Changes of the rst Schumann resonance frequency during relativistic solar proton precipitation in the 6 November 1997 event

Transcription

1 Ann. Geophysice 17, 1293±1297 (1999) Ó EGS ± Springer-Verlg 1999 Chnges of the rst Schumnn resonnce frequency during reltivistic solr proton precipittion in the 6 November 1997 event V. C. Roldugin, Ye. P. Mltsev, A. N. Vsiljev, E. V. Vshenyuk Polr Geophysicl Institute, Acdemy, Aptity, Murmnsk reg., Russi Received: 23 September 1998 / Revised: 26 Mrch 1999 / Accepted: 26 April 1999 Abstrct. The vritions of the rst mode of Schumnn resonnce re nlyzed using dt from Kol peninsul sttions during the solr proton event of 6 November On this dy the intensive ux of energetic protons on GOES-8 nd the 10% increse of the count rte of the neutron monitor in Aptity between 1220 nd 2000 UT were preceded by solr X-ry burst t 1155 UT. This burst ws ccompnied by simultneous increse of the Schumnn frequency by 3.5%, nd the reltivistic proton ux increse ws ccompnied by 1% frequency decrese. These e ects re explined by chnges of the height nd dielectric permebility of the Erth-ionosphere cvity. Key words. Ionosphere (ionospheric disturbnces; solr rdition nd cosmic ry e ects) á Rdio science (ionospheric propgtion) rdius of the outer sphere. Expression (1) cn be rewritten s follows (tking into considertion h ) f n x n 2p c r p n n 1 n n 1 ˆ 10:6 Hz ; 2 2p 2 The simplest model predicts the rst order frequency f1 ˆ 10.6 Hz. Observtions yield on verge f 1 ˆ 7.8 Hz (Blser nd Wgner, 1962). The frequency becme smller becuse the ionosphere is not perfect conductor. Introducing the impednce Z of the upper edge (r =b), the following eqution results for the resonnce frequencies (providing jzj1, h ) (Bliokh et l., 1980) x n ˆ c r n n 1 Z2 2 4 h 2 iz 2! h : 3 If the dielectric permebility t r>b is equl to e = const, the impednce is: Z ˆ p 1 : 4 e 1 Introduction The cvity between the Erth nd ionosphere is resontor for electromgnetic wves. The Schumnn resonnce frequencies cover severl Hz. The simplest vcuum model con ned with two concentric perfectly conducting spheres yields the frequency (Schumnn, 1952) s x n ˆ c n n 1 1 h ; 1 where n is n integer number, c is the light velocity, is the rdius of the inner sphere, h=)bis the distnce from the inner sphere to the outer one, nd b is the Correspondence to: V. Roldugin, E-mil: roldugin@pgi.kolsc.net.ru For n isotropic ionosphere, the dielectric permebility hs the following form x 2 0 e ˆ 1 5 x x im where x is the wve frequency, x =(4pe 2 N/m) 1/2 is the electron plsm frequency, e nd m re the electron chrge nd mss respectively, N is the electron number density, m is the frequency of collision of electrons with neutrls nd ions. One cn choose such vlues of e nd h tht severl resonnce hrmonics clculted with the help of Eq. (3) coincide su ciently well with the observed ones. A more complicted distribution of e with the height llows us to explin the observed width of the resonnce bnd relted to the resontor qulity. The ionospheric D region (heights from 60 to 90 km) ppers to determine the resonnce spectrum. In the rel ionosphere, e vries both with the height nd in horizontl directions. In ddition, the dielectric

2 1294 V. C. Roldugin et l.: Chnges of the rst Schumnn resonnce frequency permebility of the ionospheric plsm in the geomgnetic eld t heights bove ~75 km is tensor. Hence Z in Eq. (3) presents some integrl prmeters. The wves in the Erth-ionosphere cvity re generted by lightning dischrges. At the sme time, the Schumnn resonnce intensity is controlled not only by thunderstorms but lso by the solr nd geomgnetic ctivity. Sentmn nd Frser (1991) showed the dependence of the observed resonnce power on the e ective height of the ionospheric D region t the observtion point. The dependence of the rst nd second order resonnce modes on solr ctivity ws discovered