Type I radar echoes from theequatorial electrojet with double peaked Doppler spectra

Transcription

1 Utah State Unversty From the SelectedWorks of Bela G. Fejer January 1, 1980 Type radar echoes from theequatoral electrojet wth double peaked Doppler spectra Bela G. Fejer, Utah State Unversty D. T. Farley P. Johnston B. B. Balsley Avalable at:

2 VOL. 85, NO. A1 JOURNAL OF GEOPHYSCAL RESEARCH JANUARY 1, 1980 Type Radar Echoes From the Equatoral Electrojet Wth Double-Peaked Doppler Spectra B. G. FEJER AND D. T. FARLEY School of Electrcal Engneerng, Cornell Unversty, thaca, New York P. JOHNSTON AND B. B. BALSLEY Aeronomy Laboratory, NOAA, Boulder, Colorado Normal type 1 radar echoes obtaned from relatvely large zenth angles have a power spectrum wth a sngle narrow peak whose Doppler shft corresponds approxmately to the acoustc velocty n the medum. On some occasons, however, ths sngle maxmum splts nto two dstnct peaks, separated n phase velocty on one occason by 270 m/s. Ths bfurcaton s most easly observed at large zenth angles durng daytme when a narrow antenna beam s used. t has also been seen n a daytme experment n whch radars at Jcamarca and Huancayo smultaneously probed the same regon from two dfferent radar zenth angles. The bfurcaton has been observed so far only to the west of Jcamarca, over the Pacfc Ocean. Ths spectral splttng could be caused by vertcal electron densty gradents, such as those assocated wth 'blanketng' sporadc E layers. A suffcently sharp (scale lengths of a few hundred meters or less) postve gradent on the undersde of the layer and negatve gradent on the topsde would cause the type 1 velocty to be decreased and ncreased, respectvely, durng the day, by amounts as large as those observed. NTRODUCTON Type 1 radar echoes from the equatoral electrojet are characterzed by spectra whch have a narrow peak at a Doppler shft.cor'espondng closely to the on acoustc velocty n the E regox2 Type 2 echoes, whch are somewhat better understood, h/tve broader spectra wth smaller Doppler shfts. These latter echoes wll not be consdered n ths paper. A consderable amount of expermental and theoretcal effort has been expended n the last decade and a half n an attempt to understand the type 1 rregulartes and the plasma nstablty whch prodaces them [e.g., Bowles et al., 1963; Farley, 1963; Cohen and Bowles, 1967; Bals!ey and Farley, 1971; Sato, 1973, 1976; Lee et al., 1974; Farley and Fejer, 1975; Rogster and Jamn, 975; W'enstock and Rognlen, 1975; Fejer et al., 1975a, b, 1976; ttanuse and CroChet, 1977, 1978]. As was mentoned above, the type 1 spectra normally have a sngle narrow peak. n ths paper, however, we dscuss some unusual radar data n whch ths sngle peak splts nto two, one wth a larger and one wth a smaller than normal Doppler shft. The measurements were made at Jcamarca and Huancayo n Peru. at approxmately 50 MHz. The most plausble explanaton for the splttng s beleved by us to be that the nstablty threshold was substantally altered by very sharp postve and negatve vertcal densty gradents assoc- ated wth a narro TM layer of enhanced onzaton, such as those responsble for 'blanketng sporadc E' echoes n the HF band. The possble mportance of densty gradents for type 1 echoes was frst suggested by Farley and Fejer [1975]. The predcted effect s neglgble at 50 MHz for normal gradent lengths of a few klometers or more but can become sgnfcant for sharp gradents and/or lower radar frequences. Recent observatons made n Afrca at several frequences between 7 and 29 MHz by Hanuse and Crochet [1977] support ths predcton. Copyrght 1980 by the Amercan Geophyscal Unon. EXPERMENTAL RESULTS Our data were obtaned wth three dfferent radar systems n Peru. The frst operated throughout most of 1975 at the Jcamarca Radar Observatory. ts antenna was a gratng array of eght lnes of dpoles, each of whch was about 8 wavelengths long algned