Simulation of ionosonde observations of ionospheric holes

Journalo/Afmospheric

and Terrestrral

Printed inGreatBntain.

Physics, Vol

54. No. 9, pp

1177-l

183, 1992.

0021~9169/92 %5.00+.oO

Pergamon PressLtd

Simulation of ionosonde observations of ionospheric holes A. D. KALIKHMAN,N. N. KLIMOV,G. K. MATAFONOVand A. V. TASHCHILIN Siberian Institute of Terrestrial Magnetism, Ionosphere and Radio Wave Propagation (SibIZMIR), Russian Academy of Sciences, Irkutsk 33, P.O. Box 4026, 664033 Russia (Received infinalform 3 October 1991 ; accepted 10 October 1991) Abstract-When conducting active experiments in the ionosphere which give rise to artificial disturbances of electron density, it is necessary to diagnose such disturbances. The study of irregularities using groundbased radio sounding methods often involves recording and analyzing ionogram traces, depending on the position, dynamics and phase of development of an artificial disturbance. Within the context of solving the problem of diagnostics, we offer a numerical model that includes blocks for calculating disturbance parameters and parameters of the recorded signal of ground-based radio sounding of the disturbance. By considering an example of the release of a plasma neutralizing agent, we study successive evolutionary stages of the disturbance and their reflection in the behaviour of additional traces on complex ionograms.

1. INTRODUCTION

The artificial ionospheric disturbance analyzed here is the effect of the release of a chemical from a spaceActive experiments in the ionosphere, which give rise craft at ionospheric heights, which leads to the forto artificial disturbances of electron density, are usumation of an extended region of decreased electron ally associated with the injection of chemical agents density, the so-called ionospheric hole (MENDILLO et or with the heating of the ionosphere by high power al., 1975, 1980; WAND and MENDILLO, 1984; MENradio waves. They often involve identifying the resultDILLO and BAUMGARDNER, 1982; ZINN et al., 1982). ing irregularities using different methods. These The modelling results on the hole are initial ones for include ionosonde observations (PAUL et al., 1968), calculating ray paths of the radio sounding signals radio probing of the disturbance region from satellites and for ray-tracing synthesis of ionograms at different (MENDILLOet a/., 1975,1980), direct measurements by stages of development of a disturbance, up to its the incoherent-scatter method (WANDand MENDILLO, maximum phase. We have identified the main sig1984), and optical observations (MENDILLO and natures of the occurrence of a depleted ionization BAUMGARDNER, 1982). The method of ground-based region on the basis of additional traces on ionograms, radio sounding involves recording and analyzing with an ambiguous dependence of the virtual height ionogram traces, depending on the position, dynamics on the sounding frequency. and development phase of the artificial disturbance. A successful implementation of the method of rayHowever, the overall resolving ability of the method tracing synthesis of ionograms to diagnose ionois limited by the uncertainty of the disturbance parspheric holes is determined by using a new approach ameters in the process of its evolution. One of the to numerical modeling of multipath ionograms (VARsolutions of the problem of diagnosing such an SHAVSKY and KALIKHMAN, 1988). The method of rayinhomogeneous structure is considered in this paper tracing synthesis was first used for studying ionoin the form of a numerical model. This includes sucspheric holes by PAUL et al. (1968) and later by HELMS cessive blocks for calculating undisturbed ionospheric and THOMPSON (1973) for investigating a large moving parameters and characteristics of the disturbance and ionisation trough, and by LOBB and TITHERIDGE of the recorded signal of ground-based radio sounding (1977) for studying the effects of travelling ionoof the disturbed region. Unlike earlier numerical spheric disturbances on ionograms. In the cited papers models which determine the effects of the release of the magnetic field has been neglected to simplify the chemically active agents in the F-region (BERNHARDT ray-tracing procedure, which limits the applicability et al., 1975 ; MENDILLO and FORBES,1978 ; ANDERSON and BERNHARDT, 1978; ZINN and SUTHERLAND, 1980), of such modelling. ionospheric parameters are calculated along field lines 2. A SIMULATION OF AN ARTIFICIALLY CREATED from 150 km height in one hemisphere to 150 km IONOSPHERIC HOLE height in the other hemisphere. This makes it possible In experiments on the creation of ionospheric holes, to calculate injection effects in the conjugate hemichemically active agents such as H,, H,O and CO, sphere, though in this paper they are not considered. 1177

