Disturbances of the western European ionosphere during the total solar eclipse of 11 August 1999 measured by a wide ionosonde and radar network

Dis u bances o he wes e n Eu opean ionosphe e du ing he o al sola eclipse o 11 Augus 1999 measu ed by a wide ionosonde and ada ne wo k T. Fa gesa, J.C. Jodogneb, R. Bam o dc, Y. Le Rouxd, F. Gau hie d, P.M. Vilae; ∗, D. Al adill , J.G. Sole , G. Mi og aCEA=LDG, BP 12, 91860 B uye es-le-Chˆa el, F ance bGeophysics Di ision I.R.M. 3, A e Ci culai e, 1180 B ussels, Belgique cRAL, Chil on, Didco Oxon, UK dCNE=DMR=TSI, F ance Telecom, 2 A e P. Ma zin, 22301 Lannion Cedex, F ance eCETP=CNRS, 4 A e Nep une, 94 107 S Mau -des-Foss6es Cedex, F ance Obse a o i d’Eb e, 43520, Roque es, TARRAGONA, Spain gEl ARENOSILLO Space Science Di ision, Mazagon-Mogue , 2110 HUELVA, Spain Abs ac The 11 Augus 1999 Sola eclipse o ali y pa h an ac oss wes e n Eu ope a nea -cons an la i udes o abou 49◦N. I occu ed a mid- ime o a sequence o h ee days wi h s eady sola wind and quie magne osphe ic condi ions. I s esponse was obse ed by a sco e o ionosphe ic acili ies, which will p o ide high- esolu ion p obing o he a ious dis u bances. Fi s esul s allow us o compa e he ime =uc ua ions a a ious dis ances om o ali y on he eclipse and adjacen days, inside a 5◦Wes o 5◦Eas longi ude a ea. In his p elimina y wo k he oF1 and oF2 ime changes a e p esen ed in con ou maps on a 50 km size g id. They show he expec ed longi ude ansi o eclipse pe u ba ion. We en u e b ie commen s on he eclipse-own signa u es as sepa a e om he a ious wa e oscilla ions de ec ed p io o eclipse ime by 12.4 MHz pano amic azimu h scans o he Losque ada nea Lannion (B i anny). Keywo ds: Ionosphe ic dis u bances; Sola eclipse 1. In oduc ion Sola eclipses a e s udied by ionosphe is s om he be- ginning o he ionosphe e s udies (Ra cliAe, 1956). Thei dis u bances p o ide a unique oppo uni y o unde s and he ionosphe e o ma ion and dynamic p ocesses. The loss a e o he F egion ha e been well analyzed by se e al au ho s (Rishbe h, 1968). Since 1970, a new in e es has isen in so- la eclipse s udies, abou he p opaga ion o a g a i y wa e gene a ed by he a mosphe ic cooling in he Moon shadow ∗Co esponding au ho . Tel.: +33-1-45-11-4237. E-mail add esses: [email p o ec ed] (T. Fa ges), [email p o ec ed] (P.M. Vila). (Chimonas and Hines, 1970). When a sola eclipse aAec s a quie -pe iod ionosphe e, we ha e a chance o d awing c u- cial es s abou he cha ac e is ic esponse o he ionosphe ic plasmas o he se o pe u ba ions imposed by he s ong and well-deGned occul a ion o he EUV and IR sola adia ion. Howe e , ecu en ionosphe ic sola eclipse s udies ha e b ough many inconclusi e esul s. No wo sola eclipses p oduce he same plasma mo ion dis u bances because, he luna cone a el geome y and i s angula eAec s on ou magne ized plasmas a e diAe en each ime (in ac ou e- sul s sugges a longi ude di e si y o he dis u bance be ween i s p e-noon and pos -noon phases). The p esen e en ben- eG ed om he dense Wes Eu opean ionosphe ic ne wo k, and is a challenge o a mosphe e and ionosphe e models Fig. 1. Map o he o al sola eclipse pa h