along the magnetc merdan. The lnes were separated by about 6 wavelengths, and the resultng pattern n the east-west plane contaned 13 narrow lobes ponted at zenth angles rangng from 60øE to 67øW. The system, operated wth ts own 10-kW peak power transmtter at a frequency of MHz and ts own dedcated mncomputer and dsplay system, ran contnuously and unmanned, recordng power and spectral data from each lobe on flm. The data of nterest here were obtaned from only one of the lobes (ponted 64øW) at delays of 1610 and 1660/ s, whch correspond to alttudes centered at about and 108 kn. The half-power beam wdth (two way) at ths angle was estmated to be approxmately 1.7 ø, and the pulse length was 50 / s, gvng an overall alttude resoluton of roughly 7 km. Ths s far from deal for our purposes, but t was suffcent to gve us some dscrmnaton between echoes from the upper and lower portons of the electrojet. The ntegraton tme for each spectrum was about 50 s. On most occasons when type 1 echoes were obtaned wth ths system, the echoes at both delays had essentally dentcal spectra. On some days (probably less than 10% of the days on whch type 1 echoes were seen), however, the type 1 spectra exhbted an unusual broadenng followed by the development of a double-peaked structure whch often dffered at the two ranges. Examples of ths sort of behavor are shown n Fgures 1 and 2. 'Low' and 'hgh' refer.to the upper and lower alttudes. Fgure 1 shows the most spectacular example yet observed of ths splttng. On ths day, type 1 echoes were observed startng at about 0920 LT (75øW), and the spectra n the two volumes were nearly dentcal untl about 1135, when the spectrum n the lower volume began to broaden. The fgure shows the subsequent hstory of the spectral splttng untl Paper number 9A /80/009A

3 192 FEJER ET AL.' DOUBLE-PEAKED TYPE ELECTROJET RADAR SPECTRA LOW 16 MAY ø WEST , , oo 8,,.._ o HGH 13 02? ', ' PHASE VELOCTY (m/s) Fg. 1. An example of the evoluton of an extreme bfurcaton of the type spectrum. The alttude resoluton was 6-7 km, and the 'low' and 'hgh' scatterng volumes were centered at approxmately and 108 km, respectvely. The normalzaton constants of the spectra dffer. The radar beam was drected 64 ø west of vertcal. the type 1 echoes dsappeared shortly after Note that (1) the separaton between the peaks s large (almost a 2:1 rato), (2) the mean of the veloctes of the two peaks after the splt s roughly equal to the velocty of the sngle peak before the splt, (3) the observatons at the hgher alttude slghtly favor the peak at the hgher velocty, and vce versa, and (4) the hgh velocty peak s the last to dsappear (the echoes at,-, 1430 are very weak; most of the power shown n the bottom panels s due to nose as a result of the normalzaton procedure used). Fgure 2 shows a second example of the development of double peaks. n ths case the hgh phase velocty peak appeared frst n the upper regon, whle the velocty of the orgnal type 1 peak remaned essentally unchanged as ts power gradually decreased. The same processes took place n the lower regon but started a few mnutes later. By 1300 LT all the power at both alttudes was at the hgh phase velocty, whch was almost 100 m/s greater than that of the ntal peak, and remaned there untl the type 1 echoes dsappeared at about n January 1977, two radar systems, one at Jcamarca and one at Huancayo, were used smultaneously to study the electrojet echoes, the prmary am beng to study neutral wnds and temperature varatons. n addton to these (as yet unpublshed) studes, we had the good fortune to obtan further examples of the spectral bfurcaton. The geometry of the experment s shown n Fgure 3. The Huancayo antenna was a multbeam array smlar to the array used n the 1975 Jcamarca measurements. Eght rows of dpoles, each about 8 wavelengths long, were separated by 3.68 wavelengths. The rest of the radar conssted of a 10-kW peak