A. D. KALIKHMAN rr al.

1178

capable of decreasing the electron density are injected into the ionosphere. As the hole is produced, O+ ions transform to molecular ions which rapidly recombine with electrons. In general this process can be represented as : O’-tM,,*M,:+N ML +e

3 neutrals + emission,

(1) (2)

where M, = H,, Hz0 and CO1 are the injected molecules, M,’ = OH+, H,O+ and 0: are the molecular ions being produced, and N,, = H, 0 and CO are neutral particles. As a result of reactions (1) and (2), there occurs a rapid decrease of the number of electron-ion pairs in the injection region because the rate of reaction (1) exceeds by many times that of the ionmolecular reactions taking place in the undisturbed ionosphere, with the N> and O2 neutrals. Let us consider a particular example of the release of 50 kg of Hz vapour to produce an ionospheric hole. The reactions (l)-(2) corresponding to this case have the form : H2+OC *OH++H, OH’+e~~~H+O,

y* = 2. IO-‘cm’*ss

(3)

a* = 2*10m7cm3*s ~I.

(4)

The ion-molecular reaction (3) proceeds more rapidly as compared with the usual exchange reaction of Of with Nz, by a factor of about 1000, which ultimately gives rise to the formation of an ionospheric hole. In the diffusion approximation the expansion process of a cloud of injected molecules obeys the equation (BERNHARDT. 1979):

(5) where N is the particle density in the cloud ; D is the diffusion coefficient; S is the source function ; H, = kT/m,g is the scale height of the injected gas ; k is Boltzmann’s constant; m, is the mass of injected molecules ; and g is gravitational acceleration. We consider the height range h > 15&200 km, in which the atmosphere is isothermal, which permits us to neglect here the thermal diffusion process. Source S on the right-hand side of (5) represents the number of particles injected at a given point of space per unit time. Injection is usually carried out from satellites or rockets, by exploding containers containing a suitably chosen explosive, or by the operation of rocket engines. In all these cases the typical size of the region, in which injection is carried out, is significantly smaller than that of the resulting cloud and the scale height

of the atmosphere. This circumstance enables us to consider the source of particles to be a (fixed or moving) point. For determining the density of charged particles in the injection region, a system of equations modelling the behaviour of the ionospheric plasma in the tubes of force of the dipole geomagnetic field passing through the injection region was solved numerically simultaneously with equation (5). It was assumed that the ionopshere consists ofelectrons and H+. O+, NO+ and N: ions, whose state is described by the equations of continuity and motion : (6)

= (V;- U,,)c “Vi+ 1 (K - I/,)“$ n /#l

(7)

Here 0 is the flux tube cross-section; V, is the velocity of an ion of the ith sort along the field line ; qi and r, are the production rate and the lifetime of the ith ion, respectively ; k is Boltzmann’s constant ; m, is the ion mass ; T, and T, are the temperatures of the ions and electrons, respectively; G,, and U,, are the projections onto the field line of the gravitational acceleration and of the horizontal velocity of neutral wind, respectively ; v,,,and vz are the collision frequencies of ions with neutral particles and with each other, respectively. Their values and temperature dependences are given in the Appendix. For the undisturbed (without injection of HJ ionosphere, the model includes the following chemical reactions (on the right, the reaction rate constants are given in cm3 s ‘) : O++H+H++O H++O+O++H o+ +02 -0: $0 O++Nz=NO++N (NO’,N:,O$)+e-

y, =2.5*10-“JT, ?r’2= 2.3. IO- ” Jr, y, = 1.1 * IO-9 T,-O-9 y,=3.6*10-‘“T;’

aM,,+N, a = 9.0*10~5T,-‘. (8)

For disturbed conditions, reactions (3) and (4) are also taken into account. It should be noted that since we are using the model equations (6) and (7) for heights h 2 150 km where the ionospheric plasma is fully magnetized, the horizontal neutral wind will drive ions only along geomagnetic field lines. Therefore, the effects associated with the polarization of the ionospheric hole are neglected in the input equations (6) and (7).

