
ThethermosphericandmesosphericCOrespondstothe27-day ever,wedidnotincludeanytimelaginthecorrelationanalysisin
solar forcing with an enhanced VMR through the photolysis at Fig. 2(c)–(f). Therefore, the amplitude of these correlation coeffi-
frequenciesneartheLyman-α.Mostoftheradiationatthisportion cients may be underestimated. Because the vertical descent is
ofthesolarspectrumdoesnotpenetratedowntothemesosphere; often fast (Lee et al., 2011), the zero lag coefficients can still
therefore there is almost no direct response to it in the meso- capturethepositivecorrelationbetweenMLSCOandTSIandthey
spheric CO. However, the mesospheric CO may have the 27-day arestatisticallysignificantoverawideregionfromtheequatorto
signal resulted from the downward transportof the 27-day solar thewinterpole.
cyclemodulatedthermosphericCO.Thisindirectresponseoccurs Noticeabledifferencesareevidentinthedownwardextension
only in the winter polar regionwhere the air is descended from of the solar signal in the December and Julycorrelations. During
102 J.N.Leeetal./JournalofAtmosphericandSolar-TerrestrialPhysics102(2013)99–104
Fig.2. DailyaverageofzonalmeanMLSCOinLog10(ppmv)asafunctionoflatitude(a)fortheDecember,2004-2011(b)forJuly,2004–2012.Thecorrelationcoefficient
betweendailySORCETSIandzonalmeanMLSCOVMRcalculatedfor(c)Decemberand(d)July.ThecorrelationcoefficientbetweendailySORCESOLSTICEFUVandzonal
meanMLSCOVMRcalculatedfor(e)Decemberand(f)July.Solidblacklineboundsthecorrelationsthataredifferentfromzerowithin95%confidencelevels.Thesensitivity
ofCOtothe%changeofFUV(inppmvper%)estimatedbyregressingCOVMRonFUVfor(g)Decemberand(h)July.Overlaidlineboundstheregressioncoefficients
estimatedfromthecorrelations,whicharesignificantwithin95%confidencelevels.
J.N.Leeetal./JournalofAtmosphericandSolar-TerrestrialPhysics102(2013)99–104 103
Fig.3. Thetime-lagcorrelationbetweenMLSCOandSORCEFUVduringthemonthofDecemberof2009atdifferentaltitudes(from0.1hPato0.01hPa).Overlaidinblue
dashlineisthefirstprincipalcomponent(PC)oftheCONAM,whichaccountsfor∼30%ofthetotalCOvariancefromDecember2009toFebruary2010inthe201N–821N
domain.
the boreal winter (December, 2004–2011) (Fig. 2(c) and (e)), the downwardpropagationofthemaximumCONAMindex(Leeetal.,
dailycorrelationsbetweensolarirradianceandCOaresignificant 2011).
in the winter mesosphere where the CO concentration is high. This lag between CO and FUV is most noticeable during
Strongpositivecorrelationhasthelargestamplitudeinthehigher December of 2009 and 2010 in the Northern Hemisphere, while
mesosphere (between 0.01hPa and 0.005hPa) throughout the the event is not obvious in other years and not seen in the
equator to the high latitude of winter hemisphere. Those strong SouthernHemisphere.TheshorttermsolarsignalsinCOlastonly
signalstendtopropagatedowntotheupperstratosphere(below a few months, at most, since CO variability depends on both
1hPa), due to the wintertime polar descent air from the upper chemicalanddynamicalprocesses.Inthemesosphere,thephoto-
mesosphere when the strong vortex is forming. Similarly, the chemicallifetimeofCO(∼30days),iscomparabletotransporttime
correlationintheaustralwinter(July,2004–2012)showsastrong scales (Solomon et al., 1985) so that solar induced regional CO
positive correlation throughout the mesosphere, but the down- anomalies may disappear by transport. The short term solar
ward extension is weaker in the stratosphere (Fig. 2(d) and (f)). signals in CO are currently in investigation by identifying major
Thecorrelationismoresignificantmorethan0.6intheSHwhen patterns of latitude-height variability of CO in the middle atmo-
theFUVirradianceisusedtocomputethecorrelation. sphereanddetailswillbediscussedinaseparatework(Ruzmaikin
No significant negative correlation between TSI/FUV and CO etal.2007),tobesubmittedinAdvancesinSpaceResearch].
