Pages 1083-1095

ESTIMATION OF ELECTRON DENSITY OF IONOSPHERIC PLASMA USING WAVE, IMPEDANCE AND TOPSIDE SOUNDER DATA

A.Kiraga1, Z.Klos1, H.Rothkaehl1, Z.Zbyszynski1, V.N. Oraevsky 2, S.A.Pulinets2, and I.S.Prutenski2

1Space Research Centre, PAS, ul.Bartycka 18A, 00-716 Warsaw, Poland, E-mail:   kiraga@cbk.waw.pl
2IZMIRAN,Troitsk, 142092, Russia 

ABSTRACT 

Instability analysis of the dispersion relation of electron plasma indicates that enhanced emission in the frequency band (fp,fu=sqrt(fp*fp+fc*fc) can be easily detected in wave spectra of space plasmas. Such emission, in passive mode spectra, can be used to determine plasma density of the major cold plasma component and points out the existence of a minor energetic component. Contrary to such expectations, it was shown that in passive measurements performed on the INTERCOSMOS19, ACTIVE and APEX low-altitude satellites, the most prominent and persistent spectral structure was centered around frequency fr fulfilling the relation fc<fr<fp and corresponded to resonant properties of the equivalent input network. The input network mismatch at frequencies around fp significantly depressed natural plasma noise and made it hardly discernible. Plasma emissions in the band (fp,fu) were prominent when propagation properties of electromagnetic excitation were dominating (topside sounder) or if the excitation introduced a sufficient amount of nonequilibrium components into the plasma in the form of an injected electron beam. The wave, impedance and topside sounder data, registered on the recently launched CORONAS low-altitude satellite, support and supplement previous observations and interpretation. Analysis of pertinent CORONAS data in various geophysical regions is presented in this paper. Consistency of wave and impedance experimental spectra with their equivalent circuit simulations and topside sounder data is obtained. Data and simulations argue against permanent observability of the upper hybrid band in ionospheric plasma. 

INTRODUCTION 

One of the key parameters in space plasma study is electron plasma density. In particular it is necessary for the identification of plasma emission modes. Since the first report of Walsh et al. (1964) on naturally occurring noise bands close to the electron gyromagnetic fc and plasma fp frequencies in the topside ionosphere, passive in-situ measurements of AC electric fields have been used to determine fp. In early research on low-altitude satellites, determination relied on the assumption that some prominent, variable frequency enhancement in spectra (Smith 1975) or transient enhancements in fixed frequency measurements (Burenin et al. (1968), Efimova et al. (1976) occur in the upper hybrid band (fp,fu=sqrt(fc*fc+fp*fp)). Both the first report and significant theoretical effort (Fejer and Kan, 1968; Basu at al., 1982) in the study of the susceptibility of cold plasma to nonequilibrium minor components bolstered such identification. Simultaneously measured or presumably present fluxes of precipitating auroral or radiation belts' electrons or even photoelectrons arriving from the conjugate sunlit hemisphere have been invoked as a free energy source, dependent on geographical positions of observation.  

To explain the permanent occurrence of the enhanced band in equatorial plasmasphere at all local times, the processes triggered by transport of the dawn-to-dusk electric field crossing the plasmasphere have been suggested by Oya et al. (1991).  

The unambiguous observation of enhanced noise below the Z mode cutoff on the Alouette and ISIS satellites (Hartz 1969) received much less attention from the scientific community because of its local origin, for example, the interaction of the satellite itself and the medium.  

Two most prominent noise bands controlled by local plasma frequency but with maxima definitely downshifted from fp were permanently present in wave spectra on the INTERCOSMOS19 satellite. They were explained (Kiraga 1986) as the resonances of equivalent circuits composed from monopoles imbedded into ambient and wake plasmas and their preamplifiers. The plasma frequency was determined from the onboard topside sounder data.  

The equivalent circuit approach to the analysis of passive mode wave data had been applied in the ACTIVE, APEX and VC36.064CE missions (Kiraga et al. 1995a). The erratic enhancement in the upper hybrid band was occasionally registered but without a convincing relation to previously reported regions or mechanisms. Matching of wave and impedance data with the aid of equivalent circuits resulted in a unified method of preliminary analysis of both types of data (Kiraga 1995b). The notion of controlled and uncontrolled measurements were introduced (Kiraga 1996) to differentiate between normal instrument operation and the situation when several crucially important circuit elements are unknown. The data registered onboard the CORONAS satellite are very suitable to further explore essential problems in the determination of local density from the electromagnetic wave spectra measured in cold plasma. It is shown that the almost permanent noise band which follows plasma frequency corresponds to equivalent circuit resonance and is downshifted from fp for weakly and moderately magnetized plasma (fp>2fc). The relation between the spectral structure of the resonance band and equivalent circuit elements is discussed. 