power transmtter, a dedcated mncomputer dsplay system, and a dgtal magnetc tape unt. The Jcamarca observatons were made by transmttng on a dpole wth a peak power of several hundred klowatts and recevng on two 50-MHz broad beam steerable (n the east-west plane) antennas. The pulse lengths were 50/ s for both systems. The ntegraton tmes were about 70 s for Jcamarca and 80 s for Huancayo. The exact poston of the scatterng volume was determned by range gatng (usng the fact that the electrojet echoes come from a narrow regon of alttude). As s shown, there were a total of 17 separate volumes sampled, 9 of whch were probed smultaneously by both radars. Actually, only the longtude of the volumes was common; the lattudes were separated by about 13 kn, snce Huancayo s not exactly magnetcally east of Jcamarca. n these measurements the echoes were obtaned from the entre alttude range of the electrojet; because of the wder antenna beam wdths t was not possble to dscrmnate between echoes from the upper and lower portons, as was done n Probably because of ths, the spectral bfurcaton observed n 1977 was less dstnct than that observed n 1975, but t s stll readly apparent, as can be seen n Fgures 4 and 5. The echoes from Huancayo labeled 67 ø, 62 ø, and 38 ø correspond to volumes 5, 7, and 9 of Fgure 3, and the Jcamarca echoes shown n Fgure 5 (gong from left to rght) correspond to volumes 1-6. The bfurcaton and shft from a moderate Doppler shft (phase velocty of the order of 300 m/s) to a substantally hgher value (velocty of,-,450 m/s) appears n the frst two columns of Fgure 4 (volumes 5 and 7) and n all columns n Fgure 5 (volumes 1-6). Note that all of these volumes, whch span about 200 km n longtude, are west of Jcamarca. There s apparently a real geographcal effect [Balsley, 1970, 1977; Crochet, 1977]. For example, compare the rght-hand columns of Fgures 4 and 5. These correspond to vrtually the same zenth angle from Huancayo and Jcamarca, respectvely, but the frst shows echoes from volume 9 over the Andes, whle the second nvolves echoes from volume 6 over the ocean. Further evdence on ths pont s gven n Fgure 6, whch t----t' 23 MAY ø WEST '"! , PHASE VELOCTY (m/s) Fg. 2. Another example of the evoluton of a double peak and subsequent ncrease n the Doppler frequency of the type radar echo. For a descrpton of the parameters see the legend for Fgure 1.

4 FEJER ET AL.: DOUBLE-PEAKED TYPE 1 ELECTROJET RADAR SPECTRA 193 Fg. 3. A schematc representaton of the radar geometry durng the smultaneous Jcamarca-Huancayo measurements of January The lattudnal separaton of the scatterng volumes, as seen from the two observatons, was about 13 km. shows echoes obtaned at Jcamarca from the east (volumes 10, l, and 12). There s no sgn of spectral splttng, even at extreme zenth angles. Note also that the spectral peaks n Fgure 6 are at approxmately 400 m/s, or about the mean of the hgh and low veloctes n Fgure 5 (,- 500 and 300 m/s). The power receved at Huancayo from the most easterly volumes (15-17) was small, and the spectra showed no sgn of splttng. The echoes from the smallest zenth angles (volumes 7 and 8 from Jcamarca and volumes and 12 from Huancayo) contaned at most a weak type 1 component and have been gnored here. The type 1 phase veloctes measured from Huancayo and Jcamarca for the same scatterng volume (but dfferent zenth angles) were observed to be n excellent agreement, as s llustrated n Fgure 7 for volumes 4, 5, and 9. When doublepeaked spectra were present, the velocty of only the domnant component s plotted. The tme at whch the ampltude of the large Doppler shft component frst exceeds that of the component wth the smaller shft n volumes 4 and 5 s somewhat later at Jcamarca than at Huancayo, but otherwse the curves are practcally dentcal, as are the curves for volume 9 (east of Jcamarca), where no splttng occurred. These results mply that