is found in this analysis, suggesting that an increase in the solar To further examine the relation between the solar signal and
irradiance would cause CO increases in the upper mesosphere the middle atmospheric CO anomaly patterns, we computed the
and thermosphere. The solar signals are brought downward by principal component (PC) indexof the first CONorthernAnnular
dynamicsathighlatitudes,causinganomalousCOvariationsinthe Mode (NAM) during December 2009–February 2010 and com-
lowermesosphereandstratosphere.ThesolarsignalinCOisweak pared it with the lagged correlation coefficients. The first EOF
inthetropicalandsummermesosphere,largelybecauseoflackof mode of geopotential height, called the NAM represents an
verticaltransportorhorizontalmixing. approximately axially symmetric annular structure of the polar
We estimate the sensitivity of CO to solar cycle variability by vortex.Similarly,theCONAM,definedasthefirstEOFofCOinthe
regressingtheCOVMRonSORCEFUV(120–160nm),andshowit middle atmosphere, is axially symmetric annular with a “dome-
in ppmv per % change of FUV. As shown in Fig. 2(g)–(h), the like” shape centered on the pole, reflecting high CO VMR at the
maximum CO response to 1% change of FUV can be as high as pole and decreasing equatorward. The positive indexof CO NAM
0.18ppmv in the upper mesosphere (∼80km). The sensitivity representsabundantCOmixingratiosinhighlatitudes,indicating
decreasesinthelowermesosphereandstratosphere.Considering strong descent within the strong vortex. The spatial pattern at
theFUVirradiancevarianceduringDecemberof2004–December different altitude and time–height development of these modes
of2009isover20%(∼2.8(cid:2)10−4to∼2.3(cid:2)10−4W/m2),thechange aredescribedinthepreviouswork(Leeetal.,2011).Asshownin
of winter time CO can be as high as ∼4ppmv during the same Fig. 3 (dashed line), the CO NAM index at 0.01hPa, derived
period.TheamplifiedCOresponsesindicatethatsubtlevariations independently from solar variability, has the 27-day variation
in solar irradiance can result in significant atmospheric changes and is in phase with the lag correlation between FUV and CO at
viachemistryanddynamics. 0.01hPa.
TheverticalpolardescentcausesadelayedCOresponsetothe
short-termsolarforcing,whichcanbeestimatedwiththelagged
correlationanalysis between FUV time series and the COVMRat 4. Conclusions
different altitudes. In Fig. 3, the lag between the CO and FUV at
56Nincreasesfrom3daysintheuppermesosphere(0.01hPa)to WeanalyzedMLSCOandSORCETSI/FUVdatatoexaminethe
14 days the upper mesosphere (0.1hPa). The maximum correla- solarinfluenceontheatmosphericCOduringtheperiodof2004
tion is decreased below 0.6 at 0.1hPa. These results suggest that to 2011. Statistically significant positive correlation between MLS
solarsignalsinCOpropagatesdownwardfrom0.01hPato0.1hPa CO and TSI/FUV is found in the upper mesosphere. The positive
within∼11days.Thisdescentrateofthesolarsignalcorresponds correlation is extended down to the upper stratosphere at the
to∼1.3km/dayandcompareswellwiththoseestimatedfromthe winter high latitudes where the strong polar vortex is present.
104 J.N.Leeetal./JournalofAtmosphericandSolar-TerrestrialPhysics102(2013)99–104
AuraMLSmonthlyaveragedCOover70NduringDecembervaries EOS microwave limb sounder on Aura: first results. Geophysical Research
by about 16% over the 11-yr solar cycle. Due to its long photo- Letters32,L14825,http://dx.doi.org/10.1029/2005GL022765.
