THE UNIVERSITY OF M ICHIGAN COLLEGE OF ENGINEERING Department of Electrical Engineering Space Physics Research Laboratory Scientific Report Noo J'S-6 ELECTRON TEMPERATTURE EVIDENCE FOR NONTHEPRMAEB EQUILIBRIUM IN THE IONOSPHERE No Wo Spencer Lo H. Brace G. R. Carignan ORA Project 05599 under contract with: NATIONAL AERONAUTICS AND SPACE ADMINISTRATION CONTRACT NASw-1 5 WASHINGTON' Do C administered, through: OFFICE OF RESEARCH ADMINISTRATION ANN ARBOR May 1962

JOURNAL OF GEOPHYSICAL RESEARCB VOLUME 67, No. 1 JANUARY 1962 Electron Temperature Evidence for Nonthermal Equilibrium in the Ionosphere N. W. SPENCER,1 L. H. BRACE, AND G. R. CARIGNAN University of Michigan, Ann Arbor, Michigan Abstract. A brief review of the theoretical and experimental aspects of the'dumbbell' electrostatic probe experiment is given. The electron temperature and positive ion density data resulting from its use in two flights at Fort Churchill, Manitoba, Canada, in the spring and early summer of 1960, and in two flights at Wallops Island, Virginia, in midsummer 1960 and spring 1961, are presented and discussed. The electron temperature profiles exhibit large positive gradients between 150 and 250 km, gradually changing to an approximately isothermal region (approximately 2800~K) above 250 km. An exception to this was the second Wallops Island measurement in which a negative temperature gradient was observed above 250 km, in a quiet ionosphere. The ion density profiles show only slight density maxima in the E region, the less distinct F1 regions, and, at Wallops Island, well defined F, maxima at about 300 kilometers. Electron density data from (1) two-frequency beacon experiments flown on the same rockets, and (2) simultaneously recorded ionograms are presented for comparison. It is noted that good agreement is obtained between the electron density measured by the two different techniques and the ion density as measured by the dumbbell technique. It is concluded, on the basis of additional temperature data obtained from a small cylindrical probe that the dumbbell experiment measures temperatures that are representative of all thermal electrons. It is further concluded that if gas temperatures derived from satellite drag measurements and other techniques are valid as generally accepted, electrons and neutral particles are not in thermal equilibrium in the daytime ionosphere to altitudes of at least 420 km. Finally, it is observed that the electron temperature varies significantly with the state of the ionosphere, electron temperatures being higher and more variable in the more disturbed ionospheres. Introduction. The earth's ionosphere, the ex- electron densities up to the level of the density tensive region of the very high atmosphere com- maximum (-300 km). posed of neutral and ionized particles, has been The ionosonde method, although providing the subject of increasingly intensive study since data that have made possible much of our presits presence was first suggested some 80 years ent knowledge of the ionosphere, has two ago by Stewart, and postulated 20 years later major drawbacks in that it is an indirect method by Heaviside and Kennelly. Until about 1925, and, for an earth-bound station, is limited to when Appleton and Barnett measured arrival measurements below the F, maximum. Further, angles of reflected electromagnetic waves, and interpretation of ionograms is sometimes a diffithus specifically demonstrated the existence of cult and controversial matter,' and, in some a region of charged particles, work was con- cases, e.g., a spread F condition, is not generally fined largely to theoretical studies. About this feasible. time, however, Breit and Tuve devised a scheme As with most scientific studies of an experifor transmitting short pulses of radio frequency mental nature, one prefers to conduct as nearly energy vertically and for studying the reflected direct measurements of the desired quantities signal to obtain quantitative measurements. This as the technology and other aspects permit. In'ionosonde' technique has been developed ex- particular reference to the ionosphere, direct tensively over the years and today provides a measurements became possible when the V-2 large amount of quantitative, world-wide data rocket was made available for performing sci(ionograms), from which one may determine entific measurements. Passage of the rocket 1 Now with the NASA-Goddard Space Flight 2 See, for example, Program of URSI Annual Ccnter. Meeting, Commission 3, Washington, D. C., 1961. 157

158 SPENCER, BRACE, AND CARIGNAN Fig. 1. The dumbbell electrostatic ionosphere probe. through the lower ionosphere made possible relatively large magnitude sawtooth voltage (in two broad types of measurements: those result- terms of now well-known electron thermal enering from (1) an analysis of the effects of charge gies) was applied between the electrode and the on the passage of an electromagnetic wave as rocket, and the resulting current was telemetered the rocket progressed through the region, and to ground. Subsequent attempts at interpreta(2) techniques that measure strictly local proper- tion of the resulting data revealed the shortties, with the rocket as the reference point. Both comings of the exploratory instrument and stimtypes were undertaken initially; the former, ulated extensive theoretical study which, on the however, not requiring a knowledge of the spe- basis of certain assumptions relative to rocket cific manner in which the rocket might alter its surface properties, probe geometry, electron environment, and being more closely allied to energy distribution, ion mass, etc., indicated the already existing ionosonde technique, re- that good measurements of electron temperaceived the greatest attention and has, over the ture and positive ion density should be realizyears, been more extensively developed [Seddon, able. These results are documented in early 1953; Bcrning, 1960]. Measurements employ- reports and papers, which