DESCRIPTION OF EXPERIMENT 

The CORONAS satellite was launched on 3 March 1994 into a circular orbit with an inclination of 82.5° and an altitude of 500km. The satellite was actively stabilized along three axis, with one of the axis pointing to the sun with an accuracy better than 10'. The exact attitude of the other axis is not available in this paper. The processor-controlled HF wave instrument PRS4 was able to measure wave spectra in the frequency range .1-30MHz, monopole self impedance in frequency range .1-10MHz and to operate as a topside sounder in the frequency range .3-16MHz. The main objective of the mission was the direct observation of processes on the sun and this objective required low data rate, prolonged telemetry sessions. For this reason and due to limited life time, PRS4 programs requiring high data rates and dedicated to the more advanced study of wave processes in ionospheric plasma were not realized.  

In low data rate telemetry, PRS4 measured almost instantaneously either background spectra and impedance or background spectra and sounder response. The 7.5m long impedance sensor was mounted on the shadowed satellite surface and was antiparallel to the sun direction. Two mutually perpendicular dipoles were mounted in the plane perpendicular to the sun direction for separate use by the receiver and the transmitter. All booms had the same length of 7.5m and a diameter of ~0.03m. The same swept frequency receiver with a bandwidth of 25kHz, sensitivity of 1microV and the time constant of 80micros was used to provide background spectra and topside sounder data. At each programmed frequency the receiver output was sampled every 67µs and an average of sixteen consecutive recordings was telemetered as a background amplitude. In the topside sounder mode the background amplitude was measured just prior to the sounder pulse. One hundred samples were registered after each pulse but only time delays of the three largest amplitude samples were telemetered in slow telemetry mode. One reduced resolution background spectrum and one ionogram were registered in 2.4s every 87s. Self-impedance measurements were performed in a voltage divider configuration. Using a constant voltage (rms amplitude of 100mV), a stepped frequency generator fed the monopole antenna through the fixed resistor. Only the voltage measured at the base of the monopole (the base voltage) was telemetered in the data presented in this paper. The single base voltage spectrum was measured in 1s with an equally spaced mesh of 200 frequencies. Just prior to impedance measurement, the single background spectrum was measured in 1.2s with a mesh of 250 frequencies in the band of .1-10MHz and a mesh 150 frequencies in the band of 10-30MHz. The measurements were repeated every 34s. If plasma parameters change sufficiently slowly, the measurements performed within 3s may be called simultaneous. 

PHYSICAL ORIGIN OF ENHANCED BAND 

Experimental Data

Dynamic HF background spectra and base voltage spectra are presented in the upper panel and lower panel respectively in Figure 1 for three consecutive revolutions. The frequency expressed in MHz changes along vertical axis and the earliest spectrum is placed at the left. Only satellite longitude (LON), magnetic local time (MLT) and universal time (UT) values at the geographic equator crossings are given with tick marks to present the geophysical environment. In the upper panel the local gyrofrequency is plotted with a thick white line. The signal intensity is proportional to the darkness. The amplitude above 200µV is shown black. The amplitude below 25µV is shown white. The white patches occupy a significant part of the dynamic spectrum pointing out moderate satellite electromagnetic interference for frequencies above ~2MHz. Intense constant frequency bands correspond to the reception of ground-based transmitters and map the global distribution of critical frequency beneath the satellite. It can be easily anticipated that intense, narrow-band-enhancement, variable frequency (fr) is related to the local plasma frequency, fp. Only on parts of the orbits close to the polar regions does its frequency fr strongly fluctuate or the enhancement merge with intense noise below 2MHz. There is a remarkable correlation in the variation of frequency, fr, along the orbit and that corresponding to the maximum base voltage in the lower panel (FR). However, a detailed examination of data, coupled with the simulation of the instrument performance reveals that the band centered around fr does not correspond to naturally excited upper hybrid emission for most of the PRS4 data. In order to minimize effects resulting from plasma depletion by the satellite body and solar panels and from corrections due to the sheath and magnetoplasma directivity, the data registered in dense plasma when the satellite has a velocity component away from the sun are analyzed.