a 200 km heigh on Augus 11, 1999, abo e Wes e n Eu ope. Loca ion o e ical incidence ionosondes used in he pape is indica ed wi h ci cles. e.g. TGCM (Ridley e al., 1984; Roble e al., 1986), CTIP (MJ ulle -Woda g e al., 1998)). The imp o emen s in coope a ion and communica ion among Eu opean obse e s g ea ly acili a ed he compila- ion o da a. This allowed di ec compa ison o apid a i- a ions in i ual heigh ionosphe ic p oGles and in he a i- ous ue heigh - equency a ia ions. Fig. 1 is a gene al map o he eclipsed ea h domain wi h he ame encompass- ing ou p esen Wes Eu opean a ea o s udy, expanded in Fig. 4. The magne ic ac i i y indices we e no ably weak on he 10;11 and 12 Augus 1999 wi h Kp alues o 9;12+ and 15+, espec i ely. The sola =ux F10.7 was o 128. This makes i easie o isualize he main p ocesses ha dis u bed he plasma on he eclipse day. We he e p esen a G s look assessmen o he simul a- neous measu emen s om 12 e ical incidence (VI) iono- sondes and some auxilia y ada s. This apid communica ion pape deals he e o e wi h: a desc ip ion o he expe imen- al de ice and a quali a i e desc ip ion o he mos salien eclipse cha ac e is ics, as well as g a i y wa e and o he oscilla ions obse ed du ing he eclipse and on Augus 12, selec ed as con olled day in Sec ion 2; Some qua e -hou ly 2D maps o he oF1 and oF2 eclipse e olu ion o e he geog aphic a ea: 5◦Wes o 5◦Eas by 35◦ o 55◦No h in Sec ion 3; commen s on he a ious pe u ba ions isible on he p esen aw da a, and on g a i y wa e ypology, com- pa ing he analogy a ia ions in 2D- ime ac oss he Wes Eu opean ne wo k in Sec ion 4. 2. Measu emen s in Wes e n Eu ope du ing he sola eclipse 2.1. Expe imen al de ice The obse a ion o his sola eclipse beneG ed om an ex- cep ional expe imen al co e in Wes e n Eu ope. We p esen in Table 1, he cha ac e is ics o he 12 ionosondes used, wi h loca ion, ime esolu ion and maximum sola obscu a- ion a 200 km heigh . The HF sounding ne wo k o he Labo a oi e de De ec- ion e de Geophysique consis s o a ansmi e loca ed in F ancou ille (FRC) and o h ee ecei e s which ha e been se in a adius o 60 km o FRC. The expe imen is based on he p opaga ion o an HF wa e be ween a ansmi e and a ecei e h ough he ionosphe e. This HF wa e is e=ec ed when i s equency is equal o he plasma equency along he p opaga ion pa h. Two sounding me h- ods a e used in pa allel. The G s me hod is e ical inci- dence ionog ams which gi es he elec on densi y p oGle abo e FRC. The second me hod measu es phase shi s due o ionosphe ic mo ions as g a i y wa es, o a elling ionosphe ic dis u bances (TID). Phase shi s a e measu ed by ada equency Dopple spec a, which de e mine he e ical speed a he e=ec ion poin . I an i ine an pe u - ba ion aAec s he h ee links successi ely, he ho izon al componen o he pe u ba ion eloci y is de e mined by iangula ion. Table 1 Loca ion o he ionosondes used du ing he eclipse measu emen s wi h hei ime esolu ion and hei maximum sola obscu a ion a 200 km heigh S a ion Longi ude La i ude Time esolu ion Maximum sola obscu a ion (min) a 200 km heigh (%) F ancou ille 1:7◦E48:4◦N 1 100.0 Lannion 3:45◦W48:7◦N 1 100.0 Hels on 