were published in the ing these radio propagation techniques have literature [Dow and Reifman, 1949; Hok, yielded data in essential agreement with ground- Spencer, and Dow, 1953; Hok, Spencer, Dow, based measurements, and the results have re- and Reifman, 1951]. ceived orderly publication in appropriate tech- Several years later, when the opportunity to nical journals [Jackson, Kane, Seddon, 1956; carry on the experiment was again at hand, the Berning, 1960]. results of the earlier work were used as a startThe second general technique has not, until ing point. It was concluded that ejecting an apmore recently, been at all adequately under- propriate probe instrument from a suitable taken, although several exploratory Langmuir rocket into the ionosphere offered the greatest probe experiments to determine electron tem- likelihood of obtaining valid measurements. This perature were carried out using the V-2 rocket approach eliminated the necessity of considering by University of Michigan experimenters. An- varying surface conditions, the effect of escapother direct measurement technique adopted ing gas, the presence of unwanted potentials, somewhat later uses the changing impedance and other uncertainties necessarily encountered of a dipole antenna with passage through the by using the rocket as one of the probe elecionosphere to provide electron density measure- trodes. Accordingly, a probe of simple geomements [Jackson and Kane, 1959]. try, appropriate to a straightforward theoretical The early experiments employing the Lang- treatment, and state-of-the-art instrumentation, muir probe technique, or, more generally, the was selected. Figure 1 illustrates the instrument electrostatic probe technique, took the form of that was developed and is being employed in the a rather simple electrode (truncated conical basic experiment reported herein. Several desleeve) mounted at the nose of the rocket. A velopmental experiments were carried out, with

NONTHERMAL EQUILIBRIUM IN THE IONOSPHERE 159 varying degrees of success [Boggess, Brace, and the sphere can be represented by the expression Spencer, 1959]. The dumbell-shaped instrument electrically NeI (i) simulates two spherical electrodes isolated in the 2-rm ionosphere connected by a suitable voltage gen- where erator and current indicating device. The dumb- J = random current density. bell, in addition to functioning as an electro- N = particle number density. static probe, acts entirely independently as a e electron charge dipole antenna for the internally containedparticle kinetic temperature. telemeter. ~~~~~telemeter. in^n = particle mass. Theory of the measurements. The theoretical m parle ma ss. k = Boltzmann constant. basis for the experiment can be described in the following manner: Assume a spherical conductor Assuming that the predominant ion is O* immersed and at rest in the ionosphere where [Johnson, Meadows, Holmes, 1958], the ratio of there exists thermal equilibrium, for each parti- electron to ion current density is about 170 or cle species at least, and a mean free path long the square root of the mass ratio of these two with respect to the sphere diameter. In this components. This is the ratio of the number of case, the random current density of a particular electrons to the number of positive ions (numspecies of charged particles some distance from ber density of negative ions is considered negligiT I Hemisphere I + ---- Hemisphere 2 LIZ_ _ Aon Saturation Region Bti Plasma I I Collector Potentials (VOLTS) -- V' ion Satu.rtion RL o - -- i +\ Fig. 2. Volt-ampere characteristics of the dumbbell probe hemispheres.

160 SPENCER, BRACE, AND CARIGNAN I~~I.^~ ~ electron current (2) and the ion current (4), assuming N, = N,; thus __-__ __ _ -I = I + I, = ANe[(a/r)2' - 2, exp ( /Vo)] (5) The volt-ampere characteristic of the sphere, Fig. 3. A composite volt-ampere chafreteristtc of calculated from (5), is shown by the solid curve the dumbbell probe. of Figure 2. The AB portion (ion saturation region) is due almost entirely to ions, as all'thermal' electrons are repelled by the negative ble) that would cross a unit area in unit time in sphere. From about the point B to the plasma the region. This will not, however, be the case at potential, the electron current, as indicated by the surface of the sphere because the predomi- I, on the figure, becomes increasingly predominance of electron flux will charge the sphere nant. negatively, thus accelerating ions and retarding Consider a second identical sphere to be some electrons. Potential equilibrium will be estab- distance from the first sphere, with a variable lished, when the electron and ion fluxes are voltage 8V connected between the spheres. The equivalent, for a sphere potential defined in this volt-ampere characteristic of the second sphere paper as the wall potential, V,. can be represented by the dashed line in Figure The magnitude of the electron current that 2. This curve is shown inverted with respect to flows to the sphere under equilibrium is given that of the first sphere, for a voltage applied by between the spheres forces the potential of one of them positive (net electron current flow) and I= A.Je. V \ the other negative (net positive ion current I,= A.J, exp V-] A,Ne 2J2rm exp (-k) (2) where'. A, = area of the spherical collector Vo = kT/e (3) To neutralize the negative surface charge, the plasma, in effect, surrounds the sphere with a positive ion sheath (a region where the ion-toelectron number density ratio is much greater (8V)f A ) B than unity) which is the order of a few centi- meters thick for a collector in the ionosphere between altitudes of 100 and 400 km. The sheath increases the effective area of the sphere for ion current by a factor (a/r)2, where a is the sheath radius and r is the sphere radius; thus the ion current is given by I, = AJ r(a/r)2 = AN~e 2yT (a/r)2 (4) Fig. 4. A schematic representation of the dumbThe net current to the sphere I is the sum of the bell probe.