Garcia, R.R., Boville, B.A., 1994. Downward control of the mean meridional
chemicallifetimeandstrongverticalgradientinthemesosphere,
circulationandthetemperatureof thepolarwinterstratosphere.Journalof
COactsasagooddynamicstracertostudyindirecteffectsofthe theAtmosphericSciences51,2238–2245.
upper-atmosphericsolarsignalandthetransportoftheseanoma- Harder,J.W.,Fontenla,J.M.,Pilewskie,P.,Richard,E.C.,Woods,T.N.,2009.Trendsin
solar spectral irradiance variability in the visible and infrared. Geophysical
liesdowntotheloweratmosphere.
ResearchLetters36,L07801,http://dx.doi.org/10.1029/2008GL036797.
The correlation between MLS CO and the SORCE TSI/FUV is Haigh, J.D., Winning, A.R., Toumi, R., Harder, J.W., 2010. An influence of solar
significant(ashighas0.6)inawideregionofupperatmospherein spectral variations on radiative forcing of climate. Nature 467, 696–699
bothhemispheres.TheobservedpositivecorrelationbetweenFUV http://dx.doi.org/10.1038/nature09426,Oct.
Jackman,C.H.,etal.,2008.Short-andmedium-termatmosphericconstituenteffectsof
andCOsuggeststhat theincreasedsolarirradiancecouldlead to verylargesolarprotonevents.AtmosphericChemistryandPhysics8,765–785.
anabundantCOVMRviatheCO photolysisthroughouttheupper Jackman,C.H.,etal.,2009.Long-termmiddleatmosphericinfluenceofverylarge
2
mesosphereduetothedeep-penetratingsolarUVspectralforcing. solarprotonevents.JournalofGeophysicalResearch114,D11304,http://dx.doi.
org/10.1029/2008JD011415.
TheenhancedCOcanalsobetransportedtothelowermesosphere Jackman,C.H.,etal.,2011.Northernhemisphereatmosphericinfluenceofthesolar
andupperstratospherebydynamicsthroughtheverticaldiabatic proton events and ground level enhancement in January 2005. Atmospheric
descentinthewinterpolarvortex,whichextendstheCO-TSIand ChemistryandPhysics11,6153–6166,http://dx.doi.org/10.5194/acp-11-6153-2011.
Jin, J.J., Semeniuk, K., Beagley, S.R., Fomichev, V.I., Jonsson, A.I., McConnell, J.C.,
CO-FUVcorrelationtoloweraltitudes.
Urban, J., Murtagh, D., Manney, G.L., Boone, C.D., Bernath, P.F., Walker, K.A.,
We also observed the CO responses to the solar 27-day Barret, B., Ricaud, P., Dupuy, E., 2009. Comparison of CMAM simulations of
variation in the polar mesosphere, which progress downward carbonmonoxide(CO),nitrousoxide(N2O),andmethane(CH4)withobserva-
tions from Odin/SMR, ACE-FTS, and Aura/MLS. Atmospheric Chemistry and
with a time lag from 3 days at 0.01hPa to 14 days at 0.1hPa. Physics9,3233–3252.
BecauseofthelongCOphotochemicallifetime,theCOsolarsignal Kopp, G., Lawrence, G., 2005. The Total Irradiance Monitor (TIM): instrument
can be transported by the polar descent to lower altitudes, design.SolarPhysics230,1–2,http://dx.doi.org/10.1007/s11207-005-7446-4.
