D</I> Region Winter Anomaly at High and Middle Latitudes Induced By

Radio Science,Volume 20, Number 4, Pages795-802, July-August 1985

The D regionwinter anomaly at highand middlelatitudes induced by planetarywaves

Kohji Kawahira

GeophysicalInstitute, Kyoto University,Japan

(ReceivedNovember 16, 1984; revisedFebruary 4, 1985; acceptedMarch 20, 1985.)

An observationalstudy of the D region winter anomaly at high and middle latitudeshas beenmade duringthe periodof a suddenstratospheric warming of the 1967/1968winter. From isoplethanalysis of the absorptionindex, fmin, systematiclarge-scale distributions of the absorptionare found at even higherlatitudes than 60ø geomagneticlatitude as well as at middlelatitudes. On the basisof evidence that geostrophicwinds induced by well-developedplanetary waves extend to the upper D region,it is found that when southward winds are dominant, the absorptionsbecome weaker at high latitude but strongerat middle latitude, indicatingthat abundant NO could be carried away from high-latitude sourceregions to middle-latitudesink regions.Thus large amplitudeplanetary wavesin the D region could inducethe winter anomaly through NO transportnot only at middle but also at high latitudes.

1. INTRODUCTION transport of NO by enhanced eddy diffusion. A The winter anomaly of radio wave absorption in recent numerical study by Jones and Avery [1984], the D region (60-85 km), which is directly due to on the other hand, demonstratesthe role of planetary remarkable enhancementin the electron density, has wave wind transport of NO from high latitudes in been an important unresolvedphenomenon in the producing the irregular components.There are some lower ionosphere [Offermann, 1979]. According to inconsistenciesin theoretical studies of irregular Arnold and Krankowsky[1977], necessaryconditions components. for the observed electron density increase are an An observationalstudy by Kawahira [1982] of the about tenfold increase of ionization sources as NO winter anomaly at middle latitudes at sudden warm- (andO2(xAg)) and a temperaturerise of a fewtens of ings has given observational evidences that a NO degreesKelvin due to a strongtemperature depen- increase by well-developed planetary wave wind dence of the conversion rate of molecular ion to clus- transports from the polar region, where NO is most ter ion. Thus it is important to identify what pro- abundant [Cravens and Stewart, 1978; Iwagami and cessescould bring about these conditionsin the Ogawa, 1980], could produce irregular components winter D region. of the winter anomaly. Recent observationalstudies With the use of a two-dimensionaldynamical and by Pakhomov et al. [1984] have also verified the chemicalmodel, Solomonet al. [1982] simulatedthe effect of large-scalewinds on the winter anomaly at middle latitudes. winter anomaly at middle and high latitudes, re- sultingfrom a NO increasedue to verticaltransport These observational studies are confined to south from the lower thermospheresource region by eddy of 60ø geomagneticlatitude, south of aurora region. diffusion and meridional circulation. Their results are However, well-developed planetary waves in the consistentwith the regularcomponent of the winter mesospherecould induce a NO change by wind anomaly [Schwentek,1971]. Furthermore, they transport effectsalso at high latitudes, as shown by a pointed out the importantcontribution of aurora numerical study by Roble and Gray [1979] in the eventsto the high-latitudeNO increaseassociated case of NO transport in the lower thermospheric with energeticelectron precipitations [Kondo and polar region. Indeed, recent LIMS satellite observa- Ogawa,1976]. Concerning the mechanismof the ir- tions of NO2 in winter high latitudesby Russelet al. regularcomponent of the winteranomaly, Offermann [1984] find the large-scale distribution at polar et al. [1982] stressedthe importantrole of vertical winter night latitude above 70-km height. This first finding verifies a planetary wave effect on the D Copyright 1985 by the American GeophysicalUnion. region chemical species,not only at middle but also Paper number 5S0256. at high latitudes. Although the high latitude winter 0048-6604/85/005S-0256508.00 anomaly has been studied mainly in relation to