the zenth angle of the radar observaton s not a crucal factor n determnng whether or not the spectrum may develop a double-peaked structure; there s apparently a genune longtudnal effect. DSCUSSON On the bass of our present partal understandng of the type 1 electrojet rregulartes, the three physcal quanttes that determne the phase velocty of the waves are east-west wnd velocty (we neglecthe vertcal component, whch unpublshe data confrm s a vald approxmaton), temperature, and vertcal densty gradent. We must therefore look to these for an explanaton of the spectral splttng that we have observed. For reasons that are stll not completely understood, the type velocty appears to be the sum of two components, the largest of whch, the threshold velocty predcted by lnear plasma nstablty theory, s ndependent of the radar zenth angle. A smaller component descrbes the convecton of the medum by the east-west neutral wnd and s proportonal to the sne of the zenth angle. The observed phase velocty V,, then, s = Vo + sn 0 where VEw s the zonal wnd, 0 s the zenth angle of the radar beam, and Vo s the threshold velocty, whch s essentally the on acoustc velocty modfed by the effect of densty gradents, as dscussed by Farley and Fejer [1975]. Ths velocty s gven by Vo = Cs[(l + F2) 1/2- g] (2) where F= t, e/2t ek2l vcs (3) Here 1, e, and e,t are the usual collson and Larmor frequences, Cs s the acoustc velocty (proportonal to rl/2), k s the radar wave number, and L v s the electron densty gradent length (postve for upward gradents and westward electron flow, or vce versa). For reasonable electrojet parameters and the Jcamarca radar (k = 2 m-l), F s approxmately 110/L v, where L v s measured n meters, and s usually qute small. Fgure 8, taken from Farley and Fejer [1975], shows the varaton of Vo wth gradent length for 50-MHz observatons. At longer wavelengths the gradent effect may be qute mportant, and Hanuse and Crochet [1977] have shown that type 1 phase veloctes observed wth a multfrequency HF radar agree qute well wth ths theory. Wnds. Could east-west wnds somehow produce the spec- tral splttng? Apparently not. Consder the example n the left column of Fgure at 1352: 16. The peak veloctes are ap O5 HUANCAYO ZENTH ANGLE (WEST) 67 ø t, j_ 62 ø -600 PHASE VELOCTY (m/s) Fg. 4. A set of spectra observed at Huancayo showng the bfurcaton effect. From left to rght, the scatterng volumes are those labeled 5, 7, and 9, n Fgure 3. The ntegraton tme was - 80 s. The ntegratons at 38 ø, 62 ø, and 67 ø, began s, 240 s, and 320 s, respectvely, after the tmes lsted at the left. 18 ø LU

5 194 FEJER ET AL.: DOUBLE-PEAKED TYPE 1 ELECTROJET RADAR SPECTRA , ! "x- 15 O0 45v ZENTH ANGLE (WEST) 65 ø 62 ø 50 ø PHASE VELOCTY (m/s) 42 ø 37 ø,,,,,, K>O 0 Fg. 5. Normalzed spectra from volumes 1-6 observed at Jcamarca. The antenna beam wdth for these measurements was much broader than t was for those shown n Fgures 1 and 2; the echoes are from the entre alttude range of the electrojet. The spectral broadenng, splttng, and shftng can be seen here but are not as well defned as n Fgures 1 and 2. proxmately 560 and 290 m/s. At the zenth angle of 64 ø a sudden dscontnuty (a smooth change would just broaden the spectrum, not splt t), occurrng wthn at most a klometer or two, of about 300 m/s n the wnd velocty would be re- 43 ø ZENTH ANGLE (EAST) 51 ø PHASE VELOCTY _, , , _, _, (m/s) Fg. 6. Spectra from volumes 10, 12, and 14 to the east of Jcamarca durng the perod covered by Fgure 5. Note the complete absence of broadenng and/or double peaks n ths case. Note also the Doppler shft of the peak n comparson to the values of Fgure 5. qured. A wnd pattern shftng from, say, 150 m/s eastward to 150 m/s westward n a few klometers or less n the E regon seems qute unlkely, although dscontnutes of the order of 100 m/s have been seen n the E regon [e.g., Tetelbaum and Sd, 1976]. f ths