Kopp,G.,Heuerman,K.,Lawrence,G.,2005.TheTotalIrradianceMonitor(TIM):
showing lagged responses in CO with height. The mean descent
instrument calibration.Solar Physics230,111–127, http://dx.doi.org/10.1007/
rate,estimatedfromthelaggedCO-TSIcorrelation,is∼1.3Km/day, s11207-005-7447-3.
whichisconsistentwiththevaluederivedbyLeeetal.(2011). Kopp,G.,Lean,J.L.,2011.Anew,lowervalueoftotalsolarirradiance:evidenceand
BecausethesolarUVspectrumiscriticaltotheCOphotolysis,it climatesignificance.GeophysicalResearchLetters38,L01706,http://dx.doi.org/
10.1029/2010GL045777.
isnotsurprisingtoseeahighercorrelationwithFUV-COthanTSI- Li,K.F.,Jiang,X.,Liang,M.-C.,Yung,Y.L.,2012.Simulationofsolar-cycleresponsein
CO.Observationsofspectralsolarirradiance(SSI)arevaluablefora tropicaltotalcolumnozoneusingSORCEirradiance.AtmosphericChemistry
more comprehensive understanding of atmospheric responses to andPhysics12,1–26,http://dx.doi.org/10.5194/acpd-12-1-2012.
Lee,J.N.,Wu,D.L.,Manney,G.L.,Schwartz,M.J.,Lambert,A.,Livesey,N.J.,Minschwaner,K.
solar spectral forcing. The SSI data from SORCE SOLSTICE will be
R.,Pumphrey,H.C.,Read,W.G.,2011.AuraMicrowaveLimbSounderobservationsof
utilized within the full wavelength range of measurement (120– thepolarmiddleatmosphere:dynamicsandtransportofCOandH2O.Journalof
320nm) if the uncertainty estimates is completed for each GeophysicalResearch116,D05110,http://dx.doi.org/10.1029/2010JD014608.
López-Valverde,M.A.,López-Puertas,M.,Remedios,J.J.,Rodgers,C.D.,Taylor,F.W.,
wavelength. Updated degradation trend at different wavelength
Zipf,E.C.,Erdman,P.W.,1996.Validationofmeasurementsofcarbonmonoxide
range in the latest SOLSTICE data products will be helpful to from the improved stratospheric and mesospheric sounder. Journal of Geo-
estimatemorereliableUVvariancewithsolarcycle.Furthermore, physicalResearch101(D6),9929–9955,http://dx.doi.org/10.1029/95JD01715.
Marsh, D., 2011. Chemical–Dynamical Coupling in the Mesosphere and Lower
a completesolarcyclewill be useful toestimatean amplitude of Thermosphere,AeronomyoftheEarth'sAtmopshereandIonosphere.Springer,
CO and other tracer changes over short term (∼27 day) and long Dordrecht,TheNetherlandspp.1–477.
term(11 year) solarcycle. Such a reliable long term SSI data is a McClintock,W.E.,Rottman,G.,Woods,T.N.,2005a.Solar-stellarirradiancecomparison
experimentII(SOLSTICEII):instrumentconceptanddesign.SolarPhysics230,225.
keyinunderstandingthesun-climateconnections,particularlyin
McClintock,W.E.,Snow,M.,Woods,T.N.,2005b.Solar-stellarirradiancecomparison
characterizingwhichspectralchangesoftheSun'sradiativeoutput experimentII(SOLSTICEII):pre-launchandon-orbitcalibrations.SolarPhysics
canplayaroleindifferentspeciesoftheEarth'satmosphere. 230,259.
Merkel, A.W., Harder, J.W., Marsh, D.R., Smith, A.K., Fontenla, J.M., Woods, T.N.,
2011.Theimpactofsolarspectralirradiancevariabilityonmiddleatmospheric
ozone. Geophysical Research Letters 38, L13802, http://dx.doi.org/10.1029/
Acknowledgments 2011GL0475612011.