Fig. 1. Background spectra (upper panel) and base voltage spectra (lower panel).

 

Fig. 2. Background spectrum (upper panel) and topside sounder data (lower panel).

An example of simultaneous measurement of background spectrum and topside sounder data is presented in Figure 2. With an assumption that the X mode cut off occurs at the frequency marked with X, the plasma frequency (marked with N) and Z mode cut off frequency (marked with Z) were calculated. With such an assignment, the greater part of strong broadband enhancement is below the Z mode cut off and there is not an enhanced signal in the upper hybrid band. Simultaneous background and base voltage measurements which were performed in the same region twenty-four hours later are presented in Figure 3. The maximum of the broad band enhancement in the spectrum centered around fr=7MHz is slightly downshifted from the frequency FR of the base voltage maximum, and there is not a convincing emission in the upper hybrid band at frequencies close to the small secondary maximum of the base voltage at 7.8MHz. Figures 2 and 3 show unambiguously that the frequency fr is definitely less than fp so the enhancement around fr can not be an upper hybrid mode of ambient plasma. To give further insight into the nature of the band in question, selected pairs of simultaneous measurements of noise and impedance are presented in Figures 4 and 5. Data were selected in order to get the closest similarity of enhanced background amplitudes around some fr. Pairs from sunlit ionospheric regions indicated with empty triangles in Figure 1 and three revolutions later are presented in Figure 4. Almost an exact overlap of both sets of data provides confidence in the instrument and satellite operation and proves that wake, attitude and sheath effects have negligible influence on spectral structures around the main maxima. The data selected from the midnight and noon equatorial ionospheres show remarkable similarity of background spectra but significant differences in base voltage spectra for frequencies less than FR. The differences in base voltage spectra are easily acceptable because they occur in the frequency range sensitive to the sheath thickness, so they manifest the difference in ambient plasma temperatures. However no reasonable arguments, based on geophysical processes, can explain the identical structure of pronounced enhancements around fr if they are assigned to upper hybrid emission. In both Figures 4 and 5, the frequency fr is downshifted from the frequency FR at which the maximum of base voltage occurs, so the enhancement in question is below fp and it is not justified to assign it to upper hybrid emission.

Fig. 3. Background and base voltage spectra registered in noon equatorial ionosphere.

 Fig. 4. Three pairs of background and base voltage spectra from a noon equatorial ionosphere.

Fig. 5. Two pairs of background and base voltage spectra registered in a noon and a midnight equatorial ionosphere.

Equivalent Circuit Simulations 

Simulation of equivalent circuits pertinent to background and base voltage spectra provides consistent explanation of interrelations between fr, FR and fp. The main assumptions and identification of crucial parameters were presented in (Kiraga et al.(1995a), Kiraga (1995b)). The inductive character of the monopole impedance in the frequency band (fc,fu) is a common ingredient in both equivalent circuits and is essential for their properties. For a fixed monopole impedance, the transfer function which relates the a.c. voltage externally imposed on the monopole (e.m.f.) to the receiver input depends on the detailed structure of the input preamplifier. For fp and fc pertinent to Figure 3, the transfer functions calculated for the preamplifiers flown on ACTIVE and CORONAS satellites are presented in Figure 6. G,N,Z,X mark the fc, fp and Z and X cut offs. Functions are marked with c and C for ACTIVE and with h and H for CORONAS. Upper case letters correspond to total stray capacitance of 30pF and lower case letters to those of 20pF. All transfer functions reveal strong underestimation of the e.m.f. in the vicinity of the upper hybrid band. At lower frequencies ACTIVE transfer functions show a very broadband but weak overestimation of the e.m.f. Differences in stray capacitances are reflected in small amplitude, broadband changes. Properties of the CORONAS transfer function below fp are dramatically different. There exists a strong overestimation of the e.m.f. in a narrow band around some resonant frequency fr. The resonant frequency itself is determined by a fine competition between the inductive and capacitive elements of the equivalent circuit. All curves in Figure 6 were computed using the formula for monopole impedance in a homogeneous cold plasma (Balmain 1964). The base voltage computed for the same stray capacitance as the function h, has a greater resonant frequency FR due to the simpler circuit structure inside the preamplifier. It shows up as a small maximum in the upper hybrid band as well. Thus, all relations between the features of the experimental data pertinent to the identification of the upper hybrid emission have been recovered in simulations. For the large part of CORONAS noise spectra, the enhancement that traces the local plasma density results from a huge overestimation of the weak e.m.f. in a narrow frequency band due to the equivalent circuit properties.