example s not suffcently convncng, consder the data of January 15, The left-hand column of Fgure 4 and the column labeled 42 ø n Fgure 5 correspond to echoes from volume 5 n Fgure 3. The Huancayo (Fgure 4) velocty peaks are at approxmately 485 and 338 m/s, gvng a separaton of 147 m/s. The correspondng values for the Jcamarca observatons are 472 and 319 m/s, a dfference of 153 m/s. The Huancayo veloctes are slghtly larger, perhaps owng to a westward wnd of the order of m/s. The mportant pont for ths dscusson, however, s that the velocty dfferences between the two peaks are essentally dentcal; f anythng, the Jcamarca dfference s larger, whereas f wnds caused the splttng, the Jcamarca separaton whould be smaller, only 73% of the Huancayo separaton, owng to the dfference n zenth angles. Temperature. Temperature dscontnutes could n prncple explan the observatons, but as wth wnds, ths explanaton s not quanttatvely reasonable. To refer agan to the example n Fgure 1 quoted n the last paragraph, the ve- locty dfference of almost a factor of 2 mples a temperature dfference of almost a factor of 4, whch seems completely out of the queston. As was mentoned above, we note also that a smooth temperature change n a few klometers of even ths magntude wll smply broaden the spectrum, not splt t. Densty gradents. Wthn the framework of exstng theory, the densty gradents are our last hope. We need large gradents to cause a substantal frequency shft and sharp dscontnutes n the gradents to explan the splttng. For example, a thn 'trangular' densty profle n the electrojet re-

6 FEJER ET AL.' DOUBLE-PEAKED TYPE 1 ELECTROJET RADAR SPECTRA 195 b OO.._ $00 3OO o--o --- HUANCAYO,, LOCAL TME Fg. 7. The varaton of the phase velocty correspondng to the spectral maxmum for the echoes observed smultaneously at Jcamarca and Huancayo. The effect of the splttng can be seen n the data from volumes 4 and 5; no such effect occurs n volume 9, whch s east of Jcamarca and over the Andes. gon wth a steep postve gradent on the bottomsde changng suddenly to an equally steep negatve gradent on the topsde would gve the results seen n Fgure 1 f the gradent length were m (see Fgure 8). Echoes from the bottomsde of the layer, where the gradent s destablzng, would have the smallest Doppler shfts, and echoes from the topsde would have the largest. When the layer s absent, the shft should be about mdway between these two extremes. These predctons are n accord wth our observatons, whch, n spte of the poor alttude resoluton, at least crudely suggesthat the echoes wth the hgher Doppler shft are comng from the upper part of the echong regon, and vce versa. The gradents requred n ths case are admttedly very sharp, but on the other hand, the example shown n Fgure 1 s the most extreme ever observed. Note also that the shft of the sngle peaks shown n the top spectra of Fgure 1 s mdway between the double peaks seen after 1300, and the sngle peaks observed to the east of Jcamarca n Fgure 6 are mdway between the double peaks of Fgure 5. So-called 'blanketng sporadc E' layers thn enough to account for our observatons have been observed wth rockets [Smth, 1966; Smth and Mechtly, 1972; Szuszczewcz and Holmes, 1977] and more crudely wth radar [oannds and Farley, 1972; Behnke and Vckrey, 1975; Mller and Smth, 1978] at md-lattudes and have been seen on onograms at equatoral lattudes, where they may persst for hours. Of course, onograms cannot gve a detaled layer profle wth the requred resoluton, but t seems lkely that vertcal gradents at the equator could be as steep or even steeper (owng to suppresson of vertcal dffuson) than they are at temperate lattudes. Unfortunately, there are no onosonde records for the regons from whch the spectrally bfurcated echoes were obtaned. Dfferences n the exact shape, poston, and moton of a thn layer would produce varous spectral shapes. A somewhat fat-topped