Ionosonde observations of ionospheric holes

i -..l-. ‘“_f‘

1

r

c

c

Ionosonde observations of ionospheric holes

1181

of the sounding frequencies which fall on the emitting point after being reflected. In the general case in a three-dimensional inhomogeneous medium, ray paths are spatial curves such that it is difficult to determine return rays. In a particular case when the magnetic field vector, the electron density gradient and the wave normal lie in the same plane, the ray paths will lie in the same plane and, consequently, calculation of the return rays is simplified (VARSHAVSKY and KALIKHMAN,1988). We analyze three moments of time in the growth stage of the ionospheric hole. They correspond to the where rc, and JC~are the heat conduction coefficients initial, middle and maximum stage of development of electrons and ions (their values are given in the of an artificial disturbance in the ionosphere which Appendix) ; Qeis the heating rate of the electron gas by assumes the form of a rather smooth large-scale photoelectrons ; and L, is the rate of energy exchange irregularity. Figure la, b and c presents the ion density between particles of different types, with indices 0 distributions in the magnetic meridional plane (the and * corresponding to elastic and inelastic collisions, positive direction of the horizontal axis is southward) respectively. The temperature dependences of the for the above geophysical conditions ; they corcooling rates L, for elastic and inelastic collisions respond to times of 107, 324 and 720 s after the injecwith neutral particles are given by SCHUNKand NAGY tion. Electron density contours are calculated in (1978). The technique of computation of the heating plasma frequency units (MHz) and are drawn with rate Q. is described in detail by KRINBERGand TASH- the step 0.5 MHz. The figure also gives the ray paths CHILIN(1980, 1984). of the radio wave which permit us to determine the The system of equations (5)-(10) was solved consistency of the individual families of ray paths to numerically for a set of 5 x 5 = 2.5 field lines, whose additional traces on ionograms. foot points are separated from each other by 1” in The origin of the horizontal coordinate coincides latitude and longitude, which approximately corwith the injection point, and the ground-based ionoresponds to a distance of 100 km. The method of sonde is located at the distance of 50 km northward. numerical computation of hydrodynamical equations Figure 2a, b and c presents the synthesized ionograms. (67,9-10) was described and discussed in KRINBERG The ionogram traces and ray paths of the radio wave and TASHCHILIN(1984). Injection of H, was carried on the ion density distributions correspond to an out at the 250 km height at a point with geographical ordinary wave. The ray paths and ionogram traces coordinates 4 = 52”N, i = 104”E at time t = 06.00 belonging to the same family are marked by the same LT at a point inside the region occupied by the field colour. lines considered. The densities of the background neuIn the first instance it should be noted that the tral particles needed for solving the system of equamain indicator of a region of decreased ionization, an tions (5)-(10) were calculated using the MSIS-86 ionospheric hole, is the presence of multipath ionmodel (HEDIN, 1987). ograms with an ambiguous frequency-dependence of As a result of the model calculations we obtained the virtual height. The so-called main trace (brown) a set of height-latitude distributions of the electron on the ionogram is associated with the background density representing the background (undisturbed) ionization distribution. It refers to northward deflectionosphere and the ionospheric hole in different ing rays (brown), for which the angular deviation phases of its time evolution. from the vertical increases with frequency and which pass over the northern edge of the disturbance region. It is evident that, in the initial stage of development of the disturbance, the main trace (brown) more closely 3. RESULTSOF RAY-TRACING SYNTHESIS corresponds to the background ionosphere, and the OF IONOGRAMS trace cutoff is observed at times of 324 and 720 s at frequencies of 4.2 and 3.4 MHz, respectively. Such a The method of ray-tracing synthesis of the ionograms used here involves determining the ray path of behaviour of the main trace characterizes an increase radio waves for the electron density distributions with of the spatial size of the disturbance within the ionoan ionospheric hole obtained in a previous stage of sonde antenna beam. That the opposite southern edge mu also lies within the beam is simulation. We calculate only return ray paths at each of the disturbance region Electron and ion temperatures are determined by solving the self-consistent, time-dependent equations of energy balance :

1182

A. D. KALIKHMANet al.

clearly observed in the green trace and related rays (green) which remain even in the maximum expansion phase at time 720 s. The main feature of ionospheric hole manifestations on the ionograms, however, are hooked traces. Such traces (red, orange and pink) for time 107 s correspond to trajectories refracting on the periphery (pink trace) and at the centre (red and orange traces) of the hole, which leads to their relative instability and to abrupt increases of the group path. Violet traces with large group paths and a weak dependence on frequency refer to characteristic ones. Loop-like violet rays refracting on the edges of the zone of decreased ionization correspond to them. The appearance of multiple hooked traces on ionograms in a limited frequency range, which differ from each other by the group path, are usually recorded in experiments on the creation of an ionospheric hole in the stage of development of a disturbance (PAUL et al., 1968). The spatial size of the irregularity is estimated rather roughly and exceeds 50 km in the case of the appearance of hooked traces. Well-defined hooks also remain on the ionogram for time 324 s, while at the maximum stage of development of the ionospheric hole at 720 s, they are observed only at considerable group ranges. This is attributable to the form of the violet rays experiencing double reflection within the disturbance region. Additional traces on the ionogram for time 324 s in the middle stage of development of the ionospheric

hole also give an estimate of the threshold value of minimum ionization in the hole, of order 2 MHz; however, the strong distortion of the main trace (brown) produced by markedly deviating (from the vertical) ray paths indicates that the disturbance is located almost over the point of radio sounding. This is still more evident on the ionogram for time 720 s. 4. CONCLUSIONS