5:3◦W50:1◦N 5 98.7 Dou bes 4:6◦E50:1◦N 10 96.2 Chil on 1:3◦W51:7◦N 4 92.3 Poi ie s 0:3◦E46:6◦N 15 92.1 Roma 12:5◦W41:9◦N 5 83.3 Cailhau 2:1◦E43:1◦N 0.5 81.2 Wa shaw 21◦W52 ◦N 5 79.1 Julius uh 13:4◦E54:6◦N 10 75.9 Eb e 0:3◦E40:8◦N 5 71.6 El A enosillo 6:7◦W37:1◦N 15 55.7 Ionog am esul s a e p esen ed in his pape while he Dopple da a will be ho oughly s udied in a u u e pape . The Losque oblique baksca e ada sounded olumes o he ionosphe e a g oup-pa h dis ances o 300–1200 km om he ada , comple ing he VI ionosonde ne wo k. 2.2. Losque ada esul s The Losque ada acili y, nea Lannion, has a ci cula a ay o 120 an ennae scanned e e y 1.5 min. Du ing he eclipse expe imen , he equency used was 12.4 MHz wi h an ele a ion angle a ound 30◦. Any deple ion in he F laye causes deepe pene a ion o he sounding wa e, hus longe g oup pa h. The oblique soundings p esen ed in Fig. 2 co espond o F egion e=ec ing a eas om h ee azimu hs (each panel showing wo azimu h plo s), espec i ely, loca ed a ound he No he n I ish Sea (00–03◦), he Sou he n Sco land Eas coas (24–27◦), Sou h-Eas o F ance (138–141◦, Hau e P o ence) and No h-Wes o Spain (198–201◦). The ange scale is gi en in e u n backsca e g oup pa h; hence he a ge is loca ed a hal his ange. Bo h no hwa d sequences show eclipse-cen e ed holes o wid h apidly changing wi h azimu h, while he Sou h- Eas wa d eco d clea ly e idences a pene a ion dis u bance o opposi e symme y o ha o he no hwa d sigh ings. Fo all azimu hs he u hes g oup pa h ime coincides exac ly wi h he maximum obscu a ion ime in he a ge e=ec ion a ea conce ned. The Sou h-Eas o Sou h-Wes -o ien ed se- ies e eals wide and shallowe eclipse deple ions. The eclipse hole is qui e clea in he No h wi h a on al slope a 0920 UT, while a simila on o No h No h-Eas s a s a 1010 UT. The hole is deepes in he No h No h-Eas whe e he a ge is loca ed nea es o he eclipse o ali y pa h. P e-eclipse oscilla ions iewed in his h ee azimu hs we e widely diAe en ; wi h a s ong (non-eclipse) wo hou quasi-pe iod wa e in he No h, a ain o h ee wa es o abou 50 min in No h No h-Eas , and wi h much weake =uc ua ions owa d 140◦E. A ound 0830 UT on 201◦and 0900 UT on 24◦, a clea inc ease o g oup pa h indica es a No h-Eas wa d mo ing deple ion. 2.3. Ionog am esul s Fig. 3 shows wo se s o da a eco ded a F ancou ille and Eb e on Augus 11 and 12, 1999. The F ancou ille ionosonde was loca ed jus unde he o al sola eclipse pa h in he ionosphe e be ween 130 and 250 km heigh on ac- coun o he sola zeni hal angle (53◦a he maximum obscu- a ion ime). The sola obscu a ion maximum o he Eb e ionosonde was 71%. F ancou ille esul s: Compa ing measu emen s eco ded on Augus 11 and 12, bo h oF1 a ia ions a e iden ical ou - side he eclipse ime in e al, ollowing a Chapman law; inside his ime in e al, oF1 alues on Augus 11 a e lowe han on Augus 12. The maximum pho ochemical oF1 eclipse all o 37% is eached a 1023 UT when he Sun is o ally eclipsed. In he F1 laye , he s a o he eco e y phase (when oF1 begins again o inc ease) is synch onous wi h he end o he o al sola eclipse (1023 UT). Howe e , he s a o he F2 laye eco e y phase is 20 min a e he