NONTHERMAT, EQUITIBRIUM IN THE IONOSPHERE 161 6V VOLTS ------ - Sec I _ Fig. 5. Typical voltage-current curves for the dumbbell probe: (a) applied voltage; (b) resulting current; (c) current as it is expected to appear on telemetry. flow). The sphere potentials, V1 and V,, adjust probe that is a very good approximation of the continuously to meet the zero net current re- above idealized sphere-pair can be achieved. quirement of the whole probe system over the Figure 4 illustrates schematically its essential entire range of the applied 8V. features. The two outer hemispherical portions The volt-ampere characteristic of the sym- are insulated from the adjacent funnel-shaped metrical pair of spheres is obtained graphically electrodes and are interconnected by a suitable from Figure 2 by plotting the net current I for sawtooth generator and series connected'ama suitable range of 8V as shown in Figure 3. meter' (solid-state current detector). The apThus, the vertical displacement of this composite plied voltage SV and the resulting current I curve represents the current that would be ob- comprise the desired'raw data' from which served to flow between the two spheres if con- electron temperature and positive ion density nected by a generator of voltage SV. As in are computed. Figure 2, the outer, linear portions represent ion The funnels, insulated from each other as well saturation current to the more negative sphere. as from the hemispheres, are connected by a A practical probe. A physically realizable generator SVA producing a voltage identical in

162 SPENCER, BRACE, AND CARIGNAN s-Ki-:-BS"..:....... X.%.::J:~ i:"" siU:~.~3 r.~:~:' Fig. 6. A telemeter record showing the volt-ampere curve of the dumbbell probe. magnitude and phase to V.. The funnels thus The rocket orientation at ejection, the ejection serve as guard electrodes for the hemispheres, technique (forward by spring action from a resulting in a system that very closely approxi-'clamshell' nose-cone) and the moment of inertia mates the system of two isolated spheres dis- ratio for the dumbbell generally result in an cussed above. The like generators and essentially end-over-end motion (typically 4-second period) identical areas of the hemispheres and funnels in a near vertical plane. Additional equipment assure that there will be no significant potential sometimes carried by the rocket, but not ejected, difference between each hemisphere and its as- includes a radio Doppler system (DOVAP) sociated guard-funnel. Thus, a series of spherical employed for precision trajectory determination. equipotential surfaces can be considered to exist On some occasions, as will be discussed, a'twoabout each hemispherical electrode. frequency beacon' is carried on the rocket to The indicated voltage source EW (Figure 4) provide additional and comparative data. It is represents the telemetry source which employs not separated from the rocket. the dumbbell as a nearly ideal dipole antenna. Electrical characteristics. To detect sheath The telemetry frequency employed is sufficiently distortion and photo-electric current high (approximately 240 Mc/s) so that the d-c effects (to be discussed later) a two-section sawprobe system and RF telemetry system can act tooth voltage waveform as depicted in Figure independently. The fact that the local charged 5A was selected. The magnitude was chosen on particles are unaffected by the RF field will be the basis of the anticipated electron energies, discussed later and the slope and repetition rate by measureThe dumbbell probe is ejected from the ment localization needs, the quantity of data launching rocket's nose-cone at about 80 km, to desired, and the frequency response of the cureffect the desired isolation in the ionosphere rent detector and telemetry system. Figure 5B represets the elemetr sorc w ic h emly o earia t e frmte ok the~: du m b el a s: a: nearyidea di~pol anen.Eetialcaatrsis.Teetset The t e l m e t y fequncye m p oye i s fici e n l ditrinan osbe ht-lti urn h~~~~~~~~iYigh ( prximaey20M/)s ha h fet t be discse later a~ tw-scto sa prb ytmadRFtlmtysse cnattohvlag aeoma eicted i F igur