795 796 KAWAHIRA' WINTER ANOMALY INDUCED BY PLANETARY WAVES

dencefor the mechanismof this winter anomaly as a NO increase due to southward transport from high latitudes by well-developedplanetary wave winds. A -4O ,/"•Z•••(A)TE MPERATURE schematic picture of the mechanism is shown in Figure 2 rKawahira, 1985]. These large-amplitude -8O • 35KM, FT. CHURCHILL planetary waves in the mesospherehave been identi- - ---o-- 50KM, W. GEIRINISH fied by satellite observations of temperatures [e.g., Hirota and Barnett, 1977]. Thus the figure shows a realistic pattern of geopotential height, especiallyits large variation along latitude circle, in the winter D region at the time of stratosphericsudden warmings. 10-2 Then strong southward geostrophic winds could transport abundant NO from high to middle lati- 10'3 tudes and here bring about an enhancementin NO concentration, resulting in the winter anomaly. On D-4 the other hand, northward winds could cause a de- A .., .... L, UNOAU crease at middle latitudes. As suggestedin Figure 2, ;(u•• •*•:'t.,tu,O•_or '150 •'. WANTE.D SIGNAL. / • ;: U•v[RSITY PARK / even at higher latitudes than 60ø geomagneticlati- ;; ,, ', •x a 75') J o tude, there would be a large-scalewind effect on NO variations in addition to geomagneticstorm effects. •00 In order to elucidate the effect of planetary wave winds on the absorption, the isopleth analysis of z•, : , Afrei., which is the deviation from monthly median 2000 value of frei., is made at high and middle latitudes 1 including aurora region. The Afain is a good indica- 3OOO , ! , I , I I , I • I •,/, • ! Jlo I0 •0 ::• I0 •0 tor of the absorption [Sinno and Higashimura,1969], DECEMOER 1967 J•.NUARY 1968 the isopleth of Afmincorresponds to that of the elec- Fig. 1. Variationsof geophysicaland ionosphericparameters tron density in D region produced by ionization of duringDecember 1967 and January 1968 [after Rowe et al., 1969]. NO due to Lyman • solar radiation, because the analysis is confined to the south of the polar night dominantgeomagnetic effects such as energeticelec- latitude (70øN). tron precipitation[see review of Thomas,1980], the presentstudy aims to elucidatethe dynamicaleffects of planetary waveson the winter anomaly at high [NO]increase latitudesfrom an observationalanalysis.

ß 2. RESULTS The present study concentrates on the winter ß anomalywhich occurred during a suddenwarming of the 1967/1968 winter from late December 1967 to early January 1968. Variations of geophysicaland ionosphericparameters are shownin Figure 1 [Rowe et al., 1969]. It is apparentthat the absorptionsen- hanceas the warmingsdevelop, as seenin Figuresl a and ld. Accordingto Roweet al., theelectron density observedat Pennsylvania(40øN) reachedits maximum at 75-80 km height during the period from December25, 1967,to January1, 1968.They further noticedthat the winter anomaly has an intimate re- nc reuse lationshipto suddenwarmings, but not to geomag- neticstorms from the long-termcomparison between Fig. 2. A hemisphericchart of geopotcntialheight in the •intcr D region at the period of suddenwarmings, and schematic ,4p index and the absorption. illustrationfor the mechanismof the •inter anomaly[after Kawahira [1982] proposed an observationalevi- hira, 1985]. KAWAHIRA' WINTER ANOMALY INDUCED BY PLANETARY WAVES 797

• '1:.•!•0.1mHz)

(c•) 3 JAN. 196•

I i

.• (b) 1 0 JAN. 1968

Fig. 3. Large-scaleisopleth analysis of the absorptionindex, Afmin,in the form of 3-day runningmeans using the 3-hour averageat 11, 12 and 13 local time of the deviationfrom the monthly median value of frei. (a) on January3, 1968,and (b) on January 10, 1968. Positivearea is an increasein absorption,and negativearea is a decrease.Dotted line indicates60 ø geomagneticlatitude (A - 60ø).

