Annales Geophysicae, 23, 1227–1237, 2005SRef-ID: 1432-0576/ag/2005-23-1227© European Geosciences Union 2005

AnnalesGeophysicae

Orientation of the cross-field anisotropy of small-scale ionosphericirregularities and direction of plasma convection

E. D. Tereshchenko1, N. Yu. Romanova1, and A. V. Koustov2

1Polar Geophysical Institute of Russian Academy of Sciences, 15 Khalturina, 183 010 Murmansk, Russia2Institute of Space and Atmospheric Studies, University of Saskatchewan, Saskatoon, Canada

Received: 15 November 2004 – Revised: 11 March 2005 – Accepted: 22 March 2005 – Published: 3 June 2005

Abstract. The relationship between the orientation of thesmall-scale ionospheric irregularity anisotropy in a planeperpendicular to the geomagnetic field and the direction ofplasma convection in the F region is investigated. The cross-field anisotropy of irregularities is obtained by fitting theo-retical expectations for the amplitude scintillations of satel-lite radio signals to the actual measurements. Informationon plasma convection was provided by the SuperDARN HFradars. Joint satellite/radar observations in both the auroralzone and the polar cap are considered. It is shown that theirregularity cross-field anisotropy agrees quite well with thedirection of plasma convection with the best agreement forevents with quasi-stationary convection patterns.

Keywords. Ionosphere (Auroral ionosphere; Ionospheric ir-regularities)

1 Introduction

The high-latitude ionosphere is an inhomogeneous media inwhich the quasi-layered distribution of electron density withheight also changes horizontally, with spatial scales fromhundreds to tens of kilometers. In addition to the large-scale structuring, much finer irregularities of the electrondensity are often observed, more frequently at the edges oflarge-scale structures (Tsunoda, 1988). Such irregularitiescan be of various scales, from kilometers to centimeters;they are often referred to as the small-scale irregularities. Itis well established that small-scale irregularities are gener-ated in the high-latitude ionosphere through various plasmainstabilities (e.g. Keskinen and Ossakow, 1983; Tsunoda,1988; Gondarenko and Guzdar, 2004). Both theory andobservations indicate that the small-scale irregularities areanisotropic; they are strongly stretched along the geomag-netic field and often have a preferential direction in a planeperpendicular to the geomagnetic field; in this paper, the di-

Correspondence to:N. Yu. Romanova(romanova@pgi.ru)

rection of this elongation will be called the orientation of thecross-field anisotropy.

Various parameters of small-scale ionospheric irregular-ities can be measured by radio methods (e.g. Gusev andOvchinnikova, 1980; Ruohoniemi et al., 1987; Afraimovichet al., 2001), and numerous results have been reported in thepast (e.g. Moorcroft and Arima, 1972; Martin and Aarons,1977; Fremouw et al., 1977; Rino et al., 1978; Rino andLivingston, 1982; Gailit et al., 1982; Eglitis et al., 1998).Despite significant progress in this field, the relationship be-tween the irregularity parameters and the conditions in thebackground ionospheric plasma is not well established.

Recently, Tereshchenko et al. (1999) developed a newmethod of satellite signals analysis that allows one to infersuch important characteristics of the ionospheric anisotropicirregularities as the degree of their stretching along andperpendicular to the geomagnetic field and the orienta-tion of the cross-field anisotropy. Further expansions ofthis method were recently presented in Tereshchenko etal. (2004). Tereshchenko et al. (2000a) applied the originalmethod to the analysis of auroral zone irregularities and, bycomparing the inferred orientations of cross-field anisotropywith the direction of plasma convection, as measured bythe EISCAT incoherent scatter radar, found their reasonableagreement. Since the joint satellite-EISCAT data set was lim-ited, Tereshchenko et al. (2002) expanded the investigation ofthe orientation of the cross-field anisotropy by involving theHeppner and Maynard’s plasma convection model (Rich andMaynard, 1989). Again, reasonable agreement was foundbetween the orientation of the cross-field anisotropy and theplasma convection direction given by the model for speci-fied conditions. It was noted that occasionally the inferredorientation of the cross-field anisotropy was quite differentat two closely spaced receiver sites (∼100 km). These in-consistencies were attributed to strong spatial variations ofplasma flow, though no supporting data were provided.

This study continues the investigation of the relationshipbetween the orientation of the cross-field anisotropy of iono-spheric irregularities and the direction of plasma convection.

1228 E. D. Tereshchenko et al.: Orientation of the cross-field anisotropy

We compare Tereshchenko et al. (1999, 2004) method pre-dictions with convection data provided by the Super DualAuroral Radar Network (SuperDARN) HF radars. The ad-vantage of the SuperDARN radars for this kind of work is intheir capability to monitor the plasma convection in spatiallyextensive areas of the high-latitude ionosphere with temporaland spatial resolutions of 1–2 min and∼45 km, respectively.We consider three different experiments. The first two werecarried out in the auroral zone, and these comparisons ex-pand the previous analysis by Tereshchenko et al. (2000a;2002). We then consider the third experiment with observa-tions in the polar cap, where the geophysical conditions forionospheric irregularity formation can be different.

2 Determination of the irregularity parameters fromamplitude scintillations of satellite signals

Tereshchenko et al. (1999, 2000a,b, 2004) presented detailsof their method that allows one to infer several characteris-tics of the ionospheric irregularities from scintillations of thesatellite signal amplitudes measured on the ground. Here webriefly give an overview of the method and demonstrate someof its features. The method is based on the so-called Ry-tov’s approach (Rytov et al., 1978). It is assumed that thereis an ionospheric layer homogeneously filled with three-dimensional (anisotropic) irregularities of electron density.The irregularity spectrum as a function of wave number isdescribed by the power law with an arbitrary index. Forsatellite signals passing such a layer, the variance of the loga-rithm of the signal amplitude relative to the signal amplitudein the irregularity-free situationσ 2

χ is predicted theoreticallyand compared with measurements. This parameter was se-lected for the comparison because it is very sensitive to anassumed shape of the irregularities, including the orientationof the cross-field anisotropy.