!!::::. - -, —: — m,:!::,:`-:-: -,,.'I" —-:-,:!,::::;.,::.I. -- —..,,. :., - l7n, 11.1.1 ll ,- ,.` 1`11...,, -., -,,,.,3F!n,,.,'.,,-,,`t.,.,, — -..: -..... 17 - - lf — ,-, I:,:... ",,, -, --'':..... ".,-,Z.'...'.......,,7.,'e.:! ,.,.. . -,:-,., ---, — :, -,.;!.:-..-j::.,7 —.11.II... -,,.,,-.I ..,..k __,"'....."..,..". —.:......I,,.. ...:::.,:::::::::.,::,.....7 :,II.I-::::::,i-:,......7,,, —: —,:::!iwi.,.,.''-.:., -..,. I. I..-_:, -,:::, .-.-I,. -,.:::....:::.:.::..,,::::.:. :.,. -...! t,,. I,-,:, "''",..::::: I.I.I::.,:,,,.-.,,,. -. -I.,.I.I , "I ,, " "" :,.....,.,!.:,.:,I-, :;:....::;i:;:.:::'- .... I.II..I..- -.. -,.:i ,:,..III..,,::..': I., %:,:,:,-i:.:::...:,,:..,.. — I...I..-,..-I I:l.- i, —i...I.." 1:.,':I.- . I:........,,::: -I.::, II-...-..::..- % —:-, :, I-,::..:::.., 0. :I1....- 1... I1..I,,,......i ,.-.-......, - -.- --—.! —,., :.:: :-::..:. .I. -...!..I — ".. 10).....''-.m. II, - -, —-,-.....::.,.. 1, 11I I.. I I -!:!, 1,.I Ii:,:.-,: I, —-,: :".,::-,..,::,."::,LN..l...('it.i",..,.:..:,....-.mll.-. I..-.1 I. i..,::,:.!,I. -I.I..:,...-:, .::,!:... -6..... ,-I. -: .::::,.:!'::,: .,. -. -: .- —.4,,::.,."...II....-.......qm.:.. .I..I — 1-11 ,. -, :..i:i,,.::;...,,:-,...-:-q- I, ,.::.!: 1:..:,:..:...::::.::,:::. -!.,,:.,,..:,::.,... I- I ...:.:;.,.1 I-... -..I11 I .,::,b. -,., I - —-.,I.- I I..: .. I,. -.1 I. .:.I,1- I -I III:. 1...III,:,:::,.. ":.-.:.,:i::.:-...-,,.. 1...,::,-:-:,.:,:,::..::i,:...::...,.,:,,,:,::.::".'..,.i..i!", I.. I.. t...:.:!:::7 -.'",,:!,:.. -: II,.,,,,:.:::::::-.:I. 11 11 .-.-I.X1, -:-:::,:::,'..] -II.-..,:I —..,.a ::-x 5-:,..,:--, -,...IIII...::..,,..q,.I.......1.1 - ::]: i:::*i::..,....". I.1 I..,..I-,..:. .,..- ,,,..,.I.. -:;::;:..ldl.-.,i:.::: : .:::: ,.,:,. 1, ::,i -,..,::x:;:::0,,,,i,!!:...,,..1.11 —:,.. —.. ----- -— ,-,,-,,,i,,,- —.,!, — :-,.. 1.: - -ii i';':.. .,&,:,::;,..I,. ::- :, i, I," :.: :": .-::::.,, ,:... ,..,,II,\ -.::iii,.,..,::.:.;, ,,:..,.,..I: ::: j.. —-,::.::.:: .,-. —:.:7 -,- -.. -. -'..", -— -:,....,.f — ,,i.- -'-:.l.. II-:I- .,;!..,,-,- -,-,i!::i::,, ,:-`%,,,.,:.: ::: I,-.,4.-, - ,, — -, —, *;:,,i,,,.,,,::".i —*O.;I,....,,.......,..-,',,, incidence:.:-,.:i, -., —;;::. IIII " I1!:. i:: W.,... A.,"..'k,,".."'-.::::.,: i, ,,.ix:: -:z`:::,-:,::.,,,,,...,..- -,, -— -...:,.i;,,:ii::::i::. -,, 0 " :..- -., ".-..%,. -., — X-0 —..:-: 11,II.,''. - -::: —..,:::.::::;,.::".::'k,'.,....,.............,:.:...:.*::i,,:....:::. —.11, -,:,"' "r, —i-O I\'I:,:%.,,.,...'t....-A; I,,,'..,.,,,,:,,:.:-.,,-...,.-..:1;.,-.,-i.-A -i -n —::,:,...'......,., ,:,,.: .;,,-:::: .....,.,`:: -. ......,.. :., ,.,.,:.. -.. I.,..,.,.I..-.j.-1It.*'.,:-:.:-,-,... -...,,-S::I:::::::-,:.:.:: X:::i,:z —,:i,:,:::: -:. -:,`:,.-##...,-, -,-, ..'J'All. :::...,-...-:,.,.... - -.. ,"",,-...-., II...IIs.4.,, —,,:....,.....'. -* -, i....I III....., -O.,. , -,..,: .- -, ,.,::,,,:i,-: :.. ; 1.:,.:,!i: -I:,,,I-,,,,,.,,,i-,,,-,,7, -,-7 -,...;..-,-,,,,,,:i,,,.-I-N,-.., X.: j.,...1 I.,I.% - -<, -, ,-,-, 4 - -,M~A "..''-..'.',:i.ii;-.. I..:::,: O- -; —,,-k-:$i. .A — -':.,.-,,,.I...:,-.: — —,.:.."..,", -,,x,.,-. "'..',,,,,,, .....'. I... I... 14 ........ 1-...-.-"*Lq,-,.,-: 11:..... —,"- ix 5 1,9 *,,,.Al —,,,, -..*. — — :,,: I.*# —-f!i >"- --. I,! -.,*:.,*..:... -i* -i::,16, — i!;!i i,, .11 - I I,.... 0,.....-M.,.l. !;, "I" ........ i.4,-.j-:..- .,,.I I J,-,-,,:,i:7:,, "",,1,.V...-,.I ..,,.,,\,,- I,.,! —,,,... I. J, V,. ..., k.,II-I,I".. I:,:, :::.,:..:.I —`..:,-11,.:.:,.I:.,,. . ...: —.x,, -!:,::,I,.: :.:l,:..!: - : :,::i.i:. : ,.:. II:I,..I. -,,,,,,, ""i"".,:]i,,,,.,,,.',.,. II:I..,.,-,..,,, -- ii ! ,- -, ,,,\ —, -..,-..,,.ii...',. ,.::i'i::::,,,,.,,-.; -.,,-,..,I.I.. :.-:.,.\ -, —1 ,:.. Q..I-::,:.... 11`"' ,.-, -: -, -,,:::,,,, ,-,.i,,*,..i,. K,,,,];-,-,:Q,,i: ..-,,.,.i,.,." -x-:-..-;-,. —,.:: ..::,., "...,-,,,:: -,.,,.K,:,,,..;:,. -. —ii-,i,,N —-, : ii... .:.,,, iii,, —,,,.,-;,..",,,,,`0-k,,,,",, "' —. I,- -.-.-.,.,il,i,.,,.ii.i,. It "......,"'.... It'I"We:,, -;:-..11,::,:,, !,: ,.". —-m,,,,:, .":.,,',,i,-.".\-.A -11l.'k",.-,.,-1".'.'..''.'."". -,,,Pi ","..'.,';.-'..i..., I,-,., -i -,', -— iii!,,.,:,-;. —,,.,- -,-`i,.-.I- -. 