Minschwaner, K., et al., 2010. The photochemistry of carbon monoxide in the
stratosphereandmesosphereevaluatedfromobservationsbytheMicrowave
ThisworkissupportedbytheNASALivingWithaStarTargeted
Limb Sounder on the Aura satellite. Journal of Geophysical Research 115,
ResearchandTechnologyProgram(NNH10ZDA001N-LWSTRT).We D13303,http://dx.doi.org/10.1029/2009JD012654.
thanktheAuraMLSandSORCEteamsforprovidingtheirdataand Parkinson, W.H., Rufus, J., Yoshino, K., 2003. Absolute absorption cross section
analysissupport.Wethanktwoanonymousreviewersforvaluable measurementsofCO2inthewavelengthregion163–20nmandthetempera-
turedependence.ChemicalPhysics290,251–256.
commentsandsuggestionsthatledtoanimprovedmanuscript.
Pilewskie,P.,Rottman,G.,Richard,E.,2005.Anoverviewofthedispositionofsolar
radiation in the lower atmosphere: connections to the SORCE snge. Solar
Physics230(1),55–69.
Pumphrey,H.C.,etal.,2007.Validationofmiddle-atmospherecarbonmonoxide
References retrievalsfromtheMicrowaveLimbSounderonAura.JournalofGeophysical
Research112,D24S38,http://dx.doi.org/10.1029/2007JD008723.
Rottman,G.,2005.TheSORCEMission.SolarPhysics230(1),7–25.
Allen, D.R., et al., 1999. Observations of middle atmosphere CO from the UARS Ruzmaikin,A.,Santee,M.L.,Schwartz,M.J.,Froidevaux,L.,Pickett,H.,2007.The27-day
ISAMSduringtheearlynorthernwinter1991/1992.JournaloftheAtmospheric variationsinstratosphericozoneandtemperature:newMLSdata.Geophysical
Sciences56,563–583. ResearchLetters34,L02819,http://dx.doi.org/10.1029/2006GL028419.
Allen,D.R.,Stanford,J.L.,Nakamura,N.,Lopez-Valverde,M.A.,Lopez-Puertas,M., Snow,M.,McClintock,W.E.,Rottman,G.,Woods,T.N.,2005.Solar-stellarirradiance
Taylor,F.W.,Remedios,J.J.,2000.Antarcticpolardescentandplanetarywave comparison experiment II (SOLSTICE II): examination of the solar stellar
activityobservedinISAMSCOfromApriltoJuly1992.GeophysicalResearch comparisontechnique.SolarPhysics230,295.
Letters27,665–668. Solomon,S.,etal.,1985.Photochemistryandtransportofcarbonmonoxideinthe
Angell,J.K.,Korshover,J.,1981.Comparisonbetweenseasurfacetemperatureinthe middleatmosphere.JournaloftheAtmosphericSciences42(1072-1083),1985.
equatorial eastern Pacific and United States surface temperature. Journal of Swartz,W.H.,Stolarski,R.S.,Oman,L.D.,Fleming,E.L.,Jackman,C.H.,2012.Middle
AppliedMeteorology20,1105–1110. atmosphere response to different descriptions of the 11-yr solar cycle in
Andrews, D.G., Holton, J.R., Leovy, C.B., 1987. Middle Atmospheric Dynamics. spectralirradianceinachemistry-climatemodel.AtmosphericChemistryand
AcademicPress,NewYorkpp.1–489. Physics12,5937–5948.
Brasseur,G.P., Solomon, S., 2005. Aeronomyof the Middle Atmosphere, 3rd ed. Yoshino, K., Esmond, J.R., Sun, Y., Parkinson, W.H., Ito, K., Matsui, T., 1996.
Springer,Dordrecht,TheNetherlandspp.1–646. Absorptioncrosssectionmeasurementsofcarbondioxideinthewavelength
Filipiak,M.J.,Harwood,R.S.,Jiang,J.H.,Li,Q.,Livesey,N.J.,Manney,G.L.,Read,W.G., region118.7–175.5nmandthetemperaturedependence.JournalofQuantita-
Schwartz,M.J.,Waters,J.W.,Wu,D.L.,2005.Carbonmonoxidemeasuredbythe tiveSpectroscopyandRadiativeTransfer55,53.