2.1 Basics of the theory

According to Tereshchenko et al. (2004), the variance of thelogarithm of the relative amplitudeσ 2

χ is

σ 2χ=

λ2r2e αβL

3−p

0 π (p−1)/2

2−p/2+3 sin(π4 (p − 2))0((p−3)/2)

∫ zu

zl

σ 2(z)√

1 + γ (z)Rp−2F (z)

×F

(1−p

2,

1

2, 1,

γ (z)

1+γ (z)

)[c(z)−

√a2(z)+b2(z)

]−p/2

dz, (1)

where

χ = lnA

A0. (2)

In Eq. (2),A is the measured signal amplitude, andA0 isthe signal amplitude that would be measured in the absenceof ionospheric irregularities.

In expression (1)

γ (z)=2√a2(z)+b2(z)

c(z)−√a2(z)+b2(z)

,

a=1

2[(α2

−1) sin2 θ(z)+(β2−1)(sin2ψ cos2 θ(z)− cos2ψ)],

b=(β2−1) sinψ cosψ cosθ(z),

c=1+1

2[(α2

−1) sin2 θ(z)+(β2−1)(sin2ψ cos2 θ(z)+ cos2ψ)], (3)

whereF(1−p2 ,

12, 1,

γ (z)1+γ (z)

) is the hypergeometric function,λ is the wave length of the radio signal,re is the classicalelectron radius,RF is the Fresnel radius,zl andzu are thelower and upper boundaries of the irregularity layer, thezaxis is assumed to coincide with the look direction from thereceiver to the satellite,0 is the gamma-function,p is thepower index,L0 is the outer scale of irregularities,θ(z) isthe angle between the satellite-receiver direction and the vec-tor of the local geomagnetic field,α andβ are field-alignedand cross-field elongations of the irregularities, and9 isthe orientation angle of the cross-field anisotropy (the angleis counted from geographic north, positive clockwise). Weshould note that the Rytov’s approach is valid for weak scin-tillations, i.e. for those withσ 2

χ<0.3. Numerous measure-ments of the amplitude fluctuations in the subauroral and au-roral ionospheres showed that this condition is met for mostof cases (Aarons, 1982).

Tereshchenko et al. (1999) proposed to plot the experi-mentally determinedσ 2

χ in terms of satellite position alongthe meridian and then to compare this curve with a set oftheoretically expected dependencies. One can then find thebest-fit theoretical curve to the measured profiles ofσ 2

χ andthus infer the irregularity parametersα, β and9. The fittingprocedure is greatly simplified by the fact that the latitudi-nal location of theσ 2

χ theoretical maximum depends solely

on the angle9 while the shape of theσ 2χ theoretical pro-

file depends onα andβ. Note that the maximum amplitudescintillations occur for satellite positions in the vicinity ofthe magnetic zenith. The success of the fitting proceduredepends on whether the satellite pass is near the magneticzenith or away from it. To characterize how far the satellitepath was from the magnetic zenith, the minimum look angleθmin (from a receiver to a satellite) and the local directionof the geomagnetic field is considered. For the case of themagnetic-zenith path (θmin<1◦), the maximum of scintilla-tions occurs exactly at the magnetic zenith, and only param-eterα can be determined since signal oscillations originatefrom isotropic irregularities withβ=1. For a non-zenith path,the shape of theσ 2

χ profile is also influenced by anisotropicirregularities (β>1), so that bothα andβ can be determined.Note that in this case, theσ 2

χ maximum does not exactly cor-respond to the satellite position withθ=θmin.

2.2 Measurements of9: zenith and non-zenith satellitepasses

Figure 1 shows experimental (solid line) and theoretical(crosses and dots) curves for the variance of the logarithmof the relative amplitudeσ 2

χ versus geographic latitude. Bothcases of a) near zenith and b) non-zenith satellite passes areconsidered, with anglesθmin of 0.5◦ and 7.3◦, respectively.

E. D. Tereshchenko et al.: Orientation of the cross-field anisotropy 1229

The scintillation data were collected at a receiver site locatedon the Kola Peninsula. For modeling, it was assumed that theionospheric irregularities filled a statistically homogeneouslayer with boundarieszl=250 km andzu=350 km and that thevariance of the electron density fluctuations was the same atall heights. Theoretical predictions are shown for two val-ues of9, in each case a) and b); one value corresponds tothe case of the best fit between the theory and measurements(dots) and the other one (crosses) corresponds to the case ofa significantly different angle9; we selected this angle tobe 40◦ away (anticlockwise) from the direction of the bestfit. This second value of9 is considered to demonstratehow sensitive the position of the theoretical maximum to thechoice of9 is, for both zenith and non-zenith satellite passes.

In case a), the best fit is achieved forα=55, β=1 and9=106◦. In case b), the best fit is obtained forα=30, β=5and9=79◦. The 40◦ offset in the angle9 changes signif-icantly (not significantly) the position of theσ 2

χ theoreticalmaximum for the non-zenith (zenith) pass. This implies thatthe angle9 can be inferred quite accurately from the experi-mental data for non-zenith satellite passes. We performed ex-tensive analysis of the satellite data and found that, for non-zenith passes, a 4◦

−6◦ variation in9 changes noticeably theposition of the theoretical maximum forσ 2

χ . We also foundthat the horizontally anisotropic irregularities become de-tectable starting fromθmin=1◦

−1.5◦. Luckily, for most satel-lite trajectories the magnitude ofθmin exceeds these mini-mum values. We estimated the uncertainty in the determina-tion of9 by finding a set of9 values for which the differencebetween the experimental and theoreticalσ 2

χ curves was notsignificant. Forθmin>1.5◦, the uncertainty is of the order of2◦

−6◦ and it increases forθmin<1.5◦. We should note thatthe uncertainty in the determination ofα andβ is larger; itranges from several units of elongation (10%–20% effect) toa difference (100%–300% effect) of two or more. What isimportant though is the fact that the larger uncertainty in thedetermination ofα andβ does not affect the uncertainty inthe determination of9.