local maximum oc- cul a ion (99%). This obse a ion ag ees well wi h p e ious esul s (Cohen (1984): 20–50 min; Salah e al. (1986): 20 –30 min; Walke e al. (1991): 20–40 min; Kolokolo e al. (1993): 30–45 min; A aimo ich e al. (1998): 10 min). Eb e esul s: The maximum oF1 dec ease is eached nea ly a he maximum obscu a ion ime wi h 24%. The 19% all in oF2 alues lags by abou 15 min behind he maxi- mum obscu a ion ime. Fo he wo s a ions, as also o he o he s a ions no shown he e, he oF2 d op du ing he sola eclipse is weak. Fig. 2. Re u n g oup pa h e sus UT om Losque Island (Lannion) in ou diAe en azimu hs: 0–3◦, op panel; 24–27◦, second panel; 138–141◦, hi d panel and 198–201◦, bo om panel. 3. oF1 and oF2 con ou maps Mo gan e al. (1978) p esen ed a s udy wi h Gne space and ime esolu ion showing he e olu ion o TID. In his pape , we do no use he echniques de eloped by Mo gan e al. Fo hese p elimina y esul s, we p esen he la ge ime (pe iod o abou 2 h) and space (abou 1000 km) scale pe u ba ions in he wes e n Eu opean ionosphe e du ing he 11 Augus 1999 sola eclipse wi h he help o in e pola ed e ical ionosonde c i ical equency maps. Wi h ou ionosondes dis ibu ed on he same magne ic me idian, Naga a e al. (1955) had d awn a 2D map o a sola eclipse pe u ba ion a midday in he E egion whe e he x-axis was ime and y-axis he dis ance be ween ionosondes along he me idian. Ha ing da a om 12 Eu opean ionosondes, we build in- s an aneous maps in 2D ex ensions o he nea -cons an heigh s o hMF1 (abou 190 km (Fa ges, 2000)) and hMF2. T acing o iso- equency con ou s is done using a 2D in e - pola ion me hod (Akima, 1978). Fig. 3. oF1 and oF2 changes abo e F ancou ille and Eb e du ing he sola eclipse and ollowing hou s, he Augus 11, 1999 and du ing he same in e al o ime on Augus 12, 1999. The s a and end imes o he sola eclipse a e shown wi h dashed e ical lines; he maximum obscu a ion ime is indica ed wi h a do ed e ical line. The indi idual oF1 eclipse ime plo s a e qui e accu- a e ( o be e han 0.1 MHz excep o b ie in e als o spo adic E occul a ion) and clea ly minimize a maximum obscu a ion ime. Thus we can use he oF1 pa ame e space =uc ua ions as indica o s o he eclipse dis o ions. Fig. 4 p esen s en such qua e -hou ly maps be ween 0830 and 1045 UT and wo mo e a 1200 and 1300 UT. The h ee si es loca ed u he Eas : Julius uh, Wa shaw and Roma p o ided da a o he acing as bounda y alues. A G s poin dese ing no ice is he wide la i ude ex en o he eclipse plasma densi y deple ion, which exceeded 1300 km adius a El A enosillo (sou he n Spain). The eclipse deple ion should appea la e han 0900 UT om he wes e n longi udes. F om 0830 o 0900 UT, he hole obse ed Wes o he 0◦longi ude is p obably he las eAec o he A lan ic la ge-scale g a i y-wa e appa en on he Losque scans. The eclipse deple ion is maximum a 1015 UT. I ecedes om 10:45 UT as expec ed by he eas wa d Moon shadow mo ion. A 1200 UT, he sola eclipse eAec s disappea comple ely. This pe u ba ion is oo wide (a quasi-ci cle o diame e much la ge han 1000 km) and oo as (a ound 790 m=s, ha is Moon shadow eloc- i y) o be impu ed o any hing else han he sola