 Fig. 6. Simulation of background and base voltage spectra.

For the CORONAS preamplifiers, any difference in equivalent circuits which results in different resonant frequencies pertinent to different dipole arms may result in significantly structured spectra, without invoking thermal or propagation components of radiation. Wakes created by satellite structure provide an example of such environmental influence. It can be easily anticipated that a superposition of signals from two monopoles imbedded into two plasmas with sufficiently different densities should manifest as two resonant bands. The sequence of four consecutive spectra presumably corresponding to such a configuration is shown in Figure 7 a-d. Two separate noise bands (Figure 7a) marked with r and w move to higher frequencies (Figure 7b) and moderately merge in Figures 7c-d at increasing frequencies. Very impressive spectral structures result from merging but they have no relation to the upper hybrid emission band. Measurements took place when the satellite moved in the plane perpendicular to the sun direction. For such orbits the monopoles separately passed through the satellite body wake in short time intervals. The doublets were occasionally registered in that period in a broad range of frequencies according to the actual value of ambient plasma frequency. The registration in Figure 7a-d took place in favorable conditions when the ambient plasma density changed in the broad range. Figure 7e gives an example of two separated bands from another revolution. The doublet occurs in the frequency range where the merging occurs in Figure 7c-d. Exact attitude calculations are planned later to refine the data analysis but both the previous observations (Kiraga 1986) and the above discussion support the relation of the broadening of a resonant band or its separation into the doublet with differences in equivalent circuits pertinent to dipole arms due to the wake effect. One of the challenges in interpretation of the background spectra may be to distinguish the structures provided by resonant band modifications from those where the resonant band and upper hybrid emission coexist.

Fig. 7. Background spectra influenced by satellite wake.

CONTROL OF ENHANCED BAND AMPLITUDE BY THE SATELLITE E.M.F. 

Each spacecraft introduces perturbation in ambient plasma parameters. In particular the service systems and scientific instruments radiate in a very broad frequency range. The radiation spectral content may be highly variable, depending on plasma parameters and spacecraft operation. The preliminary characteristics of the spacecraft e.m.f. can be obtained in the limits of small and large plasma frequencies relative to the frequency band in question. Such preliminary characteristics are shown in Figures 8a and 8b respectively. The low plasma frequency limit (fp<fc) shows three spectra registered in the south polar cap in order to avoid contamination by the broadcasting radiation. The CORONAS electromagnetic background shows nice temporal stability and rather flat spectrum above 4.5 MHz with the exception of 6-7.3MHz range. The spacecraft interference is the lowest for MHz<f<4MHz but there are two noise bands there. The high fp limit for the frequency range 2-4 MHz is illustrated by two spectra with resonant frequency fr~7MHz. There is some increase of amplitude below fr but the other characteristics are the same as in the fp<fc case. So, the stability of the CORONAS e.m.f. and its weak coupling to the plasma characterizes both fp limits.

 Fig. 8. Influence of spectral characteristics of the satellite e.m.f. on echanced band.

The CORONAS e.m.f. significantly controls the amplitude of the resonant band for the large fraction of data. An illustrative example is given in Figure 8c where several spectra from the time interval 11:10-11.27UT in Figure 1 are superimposed. The values of amplitude maxima follow fairly closely amplitude of the CORONAS e.m.f. in Figure 8a. It would be oversimplification to expect exact similarity because spacecraft intereference should depend on plasma parameters.  

CONTROL OF ENHANCED BAND AMPLITUDE BY THE CIRCUIT RESISTANCES 

Observations 

So far, background spectra with a very pronounced resonant band have been presented. Base voltage spectra were smooth in the frequency range (fc,fp) and had almost equal amplitudes at resonant frequency FR. They can be called the standard spectra. In a large fraction of data the resonance has smaller amplitude than that expected from the superposition of the spacecraft e.m.f. and the equivalent circuit transfer function and occasionally it completely disappears. Selected cases which show main steps in such resonant band evolution are presented in Figure 9. The presentation is organized according to decreasing excursions from the standard spectra. The simultaneously measured pairs of standard spectra are shown with the thin lines for references. The pairs in which the background spectrum have the suppressed resonant band are given with the bold lines. In Figure 9a, the bold line background spectrum shows the white noise in the band 1.5-6 MHz. The spectrum is typical for the regions strongly influenced by ground based HF transmitters. The similar white noise band but with an amplitude smaller by 10dB is shown in Figure 9b. The isolated peak at 3.8 MHz can originate due to either the ground-based transmitters or to the interference line. In order to not obscure the doublet structure in the standard spectrum, the respective