The analysis performed allowed us to conclude that thevalue of9 can be determined very reliably from the ampli-tude scintillations of the satellite signals.

2.3 Determination of the parametersα andβ

Now we demonstrate how the magnetic field elongationof ionospheric irregularities (parameterα) and cross-fieldanisotropy (parameterβ) can be determined from satellitescintillation data. Figure 2 shows the theoretical profiles forσ 2χ versus geographical latitude for two values ofβ, dots for

the optimal value ofβ>1 and crosses for the case of isotropicirregularities (β=1). The experimental data were obtainedon 16 November 1997, 21:34 UT at three receiver sites inNorway: Karvika (69.87◦ N, 18.93◦ E), Tromsø (69.59◦ N,19.22◦ E) and Nordkjosbotn (69.22◦ N, 19.54◦ E). The satel-lite trajectories in all three cases were of non-zenith type; theminimum angles between the line of sight to the satellite andthe geomagnetic field wereθmin=7.2◦ in Karvika,θmin=7.2◦

Fig. 1. Experimental (solid line) and theoretical (dots and crosses)latitudinal profiles (the geographic latitude is used) of the logarithmof the relative amplitude of a satellite signal (σ2

χ ) for (a) near zenithand (b) non-zenith passes over a receiver site on the Kola Penin-sula. Dots correspond to the case of the best fit between the modeland experiment and crosses correspond to the case of the cross-fieldanisotropy orientation rotated by 40◦ anticlockwise from the direc-tion of the best fit. Also shown is the minimum angleθmin betweenthe look direction from the receiver site to the satellite and the localgeomagnetic field direction at the F region heights.

in Tromsø andθmin=7.4◦ in Nordkjosbotn. We indicate oneach panel the irregularity parameters for the case of the bestfit between the experimental and theoretical curves. One cansee that the position of the theoretical maximum is not verysensitive to the choice ofβ in all three cases (whether it is 1or 7). On the other hand, the width of the curves is stronglyaffected byβ. Detailed analysis shows that variations of theparameterα change the shape of theσ 2

χ curve near the max-imum while variations of the parameterβ strongly controlthe “tails” of theσ 2

χ curve; generally, a decrease (increase)

of eitherα or β makes theσ 2χ profiles broader (narrower).

Importantly, the analysis shows that a uniquely defined setof α andβ can be found for each satellite pass, if the bestfit between the theoretical and experimental profiles ofσ 2

χ issought. One can also conclude from Fig. 2 that the experi-mental curves cannot be described by the model of isotropicirregularities for non-zenith passes; this is in contrast to thecase of the almost-zenith pass considered in Fig. 1a.

1230 E. D. Tereshchenko et al.: Orientation of the cross-field anisotropy

Fig. 2. Experimental (solid line) and theoretical (dots and crosses)latitudinal profiles of the logarithm of the relative amplitude of asatellite signal (σ2

χ ) for a pass over Karvika, Tromsø and Nordkjos-botn (all in northern Norway) on 16 November 1997 at∼21:34 UT.Geographic latitudes are used. Dots (crosses) correspond to the caseof anisotropicβ>1 (isotropic,β=1) ionospheric irregularities. Theirregularity parameters for the best fit between the experimental andmodel profiles are given in the upper right corner of each diagram.

2.4 Multi-receiver observations: some conclusions on theionospheric conditions

The Tereshchenko et al. (1999) method allows one to inferthe irregularity parameters in the ionospheric region abovethe receiver location. If data of several receivers are com-pared, conclusions on the spatial homogeneity of the irregu-larity layer can be drawn.

Consider observations presented in Fig. 2. The estimatedirregularity parameters areα=40, β=7, 9=58◦ (±2◦) inKarvika, α=40, β=7, 9=60◦ (±2◦) in Tromsø andα=40,

β=7,9=60◦ (±2◦) in Nordkjosbotn. For all three sites, thedata show only one maximum well described by the samevalue ofα and the same value ofβ. This implies that theelectron density fluctuations (anisotropic irregularities) areof the same character (shape) above these sites, and their dis-tribution is quite uniform. Certainly, this is a very special sit-uation; generally, one cannot expect such a spatial uniformityof the density fluctuations over distances of tens to hundredsof kilometers in the high-latitude ionosphere. In the case ofnon-uniform irregularity spatial distribution, one can observemore than one peak in the latitudinal profiles ofσ 2

χ . Also, forthe case of a single maximum in the profile, different valuesof α andβ can be obtained even at close receiver locations.

Figure 3 presents experimentalσ 2χ curves obtained at

Karvika, Tromsø and Nordkjosbotn on 14 November 1997at 18:28 UT and corresponding theoretical profiles. The bestparameters describing the data areα=20,β=4,9=91◦(±3◦)for Karvika, α=20, β=4, 9=120◦ (±3◦) for Tromsø andα=25, β=5, 9=125◦ (±3◦) for Nordkjosbotn. In this case,only Tromsø and Karvika data can be described by the samemodel of irregularities, though the orientation of the cross-field anisotropy is different at these locations. Different ori-entation of the cross-field anisotropy over Karvika occurred,very likely, because of a change in the direction of plasmaconvection, as suggested for similar cases by Tereshchenkoet al. (2002).