The results of January 3 and 10, 1968, are shown tudesis clearly detected;the scaleis similar as that of in Figure 3, when the warming reachedits peak and planetarywaves. It is noted that the distributionof the absorption enhancementwas remarkable [Rowe Afminalong 60øN is not uniform but showsa "wavy" et al., 1969; Kawahira, 1982]. A systematicand large- pattern.Furthermore, there is a negativecorrelation scaledistribution of the absorptioneven at high lati- of the absorptiondistributions between middle and 798 KAWAHIRA' WINTER ANOMALY INDUCED BY PLANETARY WAVES

Geopotentic[i Height 60"1'4 3 JAN.196 6 (a) 10 JAh[1966 (b)

(College)Fort Churchill West Gelrinish

I I I I I I i - Height 180 90W 0 180 90W 0 --.,,. (km) ß1 km F t 'x O.056mb --70 68km--•-- • -1 kmm-- • ß ß1 km F 62kmT '' -• ''' x..0.15rob -- 60

-1 krn •'-- -- 50

ß1 km F

-1 km L-- --40

-- 30

Fig. 4. Vertical profile of geopotentialheight along 60øN from 0øW to 180øW. Real height is shown in the right-hand side,and the scaleof the geopotentialheight in the left-hand sideis twice the real height. The thick line is after the Upper Air SynopticMap (NOAA), and the dotted lines are estimatedfrom hydrostaticbalance with use of observedtemperature at Fort Churchill and West Geirinish, (a) on January 3 and (b) on January 10, 1968 [after Kawahira, 1982]. high latitudes;e.g., along 90øW positiveAfmin is lo- by planetary waves as shown in Figure 4. Thus the cated south of 60øN and negative Afmin, north of planetary wave induced winds would be dominant 60øN. even up to the upper D region, becausethe data at Then the evidence of large amplitude planetary College are the average between 75 and 105 km waves in D region during the 2 days was given by height layer. Kawahira [1982] as shown in Figure 4, where geopo- Although there are no wind data in Europe,it can tential height distributions along 60øN in western be suggestedfrom Figure 4 that northward winds longitude from 0øW to 180øW are shown from 30 to may have been dominant on January 3, sincelowest 70 km height. It is apparent that over North geopotentialheight at 70 km is located near 10øW, America, geostrophicsouthward winds are dominant but southward wind on January 10, since the mono- during the 2 days, sincegeopotential height become tonic increase toward west of geopotential height at higher in west side than east side. The southward 70 km is seen. winds over North America are consistent with the The comparisonbetween Figures 3 and 4 indicates winds observedat College by meteor radar by Hook a tendencythat southwardwinds (northward) would [1972] as shown in Figure 5, where zonal and merid- cause an increase (decrease)of the absorption at ional componentsof the winds averaged between 75 middle latitudes, but a decrease(increase) at high and 105 km height reverseduring the period of peak latitudes.Large-scale wave winds thus affect the dis- stratosphericwarmings (Figure 1). In Figure 5, wind tribution of D region NO even in the auroral region changes at White Sands by meteorological rocket as has theoretically been studied by Roble and Gray soundingsare also shown, which represent similar [1979], becausethe life time of NO is comparableto changes as those at College. It is evident that the that of atmosphericmotion. It is to be noted that the southward winds in meteor height at College are wind directions at College (Figure 5), being south- consistentwith geostrophicsouthward winds induced and westward, are consistentwith the distribution of KAWAHIRA: WINTER ANOMALY INDUCED BY PLANETARY WAVES 799

((3) College(64ø51'N 147ø50'W) W,nd (75-105 km)

40-

20 -

40 T

:::) w,ø - 20 t e• e' - 40b '• •0I ;0 30•!! •0I 20! 30/ Dec 1967 Jon 1968

(b) Wh,te SonOsM,ss,te Ronge(32ø23'N 106ø36'W) W,nd (70-80 krn)

40-

N 20- I _

i S - 20 -

- 40 -

E 40 I 20 '----_..__ W - 20 I I I !l I i II I i I I io 20 301 io 20 301 io 20 29 Dec 1967 Jon 1968 Feb.