by Fullekrug nd Frser-Smith (1996). They found 27-dy mplitude vritions coinciding in phse with the solr spot number. From the mesurements mde in Hungry, Mrcz et l. (1997) found similrities between the resonnce mplitude nd the tmospheric electric eld despite the locl chrcter of the ltter. The similrities re observed t di erent temporl scles, from six months to severl hours. Diurnl vritions of the resonnce frequencies cused presumbly by the eccentricity of the geomgnetic dipole were found by Blser nd Wgner (1962) t the very strt of the investigtions. So et l. (1973) revel tht the rst mode frequency depends not only on the zenith ngle of the Sun long the signl propgtion pth from lightning sh to the observtion point, but lso on the plnetry mgnetic disturbnce. Choosing Ap s disturbnce index, So et l. (1973) found tht the second order mode frequency decrese t 0.2 Hz by Ap increses from 0 to 50. They found lso noticeble positive correltion of the rst order mode frequency with 1±8 A Ê solr X-ry intensity. According to Eqs. (3) nd (4), the Schumnn resonnce frequency depends on the dielectric permebility of the ionospheric D region. The powerful gent disturbing this region is solr protons penetrting into the tmosphere during solr proton events (SPEs). In this study we exmine vritions in the rst order frequency of the Schumnn resonnce during such events. 2 Dt The rst resonnce frequency behvior hs been studied for SPE of November 6, The event ws preceded by nother SPE on November 4 relted to the clss 2 opticl re t region with heliocoordintes S14 W33. GOES spcecrfts registered n increse of X-ry ux t 0552 UT tht mximized t 0556 UT nd ended fter 7 h. Solr protons with energy more thn 10 MeV strted t 0830 UT, reched mximum t 1120 UT nd ended t 1123 UT on November 5. On November 6 the re under considertion 2B/ X9.4 occurred in the region S18 W63 on the Sun. The intensities of the solr X-rys nd of the energetic proton ux obtined from GOES-8 stellite re shown in Fig. 2. The 1±8 A Ê X-ry burst with impulsive pro le strted t 1149 UT nd reched its mximum of 7.5 á 10 )4 W/m 2 t 1155 UT. The proton ux with Fig. 1. The spectrum of 2-min uxmeter record (dotted line), its pproximtion by Gussin function (thick line) nd the mplitude versus frequency chrcteristic of uxmeter (thin line). b The spectrum fter frequency correction nd its new pproximtion energy of more thn 100 MeV strted to increse t 1245 UT, mximized t 1640 UT nd remined high till 24 h lter. The rst Schumnn resonnce behvior ws studied with the help of dt from four uxmeter chnnels in Lovozero (68.0 N, 35.1 E). The mgnetic induction coils consist of turns wound over ferrite rods. All the chnnels hve bndwidth between 1±10 Hz. Two chnnels with numbers 10 nd 11 registered D- nd H-components correspondingly nd two others, 14 nd 15, registered the circulrly polrized components, left nd right correspondingly. The clibrtion monochromtic signl dvnced t frequency of 5 Hz. Signls of ll four chnnels were registered in digitl form with 50 ms time step. To obtin the time vritions of the rst Schumnn resonnce frequency, the spectr of two-minute intervls, contining 2400 points, were clculted with the FFT method for ll four chnnels. The typicl exmple is shown in Fig. 1 s sweeping dotted curve. The thick line is its pproximtion by Gussin function inside the intervl of 5.8±9.6 Hz with four unknown prmeters: the frequency of mximum, hlfwidth, mplitude nd the constnt level. The frequency of the mximum is bout 7.4 Hz. This too low vlue nd symmetricl slope structure of the pek re cused by