We should note that some observations do not show a clearmaximum for theσ 2

χ profile, so that the irregularity param-eters cannot be determined at all. One example is given inFigure 4 for 9 November 1997 at 15:10 UT. Here the well-defined isolated maxima are seen at Karvika and Tromsø;best fitting for these data givesα=30,β=6,9=41◦ (±3◦) atKarvika andα=30,β=6,9=60◦ (±3◦) at Tromsø. We can-not determine the irregularity parameters over Nordkjosbotn.The most likely reason is that the irregularities were veryweak or patchy. During the period of 15:08–15:18 UT, theTromsø heater was producing artificial irregularities near thezenith of the station. This allowed us to reliably determinethe irregularity parameters at this location. Since plasma con-vection was directed poleward, the artificially generated ir-regularities were drifting poleward and strong satellite sig-nals scintillations were seen at Karvika. The artificial ir-regularities were not able to reach Nordkjosbotn while thebackground fluctuations were probably too weak to producestrong scintillations. Similar situations were described byTereshchenko et al. (2000a,b); we present here the additionalcase to illustrate and stress some limitations of the method.

The data presented in this section demonstrate that a net-work of satellite signal receivers separated by less than onehundred kilometers can provide important information on thefine structure of the high-latitude ionosphere.

2.5 On the role of time-averaging in the model

The approach based on Eq. (1), that we have discussed sofar, has a minor inconsistency in terms of data handling andmodelling. When experimental data are processed, theσ 2

χ

E. D. Tereshchenko et al.: Orientation of the cross-field anisotropy 1231

Fig. 3. The same as in Fig. 2 but for 14 November 1997, 18:28 UT.

profile is obtained by computingσ 2χ for every 8–12-s period

and merging all obtained values into one latitudinal profile.The theoretical curve is obtained by computingσ 2

χ at everyinstant of time, for example, for every second (below we willcall such a curve/profile “the instantaneous curve/profile”).Clearly, it is desirable to produce the theoretical curve in thesame fashion as the experimental one, i.e. instead of an in-stantaneous value ofσ 2

χ for every second, we consider theσ 2χ

value averaged over 8–12 s. In this section we investigate thesignificance of this averaging effect and its potential impacton the determination of parametersα andβ. Our analysisshowed that the time averaging does not affect the determi-nation of9.

Figure 5a shows averaged (crosses) and instantaneous(solid line) theoretical curves for a near zenith satellite passover a receiver at Tromsø. For the purpose of illustration,we selected typical values ofα=40 andβ=6. The value of

Fig. 4. The same as in Fig. 2 but for 9 November 1997, 15:10 UT.At Nordkjosbotn, the maximum ofσ2

χ at ∼68.7◦ was not strong,and the irregularity parameters were not determined.

9=71◦ was selected so that the maximum of theσ 2χ instan-

taneous profile (solid line) is achieved exactly at the angleθmin. We show the latitudinal variation forθ by the dottedline in Figure 5a, and one can see that its minimum coincideswith the maximum of theσ 2

χ curve. The model values of theanisotropy parameters are also shown in the figure. The valueθp=7.7◦ indicates the zenith angle of the satellite positioncorresponding to the peak inσ 2

χ . One can see in Fig. 5a that

the averaged profile ofσ 2χ (crosses) reaches its maximum at

θ=θmin (at the same latitude as for the instantaneous profile),but its magnitude is smaller than the maximum of the instan-taneous profile. Varying the parameter9 only shifts boththe instantaneous and averaged curves horizontally (withoutthe curves’ distortion), indicating that the averaging effect iscontrolled by onlyα andβ. We found that the case of Fig. 5ais a very typical situation for many passes.

1232 E. D. Tereshchenko et al.: Orientation of the cross-field anisotropy

Fig. 5. Averaged (crosses) and instantaneous (solid line) theoreticalvariations ofσ2

χ versus geographic latitude for a receiver at Tromsø.

Panel a) corresponds to the case of theσ2χ maximum (achieved at

the angleθ=θp=7.7◦) which is exactly at the latitude ofθmin whilepanel b) corresponds to the case of theσ2

χ maximum (achieved atthe angleθ=θp=20.2◦) located at the latitude lower than the lati-tude ofθmin. Computations were performed for the integration timeof 10 s and typical irregularity parameters wereα=40 andβ=6. Dotsshow the latitudinal variation of the angleθ .

For some passes and irregularity parameters the differencebetween the averaged and instantaneous curves is not signif-icant. Figure 5b illustrates such a situation for observationsover Tromsø. Here we consider the pass withθmin=7.7◦ andthe same parametersα andβ as in the previous case, butthe orientation of the cross-field anisotropy (parameter9A)is different. One can see that the instantaneous (solid line)and averaged (crosses)σ 2

χ curves coincide and both maximaare achieved at the look angle ofθp=20.2◦, i.e. significantlyaway from the magnetic zenith.

By considering various satellite passes and varying theirregularity parameters we were able to draw three generalconclusions. First, if an instantaneousσ 2

χ curve has its max-imum near the pointθ=θmin, then the averaging effect is notsignificant for smallα and whenβ<α. For example, if aver-aging is done over 10 s,α should be less than 10–12. This im-plies that the irregularities should be moderately anisotropicto neglect the averaging in the model. Second, for stronglyanisotropic irregularities (for example,α more than 10 and

β<α), the averaging effect is not significant for satellitepasses with theσ 2

χ maximum achieved at large anglesθ of∼15◦–20◦. Third, the averaging effect is less significant ifan instantaneous curve has its maximum away from the pointθ=θmin. Finally, we found that consideration of the averag-ing effect is more important for the determination ofα; val-ues ofβ usually do not change significantly.