eclipse. Fo ins ance, he sou he nmos s a ion El A enosillo, abou Fig. 4. Con ou maps o oF1 qua e -hou ly alues be ween 0830 and 1045 UT and be ween 1200 and 1300 UT d awn on a 50 ×50km g id. An open ci cle is d awn on he eclipse pa h when he Moon shadow is p ojec ed on he ionosphe e (a 1015 and 1030 UT) a 200 km heigh abo e Wes e n Eu ope. 1300 km om o ali y, measu ed deple ions o −20% on oF1, −17% on oF2. The peak a ound Cailhau a 1200 UT is possibly he G s sign o an eme gen bow wa e. A 1300 UT abo e Eb e, 5.75 MHz con ou e eals a on al peak in F1 laye elec on densi y, ha was obse ed on he indi idual Eb e plo in Fig. 4. (Con inued). Fig. 3. This wa e p opaga es sou hwa d a a ound 80 m=s. I s p ope ies a e discussed in a companion pape (Al adill e al., his issue). Fig. 5 p esen s 12 qua e -hou ly oF2 maps be ween 0845 and 1130 UT. The Moon’s shadow a 300 km is plo ed on he map. The eclipse dis u bance is delayed by abou 30 min on he maximum obscu a ion ime, as isible in Fig. 3. The 1030 UT map shows a minimum oF2 a ea cen e ed on 4◦W when he Moon’s shadow has al eady eached 7◦E. Two episodes o oF2 deple ion occu , espec i ely, a 0945 and 1100 UT nea he 0◦longi ude and be ween 45 and 55◦la i ude. 4. Discussion The main pu pose o ou discussion is o ela e he wo 2D c i ical equency maps o he Losque ada scans. Consis ency be ween he wo se s o measu emen s would conG m he a ia ions and allow a Gne esolu ion mapping o dis ances un il 1200 km om he Losque ada . On Losque scans, he majo phenomenon is obse ed be ween 0930 and 1130 UT as a la ge inc ease o g oup pa h o 300 km. This deple ion is ob iously due o he eclipse elec on densi y loss. Fig. 5. Con ou maps o oF2 qua e -hou ly alues be ween 0845 and 1130 UT d awn on a 50 ×50 km g id. An open ci cle is d awn on he eclipse pa h when he Moon shadow is p ojec ed on he ionosphe e (a 1015 and 1030 UT) a 300 km heigh abo e Wes e n Eu ope. We will e e o wo ea u es mos no iceable on he qua e -hou ly maps (0845 UT on oF1 and 1100 UT on oF2), and o ano he eAec mo e appa en on he Losque scans be ween 0930 and 1000 UT on he no he n azimu hs. (1) We no iced in Sec ion 3 (Fig. 4) he impo an oF1 hole s a ing a 0845 in No h-Wes Spain which seems o mo e o No h-Eas ac oss B i any owa d cen al England. Fig. 5. (Con inued). A esidual o he oF1 densi y hole obse ed on he wes a ea om 0830 o 0930 UT is measu ed a FRC. The Dopple measu emen s conG m he No h-Eas wa d displacemen . The Losque scans on 198–201◦(0830 UT) and on 24–27◦ (0900 UT) conG m his deple ion and i s mo ion. We canno p esen ly ela e his singula i y o any di ec eclipse eAec . The non- i ial cha ac e o his hole and o i s displacemen is p esen ly diNcul o in e p e . (2) The oF2 deple ion a ound Poi ie s a 1100 UT (Fig. 5) can be compa ed wi h he 138–141◦scans om Losque (Fig. 2). Fo his a ea and ime, he oF1 a ia ions a e uni o m; he abno mal e=ec ion is he e o e aking place a F2 laye le els be ween 650 and 900 km g oup pa h dis ance om Losque . Looking back om Fig. 5 a 1100 UT, we no ice a s eep densi y g adien in his di ec ion a ound he 700 km dis ance (sligh ly No h o Cailhau). P opaga ion