or rounded densty profle would gve a rather broad spectrum, wth perhaps two (or more) relatvely weak peaks correspondng to regons of approxmately constant gradent. f the layer moves slowly downward th, rough the thn (-- 3 km for daytme type 1 electrojet echoes; see Fejer et al. [1975a]) electrojet regon, as would be the case for a layer drven by gravty-wave-assocated wnd shear, we would expect the type 1 Doppler peak to frst decrease n frequency, then broaden or splt, and then ncrease to above the orgnal value. A tlted layer drftng through the beam would produce the same sequence f the tlt were upward n the drecton of travel. Ths s a very typcal scenaro; the converse has never been observed. Part of ths sequence mght not be observed f the layer s a localzed patch (as s often the case at temperate lattudes; see Mller and Smth [1978]) whch drfts horzontally nto or out of the radar beam. The fact that a dstnct bfurcaton of the type 1 spectral peak s qute rare at Jcamarca but a moderate broadenng of the peak s not seems consstent wth the gradent explanaton; only profles wth unusually sharp gradents and dstnct knks at the rght alttudes wll produce an obvous splttng. One way to confrm our hypothess that gradents cause the splttng would be to observe the effect smultaneously at two well-separated radar frequences. The magntude of the separaton n phase velocty would be largest for the lowest frequency. As yet, no such data exst, unfortunately. We do not understand why ths effect has been observed only to the west of Jcamarca, over the ocean. t s doubtless related to other east-west electrojet asymmetres whch have been observed at Jcamarca [Balsley, 1970, 1977] and whch are also not understood. Afrcan observatons. Crochet et al. [1979] reported examples of smlar spectral splttng observed n Afrca durng very strong daytme counterelectrojet (eastward electron drft) condtons usng a 14.2-MHz radar. They suggest that ther observatons can be accounted for by strong wnds and refracton effects. We wll not comment further on ths paper here; we merely wsh to emphasze that ths explanaton cannot account for the Jcamarca observatons, for whch refracton s not mportant and unrealstc wnd patterns would be requred. 600 g 500 _o 400 (n _ - - ll ELECTRON DENSTY GRADENT LENGTH LN (m) Fg. 8. The effect of an electron densty gradent on the type 1 threshold phase velocty for 3-m rregulartes near the center of the electrojet scatterng regon. The length L v s No(dN/dh) -, and the electrons are assumed to be drftng to the west (daytme condtons; upward electrc feld).

7 196 FEJER ET AL.: DOUBLE-PEAKED TYPE 1 ELECTROJET RADAR SPECTRA CONCLUSONS Densty gradents appear to be the most plausble source of the splttng of type echo spectra whch s occasonally observed to the west of Jcamarca, over the Pacfc Ocean, but not to the east, over the Andes. n any case, t seems most unlkely that sudden dscontnutes n temperature or wnd velocty could explan the observatons; the dscontnutes requred are just too extreme. Ths gradent explanaton, dscussed n the prevous secton, may not be the whole story, however; n the absence of a quanttatve nonlnear theory whch explans the smple type 1 spectrum, t s premature to assume that we completely understand ths more complcated observaton. n partcular, we need to understand why the spectral peak (or peaks) appears to correspond always to the threshold velocty, ndependent of radar zenth angle, even when the threshold s strongly affected by a vertcal gradent. The fact that ths does appear to be the case, however, further supports the hypothess of Farley and Fejer [1975], outlned n (1)-(3). Ths gradent effect may make t dffcult to determne temperature varatons accurately by studyng varatons n the type 1 phase velocty. There seems to be no easy way to separate the temperature and gradent effects. Acknowledgments. We thank the staffs of the