To give a sense of the error in estimation ofα andβ weconsider the case of Figure 5a. Application of our standardprocedure (without considering the averaging effect) to theaveraged curve (crosses) givesα=31 andβ=6. We see thatβ did not change whileα is now smaller by 9 (31 versus40). This means that if the instantaneous theoretical curve(solid line) is fit to the experimental curve for the consideredpass (so that the averaging effect is ignored), then the derivedvalue ofα is smaller than it should be.

3 Results of joint SuperDARN - satellite signal observa-tions

In this study we consider data collected in three indepen-dent experiments. The first experiment was run between9 and 15 November 1997 in northern Norway, in con-junction with ionospheric HF heating (Tereshchenko et al.,2000b). The satellite signal reception was conducted atthree sites, Karvika (69.87◦ N, 18.93◦ E), Tromsø (69.59◦ N,19.22◦ E) and Nordkjosbotn (69.22◦N, 19.54◦ E), separatedby ∼100 km. The SuperDARN radars were operated in thestandard mode with 2-min scanning through the field of view.Thirteen events of joint radar-satellite data were identifiedand studied.

The second experiment of a similar type was run in June2001, with the exception that the satellite signal receptionwas performed at Futrikelv (69.80◦ N, 19.02◦ E), Tromsø(69.59◦ N, 19.22◦ E) and Seljelvnes (69.25◦ N, 19.43◦ E),also separated by∼100 km. Seven events were consideredfor this experiment.

The third experiment was conducted on Spitsbergenarchipelago, at a settlement of Barentsburg (78.1◦ N,14.21◦ E) from September 2000 to April 2001. We obtained28 joint events for this experiment.

For the readers convenience, we remind one that Super-DARN is a network of HF radars continuously monitoringechoes from the high-latitude ionosphere (Greenwald et al.,1995). Currently, SuperDARN consists of 9 radars in theNorthern Hemisphere and 7 radars in the Southern Hemi-sphere. It is assumed that the Doppler frequency of theechoes is the line-of-sight (cosine) component of the plasmaconvection vector. This assumption is justified by the factthat the phase velocity of the F region decametre irregulari-ties is very close to the drift of the bulk of the plasma (Ruo-honiemi et al., 1987). To obtain a map of plasma convectionvectors, all available velocity measurements are fit into theconvection model, and the optimal solution is found by theleast-squares fit procedure (Ruohoniemi and Baker, 1998).Utilization of SuperDARN data is very convenient for the

E. D. Tereshchenko et al.: Orientation of the cross-field anisotropy 1233

Fig. 6. SuperDARN convection map (thin lines originated at dots)for 15 November 1997 between 12:16 and 12:20 UT and the ori-entation of the cross-field anisotropy according to measurements atKarvika and Nordkjosbotn (thick lines) at∼12:18 UT. The coordi-nates are geographic longitude and geographic co-latitude.

purposes of the present work because of the good temporal(∼1–2 min) and spatial (45 km) resolution of the measure-ments. In this study, we considered data gathered by all Su-perDARN radars in the Northern Hemisphere but the majorcontribution was always made by the Pykkvibaer (Iceland)and Hankasalmi (Finland) radars observing directly in thearea of scintillation measurements.

3.1 Auroral zone observations: the case of stationary con-vection

We first consider results for the auroral zone observationsin November 1997. Joint satellite-SuperDARN data wereavailable for various periods in between 12:26 and 22:56 UT(roughly 10:30–21:00 MLT). The orientation of the cross-field anisotropy varied in between 62◦ and 125◦, clusteringat 75◦–90◦. We remind one that the angles are counted fromgeographic north, clockwise. We should note that the orien-tation of the cross-field anisotropy can only be determinedup to the constant of 180◦. We conclude that the overall ori-entation of the cross-field anisotropy is consistent with theprevailing direction of the plasma convection at the latitudesof the auroral oval.

Let us now show some individual measurements. In 8cases out of 13 events, the convection and satellite data werefor the same area and the satellite-radar data were comparedquantitatively. Figure 6 gives an example of such a compar-ison for 15 November 1997. Here the SuperDARN convec-tion maps (thin vectors originating from the dots) are givenfor 12:16, 12:18 and 12:20 UT, together with the orienta-tion of the cross-field anisotropy of ionospheric irregularitiesat Karvika and Nordkjosbotn (thick vectors). The data onthis map (and all others, considered in this study) are pre-sented in geographic coordinates. In the past, Tereshchenkoet al. (2000a) used geomagnetic coordinates. This differenceis not important for this study, as the target of the investiga-tion is the azimuthal difference between the irregularity elon-gation and the convection direction, which is independent ofthe coordinate system used.

At Tromsø there was no strong maximum in theσ 2χ curve,

and these measurements are not considered. The scintilla-tion measurements refer to 12:18 UT. The orientation of thecross-field anisotropy was9K=257◦ (±3◦) in Karvika and9N=262◦ (±3◦) in Nordkjosbotn. The convection direc-tion obtained at the nearest SuperDARN point at 12:18 UTwas9SD=267◦. The difference between the orientation ofthe cross-field anisotropy and the convection direction is19K=−10◦ for Karvika and19N=−5◦ for Nordkjosbotn.Importantly, these differences1ψ are small. This signifiesthat the small-scale ionospheric irregularities were elongatedin the direction of the plasma convection. The fact that the ir-regularity anisotropy orientations were the same implies thatthe plasma flow was spatially uniform and this conclusion isconsistent with more coarse SuperDARN measurements.

A good agreement between the orientation of the cross-field anisotropy and the convection direction was observedin other cases; we present statistics in Fig. 7 for all 8events. The data were binned with a 5◦-step in the orientationangle. The positive (negative) values denote those measure-ments for which the satellite-inferred value of9 was larger(smaller) than9SD. The histogram shows that the differ-ences are less than 5◦ in most cases.