In this paper we have offered a solution of the problem of diagnosing an artificial ionospheric disturbance such as the ionospheric hole in the form of a numerical model This successively includes blocks for calculating the background ionosphere, disturbance parameters and characteristics of the radiosonde-recorded signals of ground-based sounding of a disturbance. We have determined the main signatures of the manifestation of disturbances as well as regularities of the variation of additional traces on ionograms, depending on the position and phase of development of the artificial ionospheric hole. It has been shown that, in order to achieve a confident recording of such a disturbance, the range of the ionosonde from the disturbance centre must not exceed the supposed disturbance size. The proposed numerical model may be useful for a preliminary analysis of effects of the release of plasma neutralizing chemicals at ionospheric heights and for predicting the results of planned active experiments.

REFERENCES ANDERSON D. N. and BERNHARDT P. A. BERNHARDT P. A. BERNHARDT P. A., PARK C. G. and BANKSP. M. HEDINA. E. HELMSW. J. and THOMPSON A. D. KRINBERGI. A.

1978 1979 1975 1987 1973 1971

KRINBERCI. A. and TASHCHILIN A. V. KRINBERGI. A. and TASHCHILIN A. V. L~BB R. J. and TITHERIDGE J. E. MENDILLOM. and BAUMGARDNER J. MENDILLOM. and FORBESJ. M. MENDILLOM., HAWKINSM. and KLOBUCHARJ. A. MENDILLOM., ROTE M. D. and BERNHARDT P. A. PAUL A. K., SMITHG. G. and WRIGHTJ. B. SCHUNKR. W. and NAGY A. F. VARSHAVSKY I. 1. and KALIKHMAN A. D. WANDR. H. and MENDILLOM. ZINNJ. and SUTHERLAND C. D. ZINNJ., SUTHERLAND C. D., STONES., DUNCANL. and BENKER.

1980 1984 1977 1982 1978 1975 1980 1968 1978 1988 1984 1980 1982

J. geophys. Res. 83,4777. J. geophys. Res. 84, 793. Geophys. Res. Leti. 2, 341. J. geophys. Res. $4649. Radio Sci. 8, 1125. Issled. PO Geomagnet. Aeronomii i jizike Solntza 16, 115. Ann. Geophys. 36,537. Ionosphere and Plasmasphere, p. 186. Nauka, Moscow. J. atmos. terr. Phys. 39, 129. Geophys. Res. Lett. 9, 215. J. geophys. Res. 83, 15 1. J. geophys. Res. 80,2217. EOS Trans. AGU 61, 529. Radio Sci. 3, 15. Rev. Geophys. Space Phys. 16,355, izv. Vuzov. RadioJizika 31, 7, 869. J. geophys. Res. 89,203. Space Solar Power Rev. 1, 109. J. atmos. ferr. Phys. 44, 1143.

I183

Ionosonde observations of ionospheric holes viz= v(M+,O) = 4.8.IO-'*[O]

APPENDIX The collision frequencies (s- ‘) (KKINBERG,1971) are: V

1,

=

VW+,H) = 4.2W’0fii ~(I-o.ls~lg(T,+T”)~[w]

V

,* = v(H+,O) = 2.2‘10-‘~[0]

v:2=V(M+

= 8.7*10-“‘[N*]

0+)=2.7*10+In* ~--~0”1 pz

I

lnit = 9.6+11n r,-- iinn,.

vz2 = v(O+ ,O) = 4.6. IO-” ,/r,tm ‘(1 - 0.15-k(C)‘Pl vz3 = “(O+,NJ

vl> = v(M+,NJ

The coefficients of thermal conductivity cm ‘*SC’) (KRINBERG.1971) are:

= 6.8. IO-‘“*[NJ “t=1+~fO-i2

(eV * K .- ’ *

7.7. 105T
K,= 1.2. IO'T-: *(4[H+]+[O+ J)/n,.