1,,:,..., k,,.r..N... l-:.,.,N,-.i: -, - -, -..,-.:...,.,,..::..- v,,","., " "...,,..,.,.i, ,:,.,,,,,.. . II,%,i — —, -,I:::..., -,.....'...... .::I -i; -,!,:m:,,:,.. — , —, 1.1,...,.1.:.:.::..,:::::.',,,i,,-..-1 -, .,,,,,&, -'..,.,'-':--"'...,!.-.r.i:.: :.. I 1i.!,;i,,,:....,;:::::-N:;:-...::.I I-i: :,I.,, —:i:,-\,::,.I, I:,,:,,., ::.... il.,I:,A,..: I.- -:,-;j:,:,:.-.,.-1- -.,"''..";"kl.,,.-i:,.i,.i-,.....,:-.....''... I:...:'...,...!",.".i.-ii,"..',."-,,'.-,,..'....'....;- 11,,,:.,,-."..- -.,,.I-..].:,.",::,.,,..-...-.i'..,-i:7:x -:.:,!-,.:::,.,. ... -,.,-,:::: 1. -...:: I — ... - I,,-b.,.,-:.Xl:-X-l::.,i::.,,-:,.1 ,,! -,,-,,,', -.. -I,....,.".,..,,.,,,.""...",., -, —,, —,.\,-,..,,. -',-',.-,,.'\".">".'-'.,.: :!, . -:::-, —,.,..,>. ,. pi:l,,,j...,:]*::::i,...I.. _, ...\ ,,...,.... I.... , ,:, i,,, ,... I.. ,-.1:;;.,..."..II1......-.... i-,::,:;:,:;:,II-,.. I I.. 1.... II-", , I ,,, 1. 1, :,.I,I ...:-;:::.. I-:........ II....; ".,.,,:- m. -.-,.:...- -.,.:." 5.:,::.,:; -,-:::m,, —-— :;!...;:.::..,;ALI-,,::::::. -..::::;t::,.I.. I... x, kX. - -....... :-...:.,::.. c. I.. —I-.-W:....,.,......................1. -. II -V,:.:,.:.,-.-.,,-,. --,,\%..... -::,", —:-,:-k --., —,".. I. -4-:,,,::::i,il,.-,i."...,:,-..,.'I.",',,:'..,.E 1',,,',.,,..'-,,.,,,.,,,,.:iFWZI',"....:.,:.:. FM ":.N.:0?...,...., I. . -N.. I1. ". MI,,..i old I I I II...... - - - -....-,-,:,.,`,-,9M.. I..,.,-, i -.........,.. - V. -a ,.... -:IN -XYl,,.!:,,, I,, -,...,,..,.:.;.-..,............ 00.-.- -— l.l,,..-.- - —:I..,......:..I,...- -- I.. -- - b... I I j,:.11 m,u.., —.1 —,..-.,..:i -:,i!, ——;i, -, -.. - I,,,I- -, ,,,.,.I- Iq..... 5., -;.,,.:i —— .: —......'',-.%'%."lii..,.,.,:..., -...,....,.,. II....:Additional -:,.... .,5:,,',..'.. - —, -, -;,k.I... —,-,..v:.is,:,......I:-,,,,:,,-:]i::.. - — ...:, —,,.\ - -, .,*:.::*..-...:. ::,:,... iiii -- - -Ali-......,,-,:i,.:..;...,,,Z.,'.i"-%.'.i.I.:I.,........ ,4. —k-M,..., ~; -,,,,....,,i<,,.\.:,-.-,,',,,.,,',..-,.-,.-,.,...., —.,,-..,,.., ,,..,%I., — I...'..-,.....,..:....,.:!1.1... -,...,.-.0.,Ni%.,,:7,,,:--,,, -k,,,-,,-.,,.,..",....iii'..":,,:,-",, —.,.i -A 0, I.,.-X.:..,VIi -,.-, R. ,, w.,,:,,.,,All...M......,s.,..'....,........... i.\R,. e,, -.....,:,... QQ,,,. u,... N.,,,, —,g, T, 1..."M A1.111,,,....'.....R.., x,\,I.,, -:,.:....... i..,. I "...:.. -,-...-,:.. . I.....,,...$ MM, I" M, -.1 - -, -.,,,,z-.-i —,ii,-!!!,. -.. I IN,. ... -:-.1..-.:.-.-., -.; - "..\. I..*..., " A,,....l.,,.;k - -..".. I::. M... ".- -,.......:., ,.d..1.,.-.. n..., I II,..:.-" Z ", . VMVKaN4,i -I R.,. —,,, " - ,MX, -,.0.11:;, I x- R,, M...., - R -, -RM";..-...i!. 5 -N-M., M.,,.\ ..... ii. .-.-.,,,..-.., SM..:.,. .... I:, -,- ,". -,%,.,.:. -, 5-:-'..,.,,..,..., - - .. %A%-MO- -,Q,'. -, 1, -—,... MI,..M—,a:..... -,, .....,,,, "",-.,-N, """` .. -!,,., - .;:- ,k,-,K-:.,,.. --, :!,..., O"..,j.,,e,,::;7 qqI -1,,,:i,`iii ".,....",.;,.. 1.......- X 11, ,-,W,xOW -.03:,-,:.:_..,::;.,,.:,:::::%:.::::,....3m....'R....,,.3,1111...,, k, .,..-"".","." -—... -.-..,..-_ - ii::."".."all.., .... i."' --- I-E-..:%.. - "...., I..". ::w.,.- .i:......I.. .!:,...I, "' $.., ,-:-i:-,-f.-,.-,,.a.e "'k........R.n......,.. 1.1m.,.., I.1..II.,-: :I, - —-S..",".",,.,......,-,--..0", .,..;:.".:'::',-'.'',...",-,,,m-.l,:,;,:,IN,-4,,:1:.I.., —,-`,. - I,-z-:,. ., ".. - -I...... ,i:,,,\N.. II. -., .i;,. -,,,. -i,,, -..I,.......:..."'. 1..I. I.,%:,-:.:: - - ,,.- ,,*,-.,.;,,g.-, %- -—!,.,.,&,I.1. .,.,..,,,";,,,- ,~,.-...-.i.` -,t g..'NO:.. .,...,.:.. Ilk -.1-1.1........... -..-.,. —-..l-..-. —..,,.:..t, ""' 1'. R —a:..,.....,...... ,:,.,.,::::, 1...# I".,.,. - —--.-.1,I.,-""' -...:\.".".\,-..'.....;,11.'.'.. ;,, 1..1:2:,,-,,,,, —-A,,-,,O,., —,:!:....,.. :,::,..;:..-.-..,,....W...:,.w i-ft.-,-,-.....'',,-.,.,.,!;;,$.:.-:..1 ,-.-.1,..,,..:I-.,.-..."..;,....,::.I 11 I. .. I:".-,-. -.... iim.,.,,,, I.: -1 I.,4.:-,-,,,-I,-.III*1... -I — -, - "-::.a. —l..:..l.: 11..::, —- —, —-V-,.%...,,.,,..: -%x.. -,,..,- I... I.a........ 111-1-11-11 Ir. -.".. .,. -..-,-: 4-;..iw.,. ..A75 40": I, -::,- V,k- K., I... —-. - -i&-..,.., -, -. " il,,i, —,. —, —-,.il-o w, S.:::,, -,.-,,,,M-,:, --.. ib . RRVA,, ,... I.:.