In 5 cases for the November 1997 experiment, the in-formation on the convection was not available for the im-mediate vicinity of scintillation measurements because the

1234 E. D. Tereshchenko et al.: Orientation of the cross-field anisotropy

Fig. 7. Histogram distribution for the difference19 between theorientation of the cross-field anisotropy and the direction of theplasma convection for observations between 9 November and 15November of 1997. 5◦ bins of the azimuth are used.

SuperDARN data were patchy. For these cases we comparedthe data only qualitatively. Figure 8 gives an example of sucha comparison for 14 November 1997. The scintillation mea-surements were performed at Tromsø at 17:02 UT. Super-DARN was continuously providing data over 10-min inter-val of 16:54–17:04 UT, but there were no convection vectorsin the region of the scintillation measurements. One can seethat the convection pattern is fairly stable, with similar con-vection vectors to the south, west and east of Tromsø. If morevectors were available, we would expect a good agreementbetween the SuperDARN and scintillation measurements at17:02 UT (note, that observations such as shown in Fig. 8were not included in the statistics of Fig. 7). We found ageneral agreement between satellite and radar measurementsfor all five events.

3.2 Auroral zone observations: a case of non-uniform con-vection

For the second auroral zone experiment, conducted in June2001, reasonable quality SuperDARN convection maps wereobtained for 2, 5 and 11 June. The orientations of the cross-field anisotropy were available for 2 June in Futrikelv andSeljelvnes, for 5 June in Futrikelv, Tromsø and Seljelvnes,and for 11 June in Tromsø and Seljelvnes. Unfortunately,for most of these events, the SuperDARN convection vectorswere quite separated from the areas of satellite measurementsand we were not able to compare the data quantitatively.Qualitatively, the orientation of the cross-field anisotropywas always in reasonable agreement with the convection di-rection in nearby regions. Interestingly, for 2 June 2001, theorientation of the cross-field anisotropy and the plasma con-vection were both in the geographically meridional direction.

Interesting results were obtained for 5 June 2001. In Fig. 9we show the convection map, together with the satellite mea-surements of the orientation of the cross-field anisotropy

Fig. 8. SuperDARN convection maps (thin lines originating fromdots) for 14 November 1997 between 16:54 and 17:04 UT and theorientation of the cross-field anisotropy (thick lines) according toTromsø measurements at 17:02 UT. The coordinates are geographiclongitude and geographic co-latitude.

for 21:08 UT. Convection maps prior to this moment werevery similar to the one shown in Fig. 9. For this event,9F=222◦ (±1◦) at Futrikelv,9T =236◦ (±2◦) at Tromsø and9S=256◦ (±2◦) at Seljelvnes. The convection direction mea-sured by SuperDARN at the nearest point was9SD=256◦ sothat the differences between the orientation of the cross-fieldanisotropy and the convection direction were19F=−34◦ atFutrikelv,19T =−20◦ at Tromsø and19S=0◦ at Seljelvnes.Clearly, only the Seljelvnes measurements at the most equa-torward site were in good agreement with the convection di-rection. However, the SuperDARN measurements show thatthe convection pattern was strongly non-homogeneous; theconvection was turning from the eastward flow at the veryhigh latitudes of∼75◦ to westward flow at latitudes of∼70◦.In other words, the azimuth of the convection vectors wasincreasing with a decrease of latitude. Thus, there was acorrelation in spatial variations of9 and9SD. The largedifferences between9 and9SD at Futrikelv and Tromsøsignify that the convection direction experienced significant

E. D. Tereshchenko et al.: Orientation of the cross-field anisotropy 1235

local variations that were not detected by SuperDARN. Thisexample illustrates the fact that multipoint scintillation ob-servations can be convenient for investigations of local struc-turing in the plasma flow.

We should say that measurements with9-9SD differencesas large as 20◦ were not rare for spatially non-uniform ortemporally changing convection patterns. For the cases ofquasi-stationary convection in time but non-homogeneous inspace (around an area of measurements), the angle9 wasvarying with latitude, in general agreement with expectedchanges of the convection pattern.

Our conclusion from the analysis performed is that multi-receiver scintillation measurements can provide additionalinformation on plasma convection and supplement the Su-perDARN maps.

3.3 Polar cap observations

Now we consider observations on the Spitsbergenarchipelago, at the settlement of Barentsburg. Theanalysis of scintillation data for this location showed thatthe Tereshchenko et al. (1999) method works quite wellfor polar cap conditions; the latitudinal profiles ofσ 2

χ showtypically a single maximum with values below 0.3 (thecriterion for the amplitude scintillations to be treated assmall-amplitude ones) and the shape and latitudinal locationof theσ 2

χ curve can well be described by a theoretical curvedefined by parametersα, β and9.

For the Baretsburg observations, measurements coveredthe time sector of 01:06–23:32 UT (almost all MLT times).The orientation of the cross-field anisotropy varied from 8◦

to 173◦ with no preferential direction. The reason for this isthat the observations were quite frequently carried out nearthe foci of the large-scale convection cells with rather circu-lar flows, contrary to the zonal flows typical for the mainlandNorway observations in the auroral zone.

We split available experimental data into two groups. Forthe first group, the SuperDARN convection maps were fairlystationary in time. For the second group, the maps showedsignificant temporal variations. A stationary convection mapto us was one for which the convection pattern in the vicin-ity of scintillation measurements did not show significantchanges within several minutes (6–12 min) prior to a momentof the comparison.

Figure 10 gives an example of the SuperDARN/satellitecomparison for a relatively stable convection pattern on 5February 2001, 13:06–13:12 UT. According to scintillationmeasurements at 13:10 UT, the orientation of the cross-fieldanisotropy was9 =258◦ (±1◦). The SuperDARN convec-tion direction at the closest point in time (13:10 UT) and inspace was9SD=259◦ so that the difference in angles wassmall19=1◦, within the error of measurements. One cansee that over a 6-min interval, the convection pattern did notchange much in the area of the comparison. The convectiondirections were9SD=259◦–262◦. It is not a surprise to seesmall differences between the convection direction and theorientation of the cross-field anisotropy for these stable pat-

Fig. 9. SuperDARN convection map (thin lines originating fromdots) for 5 June 2001 at 21:08 UT and the orientation of thecross-field anisotropy (thick lines) according to measurements atFutrikelv, Tromsø and Seljelvnes at 21:08 UT. The coordinates aregeographic longitude and geographic co-latitude.