base voltage spectra are given in Figure 9c. In Figure 9d the bold line spectrum remains flat between the weak resonant band and the weakened broadcasting band. In Figure 9e the bold line spectrum only occasionally surpasses the standard spectrum at frequencies greater than fr. Finally in Figure 9f both spectra almost overlap for frequencies outside the resonant band up to the broadcasting bands.  

In all cases the bold line base voltage spectra show excursions from their standard counterparts. It follows from the previous simulations that excursions are related to the departure of equivalent circuit resistances from those in the standard spectra. All spectra in Figure 9 were registered at middle and low latitudes in order to exclude influence of precipitating particles.  

Equivalent Circuit Simulations 

The preamplifiers flown on the CORONAS were designed to provide moderate protection against the real time telemetry transmitters that resulted in high-input resistance and consequently in very pronounced passive resonance at the input of operational amplifier. The pair of diodes was inserted there with a conduction threshold of about 0.5V. The similar pair of diodes was suitably inserted into the input stage of the impedance circuit as the only protection against external hazards. The simpler protection resulted in greater vulnerability to d.c. and a.c. fields and finally in the development of equivalent circuits which may be helpful in the analysis of data with very peculiar spectral structures.

 Fig. 9. Influence of circuit resistances on resonant band.

It was suggested (Kiraga 1995b) that the intense white noise in the band (fc,FoF2) like that in Figure 9a may result from the transition of protecting diodes into conduction regime due to strong radiation from the ground-based transmitters. The flattening of the transfer function and generation of noise at the input of operational amplifier result in a white-noise-like spectrum. It was shown in a recent simulation of preamplifiers for the COMPAS project that monopole resistance can suppress the enhanced band around fr as well as another resonance sensitive to the sheath around the antenna. Modification of the CORONAS transfer function by the resistances variability is shown in Figure 10. The effect of reduced resistance of the diodes is shown in Figure 10a and 10b for fp=7.7MHz and fp=3MHz respectively. In both cases the values of 200kOhm, 11kOhm and 1540Ohm were used which correspond to parallel-to-the-ground resistances in the preamplifiers flown on CORONAS, APEX and ACTIVE satellites. The highest peak corresponds to an unmodified transfer function calculated with a series antenna resistance of 5Ohm. Reduction of preamplifier input resistance to that of APEX results in 8dB reduction of the resonance on fr. Reduction of preamplifier input resistance to that of ACTIVE results in quenching of the resonance and significant underestimation of e.m.f. below 1MHz. The sensitivity of resonance amplitude to the series antenna resistance is shown in Figure 10c. For fp=7.7MHz the resistance of 50Ohm reduces resonance amplitude to the same value as that obtained for APEX input resistance. The antenna resistance of 400Ohm quenches resonance like the ACTIVE input resistance. For fp=3 MHz the transfer functions calculated with monopole resistances of 50Ohm, 400Ohm and 1400Ohm are presented.

Fig. 10. Inflight modification of circuit transfer function by resistances' variability.

Larger values of monopole resistance are required to get the same reduction of resonance amplitude. Simulations and observations consistently point out the relevance of in-flight variability of circuit resistances on the amplitude of resonance at fr. Both resistances may be relevant in various geophysical regions. The data in Figure 9f pertain to the case when the diodes have infinite resistance. The large modification of preamplifier parameters and of the spacecraft e.m.f. influence the data in Figure 9a. It is worth noting that the increase of monopole resistance in Figure 9d-f may be influenced by satellite motion in the Earth’s magnetic field. Noticeably greater cyclotron frequency and angle between satellite velocity and the magnetic field pertain to the bold line measurements. In this paper we make our conclusions from consistent simulations of the base voltage and background spectra for some attitudes. Exact attitude data are necessary to perform such simulations as a general rule. The essentially different vulnerability of both instruments to d.c. and a.c. fields may impose severe constraints on the uniqueness of analysis for some data sets putting them in the class of uncontrolled measurements. 

