Ionospheric anomaly due to seismic activities Part 2: Evidence from D-layer preparation and disappearance times

Transcription

1 Nat. Hazards Earth Syst. Sci.,, 7 77, doi:.9/nhess--7- Author(s). CC Attribution. License. Natural Hazards Earth System Sciences Ionospheric anomaly due to seismic activities Part : Evidence from -layer preparation disappearance times S. K. Chakrabarti,, S. Sasmal, S. Chakrabarti, S. N. Bose National Centre for Basic Sciences, J Block, Salt Lake, Kolkata 798, India Indian Centre for Space Physics, Chalantika, Garia Station Road, Kolkata 78, India Maharaja Mandra Chra College,, Ramkanta Bose Street, Kolkata 7, India Received: November 9 Revised: May Accepted: July Published: August Abstract. We show evidences for anomalous ionospheric behaviour Indian navy VLF transmittg station named VX due to South Asian region. We concentrate on variation -layer preparation time () -layer disappearance time () a period sixteen months study ir average behaviors. We identify those which exhibit significant deviations. Separately, we compute energy release by durg this period show that anomalous VLF associated anomalous energy release. We fd that anomaly deviation s from mean learly correlated. We discuss predictability this approach comp termator shift approach usg same set data. Introduction In Sasmal Chakrabarti (9, hereafter Paper ), behaviour from Indian Navy station VX (transmittg at 8. KHz), especially behaviour termators, as received at Kolkata, was presented. ata four solar quiet years from to 8 were used to obta averagg. From s received on seismically active, we specifically studied behaviours sunrise sunset termators, found that re is a higher possibility detectg anomalous termator shifts typically two prior to actual. hat sunrise sun-set termators exhibit shifts towards night before were known quite a while (Molchanov et Correspondence to: S. K. Chakrabarti (chakraba@bose.res.) al., 998; Molchanov Hayakawa, 998; Hayakawa Molchanov, ; Hayakawa et al., ; Chakrabarti et al., ; Maekawa et al., 6; Rozhnoi et al., 9). Prior to this approach, workers such as Gokhberg et al. (989) Gufeld et al. (99) showed that several before earthquake re were anomalies nighttime radio. However, relationship between anomalous termator times (or any or anomaly for that matter) seismicity, (which could be due to a combation lithosphericionospheric couplg, sono-lumescence, breakg bonds tectonic plates, etc.) is poorly understood as now (e.g., Rodger et al., 999; Hayakawa et al., ). Some oretical models have been advanced numerical simulations were carried out. It was found usg simple models that lowerg ionosphere by km could mimic termator time shift (Hayakawa et al., 996; Molchanov et al., 998) although a more cful realistic model showed that it would require a lowerg ionosphere by km which would have or observational effects as happens cases solar fls (Rodger et al., 999; Soloviev Hayakawa, ; Soloviev et al., ). It was concluded that observed shifts termators possible only if propagation path is short (< km). In order to have more put to oretical studies, we feel that it is essential to look for or types correlations which may be present. For stance, time taken to lower - layer boundary early morng (we call this as - layer preparation time or ), time taken to raise it aga eveng (we call it -layer disappearance time) may also be affected by seismic activities. his is because, presence extra ionizg agent, times taken for such activities may be altered significantly. Published by Copernicus Publications on behalf European Geosciences Union.