Fig. 10. The same as in Fig. 8 but for SuperDARN observationson 5 February 2001, 13:06–13:12 UT and for measurement of theorientation of the cross-field anisotropy at 13:10 UT.

1236 E. D. Tereshchenko et al.: Orientation of the cross-field anisotropy

Fig. 11. The same as in Fig. 9 but for SuperDARN observations on22 March 2001, 02:48–02:56 and the anisotropy orientation mea-surements at Barentsburg, Spitsbergen at 02:52 UT.

Fig. 12. Histogram distribution for the difference19 betweenthe orientation of the cross-field anisotropy and the direction of theplasma convection for observations at Barentsburg, Spitsbergen be-tween September 2000 and November 2001. 5◦ bins of the azimuthare used.

terns. This was not the case for non-stationary convectionpatterns.

Figure 11 compares SuperDARN and satellite data for anevent of 22 March 2001, 02:48–02:56 UT for which the con-vection pattern was significantly changing. The scintillationmeasurements were performed at 02:52 UT. For the closestpoints,9=205◦ (±5◦) and9SD=205◦ (02:52 UT), meaningthat19=0◦. For other frames, significant differences are ob-vious. This event demonstrates the importance of comparingradar and satellite data for as close as possible spatial areasand minimal difference in time. Clearly, utilization of the Su-perDARN radars that can monitor the convection dynamics

with a 2-min resolution is advantageous for the purposes ofthe present work; averaging over longer intervals can lead tosmoothing out the short-lived local features in the convectionmap and more significant differences between the convectiondirection and the orientation of the cross-field anisotropy. Onthe other hand, this event lends additional support to the no-tion, first expressed by Tereshchenko et al. (2002) and fur-ther discussed in this study (Sect. 2.4), that significant dif-ferences in the orientation of the cross-field anisotropy atclosely spaced points occur because of the local non unifor-mity in ionospheric convection flows.

Figure 12 presents statistics of the Barentsburg’s compar-isons for 28 events. The dark (grey) columns refer to thoseevents for which the convection pattern was stable (non-stable); out of all events, in exactly half of them, the con-vection pattern was stable. One can see that the differences19 were smaller for the stable convection patterns; for mostof the events,19 was within an interval of−5◦

÷0◦. For thenon-stable convection events, the majority of the events alsodemonstrated relatively small differences,19 was within±10◦. For some events the differences were as large as 40◦.

Our overall conclusion for the polar cap is that there is areasonable agreement between the orientation of the cross-field anisotropy and the convection direction.

4 Summary and conclusions

In this study we further investigated the relationship betweenthe orientation of the irregularity cross-field anisotropy andthe plasma convection direction in the high-latitude iono-sphere. The orientation of the cross-field anisotropy of ir-regularities was inferred from the amplitude scintillations ofthe satellite signals received on the ground. To achieve this,the latitudinal profile for the variance of the logarithm ofthe relative amplitude of the received signal was comparedwith the theoretically expected profiles, and the irregularityparameters corresponding to the best fit of the theory andexperiment were obtained. The ionospheric plasma convec-tion measurements were performed with the SuperDARN HFradars. We considered three different experiments, two in theauroral zone and one in the polar cap.

We demonstrated that the method of the irregularity pa-rameters determination works well for observations not onlyin the auroral zone but also in the polar cap. To further im-prove the method, we investigated the effect of time aver-aging on the quality of model predictions; in previous stud-ies, one instantaneous theoretical profile of signal fluctua-tions versus latitude has been used. We demonstrated that thetime averaging has to be considered if one needs to estimatethe elongation of the irregularities along and perpendicularto the geomagnetic field with better accuracy than from thetime- independent model. Importantly, the time averagingdoes not affect the model estimates for the orientation of thecross-field anisotropy.

By comparing the satellite and SuperDARN data for threeindependent experiments we showed that the orientation of

E. D. Tereshchenko et al.: Orientation of the cross-field anisotropy 1237

the irregularity cross-field anisotropy was fairly close, within±10◦, to the direction of the plasma convection for the eventswith quasi-stationary convection pattern in the area of com-parison. For the cases with quickly changing convection pat-terns, the agreement was satisfactory, with maximum differ-ences of the order of 40◦, if the comparison was performedfor nearly the same moments. We can conclude that theionospheric small-scale irregularities are elongated with theplasma convection direction in the plane perpendicular to themagnetic field. This conclusion is in line with expectationsfrom the theory of the gradient-drift plasma instability in thehigh-latitude F region; it is predicted that as the instabil-ity progresses, plasma blobs experience stretching along theconvection direction.

We also demonstrated that the multipoint satellite signalobservations with the site separation of less than 100 km canbe useful for studying the small-scale structures in the iono-spheric plasma flows and thus provide additional informationto the large-scale SuperDARN convection maps. The satel-lite scintillation data can successfully supplement the Super-DARN maps in those regions of the high-latitude ionospherewhere the HF echoes are not detected. Finally, the multipointsatellite data may give information about the spatial unifor-mity of flows and their temporal stability.

Acknowledgements.The authors are grateful to J. A. Gorin for hishelp with SuperDARN data processing and to B. Z. Khudukon, Y.A. Mel’nichenko and V. M. Sukhorukov for their help in scintil-lation measurements. We thank all national funding agencies andindividuals who make the SuperDARN convection maps available.This work was supported by the NSERC (Canada) grant to AVKand by the Russian Foundation of Fundamental Research grant No03-05-64937.