%R 1-1-:1-:...,,.......I.7..1 I I. I9,-...... I I......... --...... "' .-..... -. — ......-I-X:....,.:.... I I,::,:., , Ili, -- - - 14 i mo..." - I.I..,;.,. '-..'.-'.",-;.,-,7,,..:.:,.ff::,-;:<;"'I ;&... -, —,-,,,!..I;k. I.:..:.. M.,,rn...7., la,. —J —— -...:....:..,, -.:. I:::I — A,%,.".., I..-::::.,.,,-, ".g.m. "I 1,:!!-,.,`.,......t.....K —-, —— k-,K.......IV I..,:,.,..`:-- -.1.-I... -.l.-.l-l.-.l. —-.:I I.-:-b.. --:-.;: - I. .. 1......... "...k. -j::.::: Mm:,... WI`... ......;.Sm"',! I ^,w::: .::.. -,.9.11.: -qllll ... .-.:.. 0.. I.W I I -,.:. -ii--:.,::?:,I,.:.0..::::::.;.i.I;l. y.-IX;....,::. I....O...71...... I.,.i:: ::. ..-....7!,-...I.w:-,.,.1q.,.-.. -..I.....:.-:-::. wl.,-l-... -...:,.. 1..I..":1I:::,i:,,*,, 1..I..,-i. 2. :,-::i-ii..:-.:!'i','': :..,.-, -::],:....,:::::i:i :...:- -.. .1..,...-.1-:1:.,..!:v. ell, ... -...... -..;.11 "...,:: -.-,"$ CD... ... --- -.. I.1.. 1... I.. ll,-,.:-. —..,-.,,.-,.- -,:. :,.....,, .. e lf,::: —,-.:! 11.1'...i: —---------—, — —,: ——.:,..,. 1..I.-I:::,:::::.., - I I..... (1) -,.- -.;.... q 1..:....,::, .,-...., -..II:-i,:::];:;::ii, 0,::,:.,..,.......... ::-........:- ,;i-,iimi.!

tt NASA 602 190Km.. Fit t.. +. *ht~ pi of teeer cod hwn h feto rb |oiyo h ubelvl-meec~ntr

NONTHERMAL EQUIIjIBlIJ UM IN TIHE IONOSPHI'IIERE 165 DENSITY (105 PARTICLES/CC) 1. The electron tlemperature determination: 350( 2 4 6 Figure 7 illustrates tie manner in which the N, Tt electron current vs. 8V is derived from a dumb-,.~~;I bell volt-ampere characteristic. From graphs,y6 /^~ ~ such as shown in Figure 7, the electron current300 - vs.-probe voltage V is obtained. Referring to r~/,-:2 equation 2 and taking the logarithm |/ | In I. = In [AN,eV/kT./2rm] - V/Vo (6) t 250 - UPLEG/:DOWNLEG one obtains a linear equation in V (probe volt>L ~ /yW~.-z~ (-~ )age) and In I. (electron current) whose slope is 03 V Vo = e/kT.. Since k and e are known constants, ^ ^ Y, —" /: — T. can be determined. Hence, each volt-ampere ~ 200 - ) UPLEG characteristic measured may be interpreted in I,,,' DOWNLEG;) terms of the electron temperature of the plasma. < ~. -; 2. The ion density determination: Since the / /' /NASA 6.01 ion-saturated region of the curve is due solely 150 _ / i! X/ FORT CHURCHILL, CANADA / - \ 1F526 CST, 6 MARCH 1960 to the collection of ions, current measured in'? a {' / 1 ELECTRON TEMPERATURE (,: / ] A,0AND this region at particular voltages is a measure of ) i' ( { POSITIVE ION DENSITY the ion flux at the sheath edge (essentially the undisturbed plasma). Thus, using the known 100'' potential V, one computes the sheath effect 7 i term, (a/r)', the current density at the sheath 0 1000 2000 3000 ELECTRON TEMPERATURE (K) Fig. 9. Electron temperature and ion density for NASA 6.01. DENSITY (105 PARTICLES/CC) 3 2 4 6 8 10 300 I 1 1 I I I i IONOGRAM illustrates a typical predicted current character- DOWN istic, and Figure 5C shows the expected current / as it would appear on the telemetry record, the 250 negative-going portion reversed in position be- / cause the ring-modulator current detector em- NI ployed is magnitude sensitive only. Figure 6 is a I portion of a flight telemetry record shown to > permit comparison of the predicted and ob- 200 O oWN served characteristics. The upper trace, as la- / beled, is the output indication of a current de- / EIa tector having about 4-microampere full-scale O sensitivity. The lower trace results from a E150 - 1( series-connected, 1-microampere detector which < is quickly saturated by the currents encountered / NASA 6.02 in the F region but which serves well for E- 1/ l FORT CHURCHILL, CANADA region measurements. The ion-saturated regions 1656 CSTT, T JUNE 1960 100 ELECTRON TEMPERATURE of the characteristic noted in the discussion rela- I/ AND POSITIVE ION DENSITY tive to Figures 2 and 3 are clearly evident in the POSITIVE ION DENSITY flight record. Data reduction. The analysis of the data loo 2000 3000 provided by curves such as shown in Figure 6 is ELECTRON TEMPERATURE f(K) straightforward, and the theoretical basis for Fig. 10. Electron temperature and ion density the analysis is illustrated by the following. for NASA 6.02.