2 spheric -, s. Res, 7 S. K. Chakrabarti et al.: Ionospheric anomaly due to Kawai, for 998. lg earthuh, K., L., Erprior to t. Haz. o Seis-.KHz ts, Nat. bionouctg io Sci., chanov, n terms hysics. Location relation to seismic circles centered. Location around first pot (FRP) relation to seismic VX circles to ICSP, centered Kolkata, arounddurg first reflection sixteen months pot under (FRP) study. VX locations to ICSP, VX Kolkata, ICSPdurg marked sixteen months filled boxes under study. FRP locations VXa filled ICSP circle. Shadowed marked circles filledrepresent boxes locations FRP is is marked marked, a filledir circle. sizes Shadowed beg proportional circles represent to dividual locations magnitudes., Concentric ircircles sizes beg have radii proportional multiples to dividual km. magnitudes. Concentric circles have radii multiples km. In present paper, we concentrate on sixteen months duration (from November 6 to February 8) from Indian Navy operated transmission station VX (located at Vijayanarayanam, Lat. 8 6 E, Long. 77 N) as received by Stford University made AWESOME receiver cross-loop antennas ICSP made Gyrator-II receiver one loop antenna. receivg station is at Kolkata (Lat. E, Long. 88 N). take a statistics all region durg this period fd that a defite correlation exists between or effective magnitudes. We use Stardized calibration curve for VX-ICSP basele, However, terms predictability, termator shift method is found to be better, especially that we discussg short propagation paths. In Paper, details were presented about receiver transmitters used our study we do not repeat m here. plan present paper is followg: next Section, we present spatial distribution, our methodology is to compute effective magnitude se quakes at mid pot between VX ICSP. In Sect., we present values for period sixteen months determe mean deviations from mean se quantities. We n fd correlation between effective magnitudes deviations. We also determe predictability seismic activities usg this method, comp that obtaed from termator shift method. Fally, Sect. 6 we draw our conclusions. Statistics seismic activities Before we proceed our results, we wish to present statistics region for period sixteen months (November 6 February 8) under consideration. We gar data, such as latitude longitude place earthquake, magnitude, depth epicenter from web-page Indian Meteorological epartment ( In, we present a map which VX ICSP marked filled squs. We consider mid-pot VX-ICSP great circle path (marked a filled circle), which is first reflection pot (FRP) at ionosphere. Sce second third reflection pots also important for VX to Kolkata, we expect that closer to VX or ICSP would also affect possible ionization VX-ICSP path refore VX amplitude. In order to focus our attention on a few which might affect observed s, we note that earthquake preparation zone size is about L.M km, where M is magnitude. Sce quakes under consideration M < 8, we assume that quakes up to a distance km from FRP may be fluential. Usg FRP pot as center, we draw six concentric seismic circles, havg radii multiples km consider which take place se regions. Shaded circles show locations which took place whose magnitudes were used to compute energy release, ir sizes beg proportional to magnitudes quakes. In a, we draw histograms total number se seismic circles. We note that majority happened between km km, thus se quakes could fluence ionosphere through second third hops. In order to compute total energy released by which may affect ionosphere, we use followg formulas (Lowrie, 7): log E =.+.M s (for a magnitude M <.), log E =.+.M s (for a magnitude M >.), where, E energy released earthquake Joule < M s >=. +.M surface wave magnitude. Usg se, we will compute energy released as seen at FRP. We assume that energy a quake drops as a cyldrical wave amplitude /r. We compute great circle path from each se to FRP calculate effective energy at FRP. We n obta effective magnitude < M s > earthquake at FRP by addg contributions from all which take place a given day usg above formula. In b, we draw a histogram effective magnitudes as seen from FRP. plot peaks at < M >.. question we wish to answer is: does this effective magnitude correlate possible anomalies VLF if so, how. Of course, major contribution to effective Nat. Hazards Earth Syst. Sci.,, 7 77,