Topical Editor M. Lester thanks two referees for their help inevaluating this paper.

References

Aarons, J.: Global morphology of ionospheric scintillations, Proc.of the IEEE, 70, 360–378, 1982.

Afraimovich, E. A., Lipko, Yu. V., and Vugmeister, B. O.: Parame-ters of small- and medium-scale irregularities in the high-latitudeionosphere as inferred from the Norilsk observations, Geomag.Aeron., 41, 120–127, 2001.

Eglitis, P., Robinson, T. R., Rietveld, M. T., Wright, D. M., andBond, G. E.: The phase speed of artifical field-aligned irregu-larities observed by CUTLASS during HF modification of theauroral ionosphere, J. Geophys. Res., 103, 2253–2259, 1998.

Fremouw, E. J., Rino, C. L., Livingston, R. C., and Cousins, M.C.: A persistent subauroral scintillation enhancement observedin Alaska, Geophys. Res. Lett., 4, 539–542, 1977.

Gailit, T. A., Gusev, V. D., Ivanov, M. I., and Perekalina, E. O.:Differential-phase study of the anisotropy of ionospheric irregu-larities, Geomag. Aeron., 22, 622–625, 1982.

Gondarenko, N. A. and Guzdar, P. N.: Plasma patch structuringby the nonlinear evolution of the gradient drift instability inthe high-latitude ionosphere, J. Geophys. Res., 109(A09301),doi:10.1029/2004JA010505, 2004.

Greenwald, R. A., Baker, K. B., Dudeney, J. R., Pinnock, M., Jones,T. B., Thomas, E. C., Villain, J.-P., Cerisier, J.-C., Senior, C.,Hanuise, C., Hunsucker, R. D., Sofko, G., Koehler, J., Nielsen,E., Pellinen, R., Walker, A. D. M., Sato, N., and Yamagishi,H.: SuperDARN: A global view of the dynamics of high-latitudeconvection, Space Sci. Rev., 71, 763–796, 1995.

Gusev, V. D. and Ovchinnikova, N. P.: Model determination of thevolumetric characteristics of ionospheric inhomogeneities, Geo-magn. Aeron, 20, 434–437, 1980.

Keskinen, M. and Ossakow, S.: Theories of high-latitude iono-spheric irregularities: a review, Radio Sci., 18, 1077–1091, 1983.

Martin, E. and Aarons, J.: F layer scintillations and the aurora, J.Geophys. Res., 82, 2717–2722, 1977.

Moorcroft, D. R. and Arima, K. S.: The shape of the F-region irreg-ularities which produce satellite scintillations evidence for axialasymmetry, J. Atmos. Terr. Phys., 34, 437–450, 1972.

Rich, F. J. and Maynard, N. C.: Consequences of using simple ana-lytical functions for the high-latitude convection electric field, J.Geophys. Res., 94, 3687–3701, 1989.

Rino, C. L., Livingston, R. C., and Matthews, S. J.: Evidence forsheet-like auroral ionospheric irregularities, Geophys. Res. Lett.,5, 1039–1042, 1978.

Rino, C. L. and Livingston, R. C.: On the analysis and interpreta-tion of spaced-receiver measurements of transionospheric radiowaves, Radio Sci., 17, 845–854, 1982.

Ruohoniemi, J. M., Greenwald, R. A., Baker, K. B., Villain, J. P.,and McCready, M. A.: Drift motions of small-scale irregularitiesin the high-latitude F region: an experimental comparison withplasma drift motions, J. Geophys. Res., 92, 4553–4564, 1987.

Ruohoniemi, J. M. and Baker, K. B.: Large-scale imaging of high-latitude convection with SuperDARN HF radar observations, J.Geophys. Res., 103, 20 797–20 811, 1998.

Rytov, S. M., Kravtsov, Yu., A., and Tatarskii, V. I.: Introduction tostatistical radio-physics. Part 2. Random Processes (in Russian).Nauka, Moscow, 1978.

Tereshchenko, E. D., Khudukon, B. Z., Kozlova, M. O., and Ny-gren, T.: Anisotropy of ionospheric irregularities determinedfrom the amplitude of satellite signals at a single receiver, Ann.Geophys., 17, 508–518, 1999,SRef-ID: 1432-0576/ag/1999-17-508.

Tereshchenko, E. D., Khudukon, B. Z., Kozlova, M. O., Evstafiev,O. V., Nygren, T., Rietveld, M. T., and Brekke, A.: Comparisonof the orientation of small scale electron density irregularitiesand F region plasma flow direction, Ann. Geophys., 18, 918–926, 2000a,SRef-ID: 1432-0576/ag/2000-18-918.

Tereshchenko, E. D., Kozlova, M. O., Evstafiev, O. V., Khudukon,B. Z., Nygren, T., Rietveld, M., and Brekke, A.: Irregular struc-tures of the F layer at high latitudes during ionospheric heating,Ann. Geophys., 18, 1197–1209, 2000b,SRef-ID: 1432-0576/ag/2000-18-1197.

Tereshchenko, E. D., Romanova, N. Yu., and Kozlova, M. O.: Com-parison of the cross-field anisotropy of small-scale irregularitieswith ionospheric convection model, Geomag. Aeron, 42, 432-436, 2002.

Tereshchenko, E. D., Kozlova, M. O., Kunitsyn V. E., and An-dreeva E. S.: Statistical tomography of subkilometer irregular-ities in the high-latitude ionosphere, Radio Sci., 39(RS1S35),doi:10.1029/2002RS002829, 2004.

Tsunoda, R. T.: High-latitude F region irregularities: A review andsynthesis, Rev. Geophys., 26, 719–760, 1988.