Fig. 5. Variationsof windsin the uppermesosphere. W and E indicatethe wind toward west and east, respectively,and likewiseN and S towardnorth and south:(a) observedby meteorradar at College[after Hook, 1972]'(b) observedby meteorologicalrocket soundings at WhiteSands [after Kawahira, 1982]. the absorptionmaxima on January 3 and 10, being influencedby NO changesowing to energeticelec- located from northeast to southwest over North tron precipitation associated with geomagnetic America as seen in Figure 3. This fact would also storms, but at Boulder they may be influenced by supportthe planetarywave wind effecton the winter NO changesowing to transport effects associated anomaly. with neutral winds. In order to see these effects, there The effect of planetary waves on high-latitude D are shown the absorptionchanges observed by ri- regionis furtherinvestigated from the time changeof ometer at Dixon Island (Figure 6a) as an index of the absorption and meridional winds as shown in energeticelectron precipitations, and the meridional Figure 6. Figure 6b showsthe time changeof Afminat windsat Collegeand White Sands,respectively (Fig- the two stations, Fort Churchill, which is located ures 6c and 6d). north of 60ø geomagneticlatitude, and Boulder, The absorptionchanges at Fort Churchill seemto which is located south of that latitude. Thus vari- correlate well with that of the riometer absorption, ations of the absorption at Fort Churchill may be which could indicate the NO enhancementthrough 800 KAWAHIRA' WINTER ANOMALY INDUCED BY PLANETARY WAVES

DEC. 1967 JAN 1968 1 10 20 30 1 10 20 30 high-latitude source regions to middle-latitude sink I I I II I I I regions. As a result of this effect, NO increasesat ) Abs.(32MHz)Dixonis.(73.3øN 80.3øE) middle latitude, but the decreaseat high latitude is induced as suggestedfrom the schematicpicture of

dB Figure 2; in other words,the meridionalgradient of NO profile between middle and high latitudes be- comesless steep. The sameanalysis over Eurasia is shownin Figure 7. Some different features from Figure 6 are seenin o that middle-latitude absorption (at Garchy) is com- •,fro,. • Church,11( 58.8•194.•W ) parable to that in high latitude (at Murmansk). This ...... Boulder(40.OøN 105 :J'W) *2 may be due to a differencein a NO distribution over (b) ' (MHz) the aurora region between North America and Eu-

*1 rasia [Cravens and Stewart, 1978]. However, during

DEC. 1967 JAN 1968 1 10 20 30 1 10 20 30 I I I II I I I 4-(0 Abs.(32 Ml-lz) Dixonis.(73.3'N 80.3'E)

MERIDIONAL WIND 3- College(645'N, 147 5•V) 75-105 km dB

20 • o 1

White Sands( 32.2øN.1064'W) 70-80km

-40

1 10 20 30 1 10 20 30 DEC JAN 1967 1968

Fig. 6. Time changesof physical quantities. (a) The riometer absorption at Dixon Island (A - 63ø) which is defined as a diurnal MERIDIONAL WIND maximum value among every hourly observedones. (b) Afmi, at Fort Churchill (A = 68.7ø) and Boulder (A -49.0ø), respectively. College(645øN, 147 5'W) 75-105 km (c) Meridional wind at College observedby meteor radar [Hook, 1972], and (d) meridional wind at White Sands observed by meteorologicalrocket soundings. energetic electron precipitations [Iwagami and Ogawa,1980]. Althoughthe absorptionat Boulderis White Sonds(32.2øN.106.4'W) 70 - 80 km not remarkablecompared to that at Fort Churchill during nearly all periods,except during the period from late December to early January when the clear ' '-'-'"-'-' '-'-" '-U,' J. L',~', ,' reversedchanges are seen,the absorptionat Boulder S-40 becomesstronger, but at Fort Churchill it becomes 1 10 20 30 1 10 20 30 weaker, and furthermore then southward winds are DEC JAN dominant (Figure 6c and 6d). These resultswould 1967 1968 indicate that the large-amplitude planetary wave Fig. 7. Same as in Figure 6, but (Figure 7b) is at Murmansk southward winds carried away abundant NO from (A -- 63.5ø) and Garchy (A -- 49.6ø). KAWAHIRA: WINTER ANOMALY INDUCED BY PLANETARY WAVES 801