MEASUREMENT IN THE POLAR REGION 

In the brief survey of the background and base voltage spectra from several passes over the south polar region, no essentially new spectral structures have been found. Two spectral signatures which could be related to the upper hybrid emission were sought. The first corresponds to the case when enhanced band in the noise spectrum occurs at greater frequency than the frequency of maximum in the base voltage spectrum i.e. fr>FR. The second corresponds to such broadening of the enhanced noise that it extends into the (fp,fu) band as inferred from the base voltage measurements. Few such configurations were found but they occurred in the regions with the rapid changes of plasma frequency so they may correspond to the shifting of base voltage spectrum due to ~sec time delay between both spectra. Only the cases when fr>2MHz were analyzed in order to avoid an interference from the spacecraft noise which predominates at lower frequencies. In any case the upper hybrid emission does not manifest itself as a permanent spectral structure in the CORONAS data. 

DISCUSSION 

Analysis of wave, impedance and topside sounder data have established the physical nature of the enhanced band which traces plasma frequency. The enhanced band originates at the receiver input as a passive resonance of the a.c. circuit due to the inductive character of the antenna below the upper hybrid frequency fh. As the result of an equivalent circuit approach to the analysis of raw telemetry data, the frequency location and the amplitude characteristics of the resonance have been related to specific elements of equivalent circuits. The main driving voltage in the receiver circuit was provided by CORONAS electrical circuits. Partly due to the absence of other instruments for low-energy plasma diagnostics, the spacecraft electromagnetic interference had weak temporal variability and essentially flat spectrum in the band 2-10MHz with few noticeable depressions. Compared to our previous missions, the spacecraft interference was very moderately influenced by plasma density. The satellite geometry and the location of antenna contributed to lower interference too. Properties of the satellite interference coupled with the high input resistance of preamplifiers resulted in narrow and pronounced resonances for some environment setups. The environments were encountered where resonance broadening, splitting, fading and quenching occurred. Although the spectral variability of the satellite interference contributed to the resonance structure, it was the response of antenna impedance to environment changes which mainly or totally controlled the resonance spectral content. The frequencies fc and fp determine the frequency band where the antenna is inductive in a cold homogeneous plasma and the Balmain formula can be used as the zero order approximation in calculation of antenna impedance. But the real antenna in the thermal space plasma destroys the homogeneity through its boundary and the generation of e.m.f. due to motion in the magnetic field. In such a case the relation between the antenna characteristics and plasma parameters encompasses ‘a very wide variety of challenging plasma antenna problems to occupy the time of theoreticians and computers indefinitely, and the judicious choice among them is necessary if the work is to be generally applicable (rather than limited to special cases), and of practical significance (rather than sterile manipulations)’ according to Crawford and Harker (1979). The real ionospheric plasma further complicates the problem due to charging by suprathermal fluxes and inhomogenities with the scale length comparable to antenna dimensions. 

The equivalent circuit approach to wave spectra coupled with well-understood measurements of divers plasma parameters offers the opportunity for systematic study of relations between the equivalent circuit elements with the plasma parameters. Once the relations have been set, the relevant plasma parameters may be inferred directly from the spectra.  