3 S. K. Chakrabarti et al.: Ionospheric anomaly due to 7 6 No. Earthquakes No. No. Earthquakes No. No. No. No. 6 6 Radius Seismic Circle Kilometer Radius Seismic Circle Kilometer Radius Seismic Circle Kilometer Effective Magnitude Earthquake at Middle Pot ransmitter Receiver. Histograms total number Effective Magnitude Earthquake at Middle Pot ransmitter Receiver seismic circles showg that majority happened Effective between Magnitude km Earthquake at Middle km Pot ransmitter effective Receiver mag nitudes as seen from FRP. plot peaks at. (a-b): Histograms total number < seismic MEffective s >.. circles Magnitude showg Earthquake that at Middle majority Pot ransmitter Receiver happened between km km effective mag-. (a-b): Histograms total number seismic circles showg that majority happenenitudes as seen from FRP. plot peaks at. between (a-b): Histograms km km total number effective magnitudes s >.. < M as seen from FRP. plot peaks at energy seismic release circlesishowg from thatstrongest majority quakes. For this reason, hap- alsobetween use dividual km magnitudes km a separate effective analysis mag- < M s >.. wepened nitudes present results. as seen from FRP. plot peaks at < M s >.. behaviour Followg Chakrabarti et al. (7), we defed two notations Paper, namely,. We defe m here aga for sake completeness. In dawn, -layer boundary takes certa time to go down from its maximum height (Pot A below) to mimum height (sunrise termator or SR, denoted by pot C ). difference C A is -layer preparation time or. Similarly, at dusk, sequence events is opposite through pots, (sunset termator SS), pots B, respectively. difference B is - layer disappearg time or. Out a total sixteen months data that we use, clear data around sunrise (Mutes) (Mutes) (Mutes) (Mutes) (Mutes) (Mutes) 9 8 y +σ 9 7 y+σ 8 6 y +σ y +σ 7 9 y +σ y 6 8 y +σ y +σ 7 y + y- σ 6 +σ y-σ y- σ y -σ y-σ y y-- σ y -σ y -σ. Plots as a function. ark. (a-b): Plots as a function circles actually observed values on a given. ark circles actually observed values day. (a-b): diamonds Plots represent whichas a function associated on a given day diamonds represent which. ark anomalous circles data, actually evenobserved when values earthquake may have taken associated anomalous data, even when earthquake place on a given. (a-b): Plots diamonds represent as a function which may have taken later. place - thick later. solid curve thick is solid average curve is associated anomalous data, even when earthquake. average ark circles values, actually observed th values, solid values curves th drawn solid curves at σ, σ may have taken place - later. thick solid curve is on adrawn given σ at apart. day σ, σ diamonds σ apart. represent which average values, th solid curves associated anomalous data, even when earthquake drawn at σ, σ σ apart. may have taken place - later. thick solid curve is average values, th solid curves were drawn obtaed at σ, σ σ, apart. clear data around sunset were obtaed. On or, eir transmitter was down or our receiver was down. We compute a b plotted m as a function. ypically, varies between m, varies between 7 m. However, a number, se values anomalous. In figures, dark circles actually observed values on a given day, diamonds represent which associated anomalous data, even when earthquake may have taken place later. thick solid curve is average values, computed by removg which show anomalies more σ. th solid curves drawn at σ, σ, σ apart. In a b we show number which various Nat. Hazards Earth Syst. Sci.,, 7 77,

4 7 S. K. Chakrabarti et al.: Ionospheric anomaly due to >σ A 6//8 Stard eviation >σ Stard eviation Stard eviation 7 No. s No. s No. s >σ Amplitude 6 >σ 6 7 >σ 6 Stard eviation. (a-b): histograms showg number which Stard eviation exhibited deviations. devia. above histograms awhich tions σ for onlyshowg 9, but number such deviation exhibited showg deviations.. (a-b):. histograms number deviations which for more 6 9 deviations. deviaabove σ for only, exhibited but such a deviation for Stard eviation tions above σ for only 9, but such a deviation more 6. for more 6.. (a-b): histograms showg number which exhibited deviations. deviations above σ for only 9, but such a deviation deviations took6 place. for more. For stance, deviations above σ for only 9, but such a deviation for more 6. //8 A Correlation seismic activities VLF data In Paper, it was discussed that termator shifts may take place two prior to. So it is pertent to ask, if anomalies also observed prior to or not. In order to give an example, we plot amplitudes (shifted by db vertically for better viewg) for eleven consecutive. On, re was an earthquake a magnitude M 6.. In data, 8, drawn here thick curves, we clearly show that near sunrise termator is totally different, normal sharp drop associated sunrise is replaced by a flatter variation. behavior near sun-set termator is also different from those or. Nat. Hazards Earth Syst. Sci.,, 7 77, B //8 6//8 //8 C A C EQ- //8 //8 Anomalous //8 //8 B //8 9//8 EQ- 6//8 //8 8//8 //8 //8 Anomalous 7//8 //8 //8 6//8 9//8 //8 7 C //8 >σ B Amplitude No. Of s No. Of s Amplitude No. Of s 8//8 //8 EQ- 6 8 ime (Seconds) Anomalous //8 7//8 //8 6//8. variation amplitudes VX as a 9//8 6 8 amplitudes VX: 6 as a funcfunction variation time seconds for consecutive. ime (Seconds) 8//8 tion time seconds for consecutive :an6amplitude //8 s stacked after shift to 6 units. 