3 V. C. Roldugin et l.: Chnges of the rst Schumnn resonnce frequency 1295 Fig. 2. Proton E>100 Mev ux nd solr X-ry 1±8 A Ê intensity from GOES-8 dt. b The count rte of Aptitin monitor (thin line) on 6 November, 1997 nd the vrition of frequency of the rst Schumnn resonnce (thick line nd points) re shown the roll-o of low pss lter. The thin line in Fig. 1 is the mplitude versus frequency chrcteristic of the chnnel. Its pek is locted t 6.9 Hz. To correct the signl, spectr were divided this chrcteristic, nd the Gussin pproximtion ws pplied to improved spectr, see the Fig. 1b). the resonnce frequency 7.8 Hz. Relibility of the noted frequency devitions my be con rmed by comprison of the frequency vritions shown in Fig. 2 with others. Figure 3 shows ll four chnnels of the uxmeters. The frequency decrese fter 1230 UT cn be seen on ll of them. The short-lived frequency increse, simultneous with the solr X-ry burst ner 1150 UT, is present lso on ll chnnels. Relibility of the rst SR frequency decrese fter onset of the energetic proton precipittion is con rmed by comprison of the frequency vritions in the tenth chnnel for November 5, 6, nd 7 s shown in Fig. 4. For every dy the frequency vlue of 7.8 Hz is shown by the dshed lines. On ll three djcent dys the frequency increse tkes plce lmost t 5 h UT nd fter middy the frequency increse to 8.0 Hz is observed only on November 5 nd 7. On the SPE dy it is bsent, the frequency remining ner 7.8 Hz. No response of the SR mplitude to the SPE is found. Figure 5 shows the mplitude vritions for 5, 6, nd 7 November in rbitrry units, determined from Gussin pproximtion. The verticl dshed lines pick out the ctive intervl 1230±1700 UT, s in Fig. 2b. The diurnl vritions of the SR mplitude for ll three dys re similr nd rther strong (with the mgnitude nerly 30%). However, the ctive intervl for 6 November shows no essentil distinctions from the sme intervls for djcent dys. We hve estimted the stochstic sctter of the frequency vlues in Fig. 2 s the squre men devition of dt in the intervl 0500±1100 UT fter excluding the liner trend. It is r = 0.11 Hz nd plotted s br t 0600 UT. Let us use this vlue to estimte the signi cnce of the mximum t The men vlue 3 Experimentl results Figure 2b shows count rtes of the neutron monitor in Aptity, nd the rst SR frequency in the tenth chnnel, D-component, displyed both by points for every 2 min intervl, nd by running-verge smoothed curve. The length of the smoothing is 11 points. The pek of the neutron monitor increse is mrked by the dotted verticl lines. Its gretest height exceeds the pre-event bckground by 10%. The dily men vlue of the frequency in Fig. 2 is bout 7.9 Hz. Except for stochstic scttering, there re long-term vritions of the frequency. Augmenttion of the neutron monitor count rte is ccompnied by the resonnce frequency decrese. The gretest decrese is between 0.05±0.1 Hz, i.e., bout 1% of the resonnce frequency. It cn lso be seen tht the X-ry re t 1145 UT ws ccompnied by n increse of the resonnce frequency by 0.3 Hz or by 3.5%. A sttisticl sctter of the frequency vlues is comprble with mgnitude of these mentioned e ects. One of the cuses of the gret sctter is too rough digitizing frequency Dt = 0.05 s, so the Nyquist frequency f N = 1/2Dt = 10 Hz is only little greter thn Fig. 3. Vritions of the rst Schumnn resonnce frequency on 6 November, 1997 t ll four uxmeter chnnels

4 1296 V. C. Roldugin et l.: Chnges of the rst Schumnn resonnce frequency of the frequency for the X-ry burst intervl 1146±1214 UT is 8.18 Hz, the number of points in this intervl is n = 15. Thus the di erence between the dily men vlue of 7.90 Hz nd the men vlue of p the pek 8.18 Hz exceeds the stndrd error E = r= n ˆ 0:028 by exctly 10 times. 