In the low-temperature ionospheric plasma the resonance frequency fr is most significantly controlled by the antenna inductance and stray capacitances in front of the input operational amplifier. If the inductance is computed with the Balmain formula, the relation fr<fp holds for our instruments in moderately and weakly magnetized plasma (fp>2fc). As fp decreases below 2fc, the resonance frequency fr may enter into the upper hybrid band (fp,fh), for an increasing range of angles between the antenna and the Earth’s magnetic field (Kiraga et. al. 1995c). The resonance amplitude depends on antenna resistance around fr. The Balmain formula predicts significant amplitude. The overall consistency of our simulations with data registered in very different environments justifies the question about the physical origin of enhanced noise in the upper hybrid band in the first report of Walsh et al. (1964), was it the plasma radiation or the circuit resonance? The equivalent circuit was considered in that study, but the antenna in plasma had been an issue at the very beginning at that time. At least due to low electromagnetic interference, the analysis was biased to the external sources of e.m.f. There were numerous subsequent reports which related variable frequency enhancement in raw telemetry data to the upper hybrid band plasma emissions. The notions of susceptibility of the plasma to radiate in this band and the high input impedance of wave instrument resulted in the assumption that the voltage spectrum inside the instrument is directly related to the spectrum of the voltage which drives the circuit. The implications of inductive character of antennas below fh were overlooked in such a case. The analysis of the CORONAS data proves that the observation of the enhanced band in the telemetered voltage on an electromagnetically very clean platform is not sufficient to assign it to the upper hybrid band emissions. In dense and very cold plasma the circuit is capable of providing a very large signal from the very weak external voltage. It was shown (Kiraga et al. 1995a,b) that the transfer functions calculated for the same ratio fp/fc are almost the same as for the high input resistance preamplifier operating in MHz and kHz frequency range. Paradoxically the first signature of plasma warming is the diminishing of the resonance amplitude due to the real part of antenna impedance. During the injection of the pulsed electron beam from the board of the APEX satellite (Kiraga et al. 1995d,e), very distinct plasma emissions were identified and the emissions related to the upper hybrid band revealed several spectral structures in dependence on the fp/fc ratio. The wave receiver on the APEX had much smaller input resistance than that on the CORONAS, and in several cases, prior to the onset of the beam injections the background spectra were flat in the frequency range below the broadcasting bands. So the standard, moderate resonance in the receiver circuit was destroyed probably by the combined effect of reduced resistance of protecting diodes and the increased resistance of the antenna. The 2µs long beam pulses were sent at a rate of 44kHz and created a highly variable environment which could further modify the elements of the equivalent circuits. Electromagnetic radiation generated due to beam plasma interaction was sufficiently strong to make the impedance issue of secondary importance at least on the preliminary stage of identification of radiation modes. It is very important for the in-situ diagnostic of plasma radiation in the natural processes to anticipate the transition of equivalent circuit parameters from those in the cold plasma to those when the plasma radiation is really detected. Separation of the total e.m.f. into spacecraft interference and external sources may be a complicated task. Without an understanding of such a transition, the interpretation of large volumes of wave data may be highly misleading. Over the Earth’s equator, there is the altitude profile bounded by observations of circuit resonance on the CORONAS and highly structured spectra on several geosynchronous satellites. Broad range of this profile was covered by the EXOS-D observations of permanent enhancement close to the upper hybrid band. The EXOS-D observations could be useful to study transition between low and geosynchronous limits. The impressive simulations of the equivalent circuit in the framework of quasi-thermal noise spectroscopy (Maksimovic et al. 1995) could provide an important ingredient to such study, namely the evolution of a preamplifier transfer function from its cold plasma approximation to that already computed for an extremely hot model. On the other hand, separation of the e.m.f. contribution from the transfer function contribution in the QNT fitting procedure and in the presentation of typical spectra may be very fruitful for extension of its applicability and refinements in data analysis. Does the discrepancy in local density provided by QNT spectroscopy and relaxation sounder follow from probing of different plasma volumes or does it originate from approximations unavoidable in calculation of the transfer function? Our wake affected spectra in Figure 7 indicate that antenna impedance is sensitive to plasma density in rather limited volume. But how large correction to the transfer function is required in order to get agreement with relaxation sounder data and how the correction affects values of other fitted parameters? It was the successful simulation of the discrepancy between wave and topside sounder data which resulted in the systematic development and understanding of equivalent circuit applicable to HF wave receivers in the ionosphere. 

There are several regions in the ionosphere where the measurements are very challenging due to rapid high variability of medium parameters. The very high data rates are required to unfold from measurements of the medium morphology and interactions. But medium variability may result in substantial modification of spacecraft-plasma interaction. In the case of wave measurements, the changes in the interference spectrum and in instrument parameters may occur. It was pointed out (Kiraga 1996) that high time resolution measurements with variable suitably adjusted preamplifier parameters may provide almost instantaneous monitoring of circuit components which are directly controlled by ambient plasma density and spacecraft charging. The instantaneous control of plasma density is required for identification of plasma modes and for verification. The equivalent circuit approach is necessary during instrument design and then in the analysis of data. 

ACKNOWLEDGMENTS:

In Poland this work was supported by the Committee on Schiencific Research Grant 2 Z6Z600304. A. K. benefited from numerous discussions with J. Krasowski, L. Plewka, G. Juchnikowski and J. Compa from the technical staff. One of us (A. K.) would also like to extend heartfelt thanks to his wife Alice, daughters Catherine, Magdalene, and Agnes, and to his sons Martin and Stanley, for their understanding and support during the days and years spent in absence in the processing and learning the divers data from spacecrafts. The Editor initiative to extend the original paper has resulted in acceleration of research and is deeply appreciated. 