166 SPENCER, BRACE, AND CARIGNAN DENSITY (1IO PARTICLES/CC) m the reduction of dumbbell probe data for the 0 2 4 6 8 10 12 14 6 simplest case as illustrated in Figure 7, which I iZ~ I|A~~ X || shows a volt-ampere curve recorded near the ^Q~400~ L \ ~~\ m~ /peak of the trajectory of the flight noted. DOWN, UN PREIMINARY At altitudes 50 to 100 km below the apogee, N \g / i\the rocket velocity becomes appreciable in comBRL \ parison with the average ion velocity. The result 2-FREO of BEACON \ is an increasing distortion of the sheath and of 350, ~ the ion current characteristic of a sphere. The iI! \ | | current becomes dependent in part upon the li/ag\~~~ 5angular orientation of the probe with respect to,' / ithe velocity vector. Figure 8 is a photograph of 300 - 1 UPLE two portions of the telemetry record of a flight Il / in a region where the probe velocity is signifiv,-~ j i/ / *I~ cantly higher than the ion velocity. These curves W.' \ I jDOWNLE should be compared with Figure 6 which shows x 250 - an apogee case where only the horizontal com|~250r l~~~ \:1ponent of the velocity remains. It should be.i"~ \ }j~~ \ noted that these distortions in the current char~o,1/ acteristic are due only to distortion of the ion 5 200 _ /component of the current so that the electron DENSITY (105 PARTICLES/CC) 0 4 8 12 16 l50, ~r@!I I I i / i 350 - NASA 603 K \ }350 - ~~/ ELE\~~5 - fT (CYLINDER) WALLOPS ISL. VA. TYLINDER) 1126 EST., 3 AUG 1960 NE ELECTRON TEMPERATURE BRL 2-FREQ. ELECTRON TEMPERATURE (OK) N er TEj Inn ~'- &/ /~ POSITIVE ION AND2 - r/! I I_ \ \ i 300- )] ^ ~ELECTRON TEMPERATURE (~K)^~o. /PRELIMINARY Reference to equation 4 shows that the N, re- 20 duction requires that the ion temperature be measured or assumed. Since the data presented Ibelow indicate nonthermal equilibrium at these altitudes, it is not entirely valid to use the meas- NASA 6.04 uredig. 11. Electron temperature in the density, andlcul- WALLOPS ISLAND VA. 1156 EST. 26 MARCH 1961 temperature available in these data, the author NASA 6.03. PITIE 2 ONLERO have elected to use it and make corrections later 00 edge, J,, and hence the ambient ion density N9. ^ // OATA Reference to equation 4 shows that the N, re- - 200 - C~" / / / duction requires that the ion temperature relations are better unmeasured or assumed. Since error thata presults is not exremely 0 2000 below indicate nonthermal equilibrium at thesef/ j f altitudes, it is not entirely valid to use the meas-n _ -y / / NASA 6.04 ured electron temperature in the density calcula- 1 / / WALLOPS ISLAND. VA. temperature available in these data, the authorsa ELECTR have elected to use it and make coelrrections laternsity for N when the temperature relations are better un-\ \ \ \ l \ \ derstood. The error that results is not extremely 10ooo 2000 3000 serious, however, since T, enters as the squareELECTRON TEMPERATURE c) root. Fig. l2. Electron temperature, ion density, and The above discussion indicates the major steps electron density for NASA 6.04.

NONTHERMAL EQUILIBRIUM IN THE IONOSPHERE 167 meters, the rocket velocity is always negligibly I )small'in comparison with electron velocity. 400 _ I Ionosphere parameters measured by the C1959 MO W.T IAG.'dumbbell' technique. The electron temperaARDC 1959 MODEL W. UG.'60 l TC ture and ion density data obtained from four W.I. MAR.61 dumbbell flights are shown in Figures 9, 10, 11, 350 1 \12, and 13. All four flights took place during the I J ) daytime, under varying ionosphere conditions, at i: / somewhat different times of the year, and at two different latitudes: in the auroral zone at Fort F. CHUR. 0 300 -_ T / MAR.60 Churchill, Manitoba, Canada, (59~N) and at W \ FT. CHUR. Wallops Island, Virginia (380N). The statistics ^ \ V~JUN~6 of the flights are summarized in Table 1, and s \ 10-km interval data are tabulated in Tables 2 - 250 L )tJ |and 3. W 250 - / \)Each profile represents a series of data points,,~~~~5 | y~ ~ ~ jj/each point being obtained from a running aver/ y^ 2 TVage of many individual measurements. Figure 14 200 r [iSUMMER'57 illustrates a typical averaging situation. Many /IFALy HOROWIT volt-ampere curves (1-3000) are obtained for.5 1 / I a l'5. B / LAGOW each rocket flight during ascent and descent, / // GAS TEMP. }/,w / yFT. CHUR. and each curve yields an electron temperature. so._.//,' The expanded sections of the curve of Figure 14 7 1/?- 0COMPARISON OF show the individual electron temperature data GAS _ ELECTRON points obtained during two typical 10-km altiTEMPERATURE u~^/J28~ \ 8)/ tude slices. It should be noted for a typical case o100 X that the 12 consecutive temperatures are measured in approximately 2 seconds so that each o 500 1000 1500 2000 2500 3000 mean temperature represents an average over 2 TEMPERATURE (~K) seconds of flight, but each individual point a Fig. 13. Summary of electron temperature data. very small altitude interval. Figure 11 also shows curves of electron density obtained through use of a two-frequency current, obtained in the same manner as shown beacon' on the same rocket from which the in Figure 7, as explained earlier, may still be dumbbell was ejected. Similar data obtained used to obtain electron temperature. On the from flight NASA 6.04 are shown for comparison other hand, ion density data reduction must in Figure 12. Additional data resulting from consider the effects of velocity. simultaneously obtained ionograms are also preRecently reported studies of volt-ampere rela- sented for comparison. tionships that take into account a superimposed Several conclusions seem apparent upon indirected ion velocity [Hoegy and Brace, 1961], spection of the data: and extensive analysis of the data obtained have (a) The electron temperatures are signifishown, as noted above, that the electron tem- cantly greater than generally accepted neutral perature data reduction can be made without particle temperatures, particularly for disturbed introducing significant error even though there ionosphere conditions (NASA 6.01, 6.02, 6.03), is an appreciable velocity component. A demon- where a factor of 2 is generally apparent in the stration of the validity of this statement is not temperatures. The authors have chosen to inappropriate to this paper and the reader is referred to the report noted. It should be observed, in addition, that although the ion veloc- Equipment supplied and operated, and data reduced and provided, through courtesy of W. W. ity and rocket velocity are of the same order in Berning, Aberdeen Proving Ground, Aberdeen, vertical sounding flights to several hundred kilo- Maryland.