5577• Em•sslon (Rayle•hs) HauteProvence sorption at Garchy (47.2øN, 3.1øE) are compared. 5oo A ½•'ø•N5 •E) Similar changesof the two parameterswould suggest /% that abundant atomic oxygen in the polar region could also be transported by southward winds as- :OoOo/v'100 • sociatedwith planetary waves. It is to be noted that roles of eddy diffusion owing +2-- Afrn,n (MHz) Garchy (47 2øN3 fE) to breaking gravity waves [Lindzen, 1981] are still

+1 important as stressedby Offermann et al. [1982], be- causeits continuous transport of NO from the lower 0 thermospherecould maintain higher NO density in the polar region.Further eddy diffusionitself may be -1- influenced at the time of sudden warmings due to 10 20 301 10 20 30 DEC. JAN zonal wind variations which affect vertical propaga- 1967 196B tion of internal gravity waves as discussedby Dunk- Fig. 8. Variationsof 5577-• airglowemission at Haute Prov- erton and Butchart [1984]. These variations will be ence[after Fukuyama,1977] and of the absorptionat Garchy. studied in more detail. In the present study, temperature rise effectson the the active period of planetary waves from late De- absorption could not be elucidated. Temperature cember to early January, a remarkable absorption changesduring the warming are seenas a rise in the increaseat middle latitude and the decreaseat high stratospherebut a decline in the mesosphere[Hirota latitude are apparent as is seenin Figure 6. and Barnett, 1977]. Thus the absorption enhance- These evidencesshow a direct effect of atmospheric ment during the stratosphericwarmings may be due motions on the lower ionosphereat high latitudes. to a NO increase.However, seasonallyaveraged temperatures of the winter D region are warmer than

3. CONCLUDING REMARKS AND DISCUSSION those of the summer. Thus the cluster ion density in winter would be far lessthan in summer even during In the present paper, from meteorological and ab- sudden warmings. Therefore a warmer winter D sorption data analysis at the high and middle lati- region could give an adequate condition for electron tudes,it is found that large-amplitudeplanetary wave enhancement[Arnold and Krankowsky, 1977]. winds transport NO-rich air from high to middle lat- The effect of atmosphericwaves on the polar lower itudes, resulting in a NO (absorption) increase at ionosphere,as discussedin the presentpaper, may be middle latitudes but a NO (absorption) decreaseat quite interesting and important in the aeronomy of high latitudes. However, during the period of less the mesosphereand lower thermosphere. Since the active planetary waves, the daily changesof the ab- polar region acts as the chemical source region of sorption at high latitude are similar to that of ener- NO and O, which are produced by geomagnetic getic electron precipitation, but those at middle lati- storms, the transport process in the polar region tudes are weak. These results show interesting effects could be one of the important factors in the chemis- of meteorological and geomagneticdisturbances on try and thermodynamicsof the mesosphereand even the high-latitude lower ionosphere. upper stratosphereat middle and high latitudes. Furthermore, the planetary wave wind transport from the polar region at the time of suddenwarming Acknowledgments. The author would like to thank the re- can play an important role not only on NO but also viewers for their constructive comments. He also expresseshis heartythanks to YoshidaFoundation for Scienceand Technology on atomic oxygen (O) changes,because its lifetime is for helping him to attend the workshop at MPAE. long enough to be affected by atmospheric motion and is abundant at higher latitudes. This effect has REFERENCES been recognized as a dramatic increase of airglow Arnold, F., and D. Krankowsky, Ion composition and electron- intensity at the time of sudden warmings as present- and ion,loss processesin the earth's atmosphere,in Dynamical ed by Fukuyama [1977], where variations of the air- and Chemical CouplingBetween the Neutral and Ionized Atmo- glow intensity at Haute Provence (44øN, 6øE) at the sphere, pp. 93-127, edited by B. Grandel and J. A. Holtet, D. Reidel, Hingham, Mass., 1977. time of the sudden warming of the 1967/1968 winter Cravens,T. R., and A. I. Stewart, Global morphology of nitric were also analyzed. In Figure 8, the variation of 5577 oxide in the lower E region, J. Geophys.Res., 83, 2446-2452, A airglowemission at Haute Provencesand the ab- 1978. 802 KAWAHIRA: WINTER ANOMALY INDUCED BY PLANETARY WAVES