8. s stacked afteran amplitude shift pots,, denoted one A B C 6//8 units. pots A denoted. one C, B = = On. variation amplitudes VX as a funca B C =C 8 Aoccurred = 6. On nd B 6.. 8, earthquake M data on, tion an time seconds for consecutive ime (Seconds) : 6 8 8, an earthquake occurred magnitude 6.. data st to 6 s clearly stacked after anthat amplitude shift on plotted 8. a thick le, on 8,, plotted a thick le, clearly that on st units. pots A C, anomalously B denoted one higher those 8, variation anomalously higher = = before. On nd C A Bthat VX or after to 6those.. also amplitudes as a func or for also that 8, earthquake occurred magnitude 6.. data st to earthquake day value before isonafter tion antime. seconds consecutive : earthquake day after value that, plotted a thick clearly on st to 8. s le, stacked an amplitude shift is 8, anomalously higher those units. pots A, denoted one C B or. also that before after to = C A =B. On nd In 6, we mutes forst all 6 earthquake day value 6.. is 8, an earthquakeplot occurred magnitude data on strong plotted havg a magnitude Mon>st (upper, a thick le, clearly that two 8, first anomalously higher thoseonly panels) which belong to three seismic circles or. also before afterin all to (i.e., a zone radius km that centre at FRP). 6 earthquake day value is cases, take place on day. It is clear that generally higher prior to seismic events not after m. In fact, if we take simple averages, we fd (third panel) that average is peaked two prior to peak for one day prior to peak for. error-bars stard deviations obtaed on each day. As Paper, we plot 7a b, cross-correlations between or effective magnitudes which takes place on day. Earthquakes all effective magnitudes were taken this graph. In 7a, peak occurred one day prior to earthquake re also smaller peaks. In 7b, we note a quite broad peak, though it also occurred one day prior to event day. In terms predictability, we fd that termator shift approach (Paper ) gives dications possible seismic events earlier or approach. In 8a d, we plot a similar result as 7a b, takg depths seismic events to consideration. In (c) we plot correlation coefficients for those quakes havg shallow depths (d < km) (d) we consider those quakes havg deeper depths (d > km). We generally fd that peak is sharper for shallower quakes.

5 7. cross-correlations between (left pane) or S. K. Chakrabarti et al.: Ionospheric anomaly due to 7 (Mutes) (Mutes), (Mutes) Correlation Coeffcient ime (s) Lag/Lea (c) km < km ime (s) Lag/Lea 6. variation as a function for a period around seismic events. first panel variation second panel variation. zero X-axis dicates day havg M >. third panel 6. variation average variation (filled circles) as a function (filled squs) obtaed from 8. first (a-d): two cro panels stard deviations as error bars for those. It is clear from third panel that durg, (c-d) value eff is maximum onfor twoa period before around value seismic events. is maximum onfirst one day panel before. before after variation second panel results for earthquak variation. zero X-axis dicates day.. Concludg remarks less km havg M >. third panel average variation (filled circles) It is long conjectured (filled that ionospheric squs) anomalies could have quakes havg depths o.8 all cases, peak o obtaed from first two panels beenstard detected prior deviations to an earthquake. as error A number groups. have been workg on this problem for last two decades shallow bars for those. It is clear from third panel that.6 some evidences have been found. Sce no Indian or subcontent is is maximum also vulnerable on two to severe, we have( < d < strong peak durg, value. before valuebeen systematically isrecordg maximumon s over also. last few. one day before. years to eir establish or to refute conjecture. In Paper present paper, we have used only VX data as.. received from Kolkata. In Paper, we concentrated on calibration sunrise sunset termator times over whole year so that anomalies may be studied 6 easily. his was. also possible because durg 8, sun was particularly quiet re was no disturbances due to solar 7. cross-correlations between or activities. re we showed that re is possibly a distct.9.9 effective earthquake magnitude occurs plotted as signature anomaly termator timgs anomaly a function before.8 after event ( th day). In, is seen almost 8 h prior to seismic events. In present.8 peak occurs one day prior to seismic event but effect contues even after event. In, peak is fairly broad, appears -layer preparation time () -layer disappearance paper, we chose an alternate measurable quantity, namely.7.7 at around. day prior to event..6.6 time (). We successfully demonstrated that more energy released due to seismic events on agiven day is,.. more is deviation from mean. In 9, we plot effective magnitude as a function stard deviation from mean. predictability is poorer. We showed that both.. However, correlation is not very tight as a result, correlation is generally. lear, i.e., on a given day,., cross-correlation is peaked only about a day prior effective magnitude or energy deposited associated.. to seismic event. We used both effective magnitude earthquake is directly related to deviation where we add energy released from smallquakes also,.. (left) (right) Effective Magnitude Earthquake Stard eviation Nat. Hazards Earth Syst. Sci.,, 7 77,