4 Discussion Fig. 4. Vritions of the rst Schumnn resonnce frequency on 5, 6, nd 7 November, 1997 in the tenth chnnel Fig. 5. Vritions of the rst Schumnn resonnce mplitude on 5, 6, nd 7 November, 1997 in the tenth chnnel Expression (3) llows us to obtin the resonnce frequency vrition. Assuming jej 1; m ˆ const, x m, neglecting the second term under the rdicl, nd tking into considertion Eqs. (4) nd (5), we obtin the vrition of the rel prt of the frequency: Re df n c dn ˆ Ref n 2pRef n 2N dh Im Z : 6 h 2 h According to Eq. (3), the rst order frequency Re f 1 = 7.8 Hz tkes plce under the following condition Im Z 0:37 : 7 2 h Substitution of Eq. (7) into Eq. (6) yields the vrition of the rst order frequency Re df 1 0:36 dn Ref 1 2N dh : 8 h Expression (8) shows tht the frequency cn chnge due to both electron density vritions in the ionospheric D region nd vrition of the ltitude of the D region's lower edge. X-rys nd solr protons ect the ionosphere in di erent wys. Growth of the X-ry ux enhnces the electron density without signi cnt chnge of the ionosphere ltitude. The 3.5% increse in the Schumnn resonnce frequency observed t 1155 UT on 6 November 1977 could be cused by the density vrition dn/n 19% verged over ll the Erth. Since X-rys ect the dyside ionosphere only, the observed 3.5% frequency increse requires ~38% increse in the dyside electron density. This estimte is within the rnge of possible density vritions during X-ry events. According to clcultions by Curto et l. (1994), solr X-ry burst of clss X9.4 rises the electron density t the ltitude of 100 km, i.e., ner the top edge of the cvity, twice or three times. Solr protons penetrting deep into the tmosphere ionize regions lower the norml D region tht corresponds to decrese of the ionosphere ltitude. According to Eq. (8), the 1% decrese in the resonnce frequency observed from 1220 to 2000 UT could be cused by the globl decrese of the ionospheric ltitude dh/h )2.8%. Assuming the protons precipitte t ltitudes bove 60, mens tht the precipittion re is bout 1/8 of the whole ionospheric surfce nd the required ltitude vrition in the high ltitudes is dh/h )2.8% 8 )22%. If the norml ltitude of the Schumnn cvity is h = 75 km we obtin h 58 km in high ltitudes during the SPE on 6 November, These estimtes re very rough, of course, but

5 V. C. Roldugin et l.: Chnges of the rst Schumnn resonnce frequency 1297 nevertheless they re consistent with commonly observed ionospheric e ects of solr protons (e.g., Mitr, 1974). 5 Conclusion The decrese of the frequency of the rst Schumnn resonnce is found for the reltivistic proton precipittion in the SPE of 6 November, Its vlue is bout 1%. The frequency increse of bout 3.5% coincides with the strong solr X-ry burst lsting 10 min. Both e ects re explined by chnges of the dielectric permebility in the Schumnn cvity. Topicl Editor D. Alcyde thnks referee for his help in evluting these pges. References Blser, M., nd C. A. Wgner, On frequency vritions of the Erth-ionosphere cvity modes, J. Geophys. Res., 67(10), 4081± 4083, Bliokh, P. V., A. P. Nikolenko, nd Ye. F. Filippov, Schumnn Resonnces in the Erth-Ionosphere Cvity, Peter Perigrinus, London, Curto, J. J., C. Amory-Mzudier, J. M. Tort, nd M. Menvielle, Solr re e ects t Ebre: unidimensionl physicl, integrted model, J. Geophys. Res., 99 (A12), ±23 296, Fullekrug, M., nd A. C. Frser-Smith, Further evidence for globl correltion of the Erth-ionosphere cvity resonnces, Geophys. Res. Lett., 23 (20), 2773±2776, Mrcz, F., G. Stori, nd B. Zieger, Vritions in Schumnn resonnces nd their reltion to tmospheric electric prmeters t Ngycenk sttion, Annles Geophysice, 15 (12), 1604±1614, Mitr, A. P., Ionospheric e ects of solr res. Astrophysics nd spce science librry, 46. D. Reidel, Dordrecht 266 p., So, K., M. Ymshit, S. Tnhshi, H. Jindon, nd K. Oht, Experimentl investigtions of Schumnn resonnce frequencies, J. Atmos. Terr. Phys., 35 (11), 2047±2053, Schumnn, W. O., UÈ ber die Strlungslosen Eigenschwingungen einer leitenden Kugel die von Luftscicht und einer IonosphÈ renhuã lle umgeben ist, Z. Nturforsch., 7, 149±154, Sentmn, D. D., nd B. J. Frser, Simultneous observtions of Schumnn resonnces in Cliforni nd Austrli: evidence for intensity modultion by locl height of the D-region, J. Geophys. Res., 96 (A9), ±15 984, 1991.