Dunkerton, T. J., and N. Butchart, Propagation and selective Offermann, D., H. G. K. Bruckelman, J. J. Barnett, K. Labitzke, transmissionof internal gravity wavesin a suddenwarming, J. K. M. Torkar, and H. U. Widdel, A scale analysis of the D Atmos. Sci., 41, 1443-1460, 1984. region winter anomaly,J. Geophys.Res., 87, 8286-8306, 1982. Fukuyama, K., Airglow variations and dynamics in the lower Pakhomov, S. V., Z. Ts. Rapoport, and V. M. Sinel'nikov,Investi- thermosphereand upper mesosphere,III, Variations during gation of the winter anomalyin ionosphericabsorption in Janu- stratosphericwarming events,J. Atmos. Terr. Phys., 39, 317- ary 1981, J. Atmos. Terr. Phys.,46, 129-141, 1984. 331, 1977. Roble, R. G., and J. M. Gray, The effectof horizontal transport on Hirota, I., and J. J. Barnett, Planetary waves in the winter auroral NO densities,Geophys. Res. Lett., 6, 703-706, 1979. mesosphere•Preliminaryanalysis of Nimbus 6 PMR results,Q. Rowe, J. N., A. J. Ferraro, H. S. Lee, and A. P. Mitra, Changesin J. R. Meteorol. Soc., 103, 487-498, 1977. electron density and collision frequency at University Park, Hook, J. L., Wind patternsat meteoraltitudes (75-105 kilometers) Pennsylvaniaduring the stratosphericwarming of 1967/68, J. above College,Alaska, associatedwith midwinter stratospheric Atmos. Terr. Phys.,31, 1077-1082, 1969. warmings,J. Geophys.Res., 77, 3856-3868, 1972. Russel,J. M., III, S. Solomon, L. L. Gordley, E. E. Remsberg,and Iwagami, N., and T. Ogawa, An antarcticNO densityprofile de- L. B. Callis, The variability of stratosphericand mesospheric duced from the gama band airglow, Planet. SpaceSci., 28, 867- NO 2 in the polar winter night observedby LIMS, J. Geophys. 873, 1980. Res., 89, 7267-7275, 1984. Jones,G. A., and S. K. Avery, The transport of nitric oxide by the Schwentek, H., Regular and irregular behavior of the winter mean circulation and planetary waves,J. Atmos.Sci., 41, 2756- anomaly in ionosphericabsorption, J. Atmos. Terr. Phys., 33, 2773, 1984. 1647-1650, 1971. Kawahira, K., An observational study of the D-region winter Sinno, K., and M. Higashimura,Development of ionosphericab- anomaly and sudden stratosphericwarmings, J. Atmos. Terr. sorption associatedwith stratosphericwarmings in 1958 and Phys.,44, 947-955, 1982. 1963,d. Atmos.Terr. Phys.,31, 1353-1357, 1969. Kawahira, K., An observationalstudy of the winter anomaly in Solomon, S., P. J. Crutzen, and R. G. Roble, Photochemicalcou- lower latitudesduring suddenwarmings, J. Atmos. Terr. Phys., pling betweenthe thermosphereand the lower atmosphere,2, D in press,1985. region ion chemistryand the winter anomaly,J. Geophys.Res., Kondo, Y., and T. Ogawa, Odd nitrogen in the lower thermo- 87, 7221-7227, 1982. sphereunder auroral perturbation,J. Geomagn.Geoelectr., 28, Thomas,L., The lower ionosphereat high latitude,in Exploration 253-282, 1976. of the Polar UpperAtmosphere, edited by C. S. Deehr and J. A. Lindzen, R. S., Turbulenceand stressowing to gravity wave and Holtet, pp. 99-112,D. Reidel,Hingham, Mass., 1980. tidal breakdown,J. Geophys.Res., 86, 9707-9714, 1981. Offermann, D., Recent advances in the study of the D-region K. Kawahira, GeophysicalInstitute, Kyoto University,Kyoto winter anomaly,d. Atmos.Terr. Phys.,41, 2446-2452, 1978. 606, Japan.