6 76 S. K. Chakrabarti et al.: Ionospheric anomaly due to Correlation Coeffcient (c) km < d < km km < d < km (d) km < d < km km < d < km cross-correlations between (a b) or (c d) effective earthquake magnitudes as a function before after event ( th day). (c) show results for for which depths (d) epicenter less km (d) show results havg depths epicenters between km. In all cases, peak occurs day prior to seismic event. For shallow (d < km) peak is sharper re no or strong peaks before after event. For deeper ( < d < ) re smaller peaks at or also. Effective Magnitude Earthquake 6 Effective Magnitude Earthquake 6 absolute magnitudes showed that both cases pre-cursors present. What is more, we found that for shallower (d < km) correlation peaks sharper quakes which occur at a higher depths (d > km). Although VX-ICSP basele might have exhibited a correlation, predictability actual event location is still not possible. We conjecture that if we carry out such observations from a multiple number receivg stations, n basele exhibitg a tighter correlation is affected more by seismic events. hus by takg data multiple stations we will possibly be a position to locate region seismic activities well ahead time. 6 Stard eviation 6 7 Stard eviation Acknowledgements. his project is supported by a RESPON grant from ISRO. 9. Effective magnitudes as a function deviations from mean (left) (right). correlation is generally lear, i.e., effective magnitude or energy deposited associated earthquake is directly related to observed deviation. Edited by: M. E. Contadakis Reviewed by: three anonymous referees Nat. Hazards Earth Syst. Sci.,, 7 77,

7 S. K. Chakrabarti et al.: Ionospheric anomaly due to 77 References Chakrabarti, S., Sasmal, S., Saha, M., Khan, R., Bhowmik,., Chakrabarti, S. K.: Unusual behaviour -region Ionization time at 8. KHz durg seismically active, Indian J. Phys., 8(, 6), 8, 7. Clilverd, M. A., Rodger, C. J., homson, N. R.: Investigatg seismoionospheric effects on a long subionospheric path, J. Geophys. Res., (A), , 999a. Clilverd, M. A., Rodger, C. J., homson, N. R.: Sunrise effects on VLF s propagatg over a long north-south path, Radio Sci., (), 99 98, 999b. Gokhberg et al.: Phys. Earth Planet. Inter., 7, 6-67, 989. Gufeld, I. L., Rozhnoi, A. A., yumensev, S. N., et al.: Radiowave disturbances period to Rudber Rachsk, Phys. Solid Earth, 8(), 67 7, 99. Lowrie, W.: Fundamentals Geophysics, Cambridge University Press, UK, 7. Hayakawa, M., Molchanov, O. A., Shima, N., Shvets, A. V., Yamamoto, N.: Seismo Electromagnetics: Lithosphere Atmosphere-Ionosphere Couplgs, edited by: Hayakawa, M. Molchanov, O. A., ERRAPUB, okyo, Japan, pp.,. Maekawa, S., Horie,., Yamauchi,., Sawaya,., Ishikawa, M., Hayakawa, M., Sasaki, H.: A statistical study on effect on ionosphere, based on subionospheric LF propagation data Japan, Ann. Geophys.,, 9, doi:.9/angeo--9-6, 6. Molchanov, O. A. Hayakawa, M.: Subionospheric VLF perturbations possibly related to, J. Geophys. Res,, 789 7, 998. Molchanov, O. A., Hayakawa, M., Ondoh,., Kawai, E.: Precursory effects subionospheric VLF Signals for Kobe earthquake, Phys. Earth Planet. In.,, 9 8, 998. Rodger, C. J., Clilverd, M. A., homson, N. R.: Modelg subionospheric VLF perturbations associated, Radio Sci., (), 77 8, 999. Rozhnoi, A., Solovieva, M., Molchanov, O., Schwgenschuh, K., Boudjada, M., Biagi, P. F., Maggipto,., Castellana, L., Ermi, A., Hayakawa, M.: Anomalies VLF radio s prior Abruzzo earthquake (M=6.) on 6 April 9, Nat. Hazards Earth Syst. Sci., 9, 77 7, doi:.9/nhess , 9. Sasmal, S. Chakrabarti, S. K.: Ionosperic anomaly due to seismic activities - Part : Calibration VLF VX 8. KHz station from Kolkata deviation durg seismic events, Nat. Hazards Earth Syst. Sci., 9, 8, doi:.9/nhess-9--9, 9. Soloviev, O. V. Hayakawa, M.: hree-dimensional subionospheric VLF field diffraction by a truncated highly conductg cylder its application to rimpi effect problem, Radio Sci., 7, 79, doi:.9/rs99,. Soloviev, O. V., Hayakawa, M., Ivanov, V. I., Molchanov, O. A.: Seismo-electromagnetic phenomenon atmosphere terms subionospheric radio wave propagation problem, Phys. Chem. Earth, 9, 69 67,. Nat. Hazards Earth Syst. Sci.,, 7 77,