REFERENCES 

Balmain K. G., The impedance of a short dipole in a magnetoplasma, IEEE Trans. Ant. & Prop., AP-12, (5), 605, (1964).

Basu B., T. Chang, J.R. Jasperse, Electrostatic plasma instabilities in the daytime lower ionosphere, Geophys. Res. Letters, 9, (1), 68-71, (1982).

Burenin H.A., G.G. Getmantsev, N.A. Mitiakov, V.O. Rapoport, A.M. Sizmin, V.N. Tiukin, Study of passive plasma resonances on Cosmos-142 satellite, Cosmic Space Res.,6, 313-315, (1968).

Crawford F.W., K.J. Harker, Antennas in plasmas: the problem of boundaries, Ann. Telecomunic., 34, (2), 284-290, (1979).

Efimova T.V., N.A. Mityakov, E.E. Mityakova, V.V. Pisaryeva, V.O. Rapoport, Yu.V. Chugunov, Investigation of passive plasma resonances by satellite COSMOS-259, Radiophysics, 19, 977-983, (1976).

Fejer J.A., J.R. Kan, Noise spectrum received by an antenna in a plasma, Radio Sci. 4,(8), 721-728, (1969).

Gaffey, J.D.,Jr. and R.E. LaQuey, Upper Hybrid Resonance in the magnetosphere, J. Geophys. Res., 81,(4), 595-600, (1976).

Hartz T.R., Radio noise levels within and above the ionosphere, Proc.IEEE, 57, No.6, 1042-1050, (1969).

Kiraga A., Wake dynamics of INTERCOSMOS-19 satellite inferred from measurement of ionospheric radio noises, Symposium on Active Experiments, COSPAR XXVI, paper 1.5.8, Toulouse, (1986).

Kiraga A., Z. Klos, V.N. Oraevsky, S.A. Pulinets, V.C. Dokukin, and E.P. Szuszczewicz, Estimation of plasma density from wave data of cold electron plasma, Adv. Space Res., 15, (12), 143-146, (1995a).

Kiraga A., Simulation of impedance and wave measurements in cold space plasma, Chapman Conference on Measurement Techniques, Santa Fe, 3-7 April, (1995b).

Kiraga A., Z. Klos, V.N. Oraevsky, S.A. Pulinets, V.C. Dokukin, and E.P. Szuszczewicz, Estimation of plasma density from wave data of cold magnetoplasma, Chapman Conference on Measurement Techniques, Santa Fe, 3-7 April, (1995c).

Kiraga A., Z. Klos, V.N. Oraevsky, V.C. Dokukin, and S.A. Pulinets, Observation of fundamental magnetoplasma emissions excited in magnetosphere by modulated electron beams, Adv. Space Res., 15, (12), 21-24, (1995d).

Kiraga A., Z. Klos, V.N. Oraevsky, V.C. Dokukin, and S.A. Pulinets, Electromagnetic emissions in the ionosphere - pulsed electron beam system, Chapman Conference on Measurement Techniques, Santa Fe, 3-7 April, (1995e).

Kiraga A., Spacecraft plasma interactions inferred from the impedance and noise measurements, to be published, available on http://www.estec.esa.nl/CONFANNOUN/96a09/main.html, (1996)

Maksimowic, M., S. Hoang, N. Meyer-Vernet, M. Moncuquet, J. Bougeret, J.L. Phillips, P. Canu, Solar wind electron parameters from quasi-thermal noise spectroscopy and comparison with the other measurements on Ulysses, J. Geophys. Res., 100, A10, 19881-19891, (1995).

Oya H., M. Iizima, and A. Morioka, Plasma turbulence disc circulating the equatorial region of the plasmasphere identified by the Plasma Wave Detector (PWS) onboard the AKEBONO (EXOS-D) satellite, GRL, 18, (2), 329,(1991).

Smith P.A., Cerenkov relations between r.f. noise and energetic particles on Ariel 4, Proc.R.Soc.Lond., A343, 241-245,(1975).

Walsh D., F.T. Haddock, H.F. Shulte, Cosmic radio intensities at 1.225 and 2.0Mc/s measured up to an altitude of 1700km, Space Research IV, North-Holland Publishing Co. 935,(1964).