Jcamarca and Huancayo observatores for ther help wth the observatons and M. erkc for hs help wth the observatons and data processng. Ths work was supported by the Aeronomy Program, Dvson of Atmospherc Scences of the Natonal Scence Foundaton, through grants ATM , ATM , and ATM The Jcamarca and Huancayo observatores are operated by the Geophyscal nsttute of Peru, Mnstry of Educaton, wth support from the Natonal Scence Foundaton and the Natonal Aeronautcs and Space Admnstraton. The Edtor thanks M. Crochet and K. L. Mller for ther assstance n evaluatng ths paper. REFERENCES Balsley, B. B., A longtudnal varaton of electron drft velocty n the equatoral electrojet, J. Geoœhys. Res., 75, 4369, Balsley, B. B., E-regon dynamcs, J. Atmos. Terr. Phys., 39, 1087, Balsley, B. B., and D. T. Farley, Radar studes of the equatoral electrojet at three frequences, J. Geoœhys. Res., 76, 8341, Behnke, R. A., and J. F. Vckrey, Radar evdence for Fe 4- n a sporadc E layer, Rado Sc., 10, 325, Bowles, K. L., B. B. Balsley, and R. Cohen, Feld-algned E regon rregulartes dentfed as on acoustc waves, J. Geoœhys. Res., 68, 2485, Cohen, R., and K. L. Bowles, Secondary rregulartes n the equatoral electrojet, J. Geoœhys. Res., 72, 885, Crochet, M., Radar studes of longtudnal dfferences n the equatoral electrojet: A revew, J. Atmos. Terr. Phys., 39, 1103, Crochet, M., C. Hanuse, and P. Broche, HF radar studes of the twostream nstablty durng an equatoral counterelectrojet, J. Geoœhys. Res., 84, 5223, Farley, D. T., A plasma nstablty resultng n feld-algned rregulartes n the onosphere, J. Geophys. Res., 68, 6083, Farley, D. T., and B. G. Fejer, The effect of the gradent drft term on type 1 electrojet rregulartes, J. Geoœhys. Res., 80, 3087, Fejer, B. G., D. T. Farley, B. B. Balsley, and R. F. Woodman, Oblque VHF radar spectral studes of the equatoral electrojet, J. Geoœhys. Res., 80, 1307, 1975a. Fejer, B. G., D. T. Farley, B. B. Balsley, and R. F. Woodman, Vertcal structure of the VHF backscatterng regon n the equatoral electrojet and the gradent drft nstablty, J. Geoœhys. Res., 80, 1313, 1975b. Fejer, B. G., D. T. Farley, B. B. Balsley, and R. F. Woodman, Radar observatons of two-dmensonal turbulence n the equatoral electrojet, 2, J. Geoœhys. Res., 81, 130, Hanuse, C., and M. Crochet, Multfrequency HF radar studes of plasma nstabltes n Afrca, J. Atmos. Terr. Phys., 39, 1097, Hanuse, C., and M. Crochet, Oblque HF radar studes n the equatoral electrojet n Afrca, J. Atmos. Terr. Phys., 40, 49, oannds, G., and D. T. Farley, ncoherent scatter observatons at Arecbo usng compressed pulses, Rado Sc., 1, 763, Lee, K., C. F. Kennel, and F. V. Coront, On the margnally stable saturaton spectrum of unstable type 1 equatoral electrojet rregulartes, J. Geoœhys. Res., 79, 249, Mller, K. L., and L. G. Smth, ncoherent scatter radar observatons of rregular structure n md-lattude sporadc E layers, J. Geoœhys. Res., 83, 3761, Rogster, A., and E. Jamn, Two-dmensonal nonlnear processes assocated wth type rregulartes n the equatoral electrojet, J. Geoœhys. Res., 80, 1820, Sato, T., Unfed theory of type 1 and 2 rregulartes n the equatoral electrojet, J. Geoœhys. Res., 78, 2232, Sato, T., On mechansms governng the electrojet plasma nstabltes, J. Geoœhys. Res., 81, 539, Smth, L. G., Rocket observatons of sporadc E and related features of the E regon, Rado Sc., 1, 178, Smth, L. G., and E. A. Mechtly, Rocket observatons of sporadc E layers, Rado Sc., 7, 367, Szuszczewcz, E. P., and J. C. Holmes, Observatons of temperature gradents n md-lattude Es layers, J. Geoœhys. Res., 82, 5073, Tetelbaum, H., and C. Sd, Formaton of dscontnutes n atmospherc gravty waves, J. Atmos. Terr. Phys., 38, 413, Wenstock, J., and T. D. Rognlen, Nonlnear saturaton and angular rotaton of nstabltes n the E layer, Rado Sc., 10, 231, (Receved July 6, 1979; revsed September 13, 1979; accepted September 13, 1979.)