168 SPENCER, BRACE, AND CARIGNAN TABLE 1. Dumbbell Probe Rocket Statistics Rocket Serial No. Date Time Location Ionospheric Condition Rocket NASA 6.01 3/16/60 1526 CST Ft. Churchill Aur. Zone, spread F Aerobee 300 NASA 6.02 6/15/60 1656 CST Ft. Churchill Aur. Zone, near quiet Aerobee 300 NASA 6.03 8/ 3/60 1126 EST Wallops Is. Magnetic storm end Aerobee 300 NASA 6.04 3/26/61 1156 EST Wallops Is. Quiet Aerobee 300 TABLE 2. Tabulated Dumbbell Probe Data NASA 6.01 NASA 6.02 NASA 6.03 NASA 6.04 Alt., Upleg, Downleg, Upleg, Downleg, Upleg, Downleg, Upleg, Downleg, km T,(~K) T,(~K) T.(~K) T.(~K) T,(~K) T,(~K) T,(~K) T,(~K) 80 90 900 900 100 1160 1270 1100 1140 970 110 1430 1360 1150 1100 750 1460 400 120 1410 1070 1155 1140 1390 1570 600 130 1490 1140 1150 1235 1870 1430 700 140 1520 1240 1245 1245 1750 1395 780 1170 150 1410 1370 1410 1360 1770 1515 880 1330 160 1370 1440 1650 1670 1800 1780 1090 1580 170 1700 1440 1820 1840 1795 1850 1200 1660 180 2040 1920 2010 2040 1875 1930 1250 1850 190 2350 1810 2160 2300 2040 2120 1310 1910 200 2330 2050 2300 2425 2200 2370 1570 1990 2i0 2450 2340 2530 2600 2300 2545 1870 2150 220 2625 2330 2600 2745 2335 2620 2090 2260 23( 2650 2530 2675 2765 2295 2630 2270 2300 240 2780 2470 2700 2745 2210 2620 2390 2440 250 2800 2780 2730 2755 2185 2625 2470 2500 260 2900 2880 2710 2760 2200 2690 2510 2350 270 2830 2640 2740 2765 2260 2775 2375 2210 280 2970 2720 2815 2880 2365 2860 2200 2180 290 2900 2880 2870 2920 2460 2910 2080 2100 300 2880 2960 2560 2935 1990 2040 310 2970 2900 2615 2960 1960 2010 320 3000 3040 2625 2980 1975 2020 330 3165 3010 2625 2980 2040 2050 340 2630 2975 2120 2120 350 2615 2955 2120 2120 360 2595 2925 2030 2030 370 2615 2880 380 2660 2845 390 2715 2825 400 2790 2870 410 2860 2910 420 2875 2920

NONTHERMAL EQUILIBRIUM IN THE IONOSPHERE 169 elude the 1959 ARDC temperature curve (see the time of a solar cycle maximum, at high latiFig. 13) for comparison as it has become a fa- tude. miliar reference curve. This curve is not believed (b) There is a significant difference between to be seriously in error, being possibly somewhat electron temperatures for a quiet and for a low, for daytime conditions in the region shown. disturbed ionosphere. Also, there appear to be The general accuracy is borne out by an increas- greater secondary (fine structure) variations at ing amount of published and unpublished data Fort Churchill under spread F as compared with obtained from satellite drag, sodium releases, quiet conditions. and ion density profiles that show somewhat higher temperatures. The direct measurement (c) There are clear differences between ascent data of Horowitz and LaGow are also included and descent vals of Figure 12 data, particufor comparison. These apparently high gas tem- larly in the F, region. It should be noted, howperatures are not explained except to note that ever, that there exist both time and spatial their data result from measurements made at differences between ascent and descent paths, TABLE 3. Tabulated Dumbbell Probe Data NASA 6.01 NASA 6.02 NASA 6.03 NASA 6.04 Alt., Upleg, Downleg, Upleg, Downleg, Upleg, Downleg, Upleg, Downleg, km N. X 106 N, X 106 Np X 10 Np X 106 Np X 106 Np X 10 N X 106 N X 10 80 0 90 0.15 0.34 1.00 100 0.60 0.61 0.80 1.40 1.60 110.71 1.02 1.15 0.85 1.48 1.80 120.83 1.43 1.40 1.00 1.56 1.71 2.10 130.72 1.55 1.37 1.23 1.70 1.90 2.30 140.86 1.58 1.84 1.60 1.95 2.18 2.55 150 1.28 1.68 2.20 2.00 2.30 2.45 2.90 160 1.48 1.94 2.20 2.30 2.70 2.90 3.40 170 1.47 2.38 2.58 2.70 3.10 3.00 4.35 180 1.75 2.15 2.90 3.01 3.20 3.00 5.00 190 1.70 3.02 3.41 3.40 3.50 3.17 5.60 200 2.20 2.90 3.56 3.40 3.70 3.40 6.40 210 2.15 3.23 3.64 3.42 3.80 3.47 7.20 220 2.45 3.54 3.81 3.81 3.75 3.40 7.80 230 2.40 3.53 4.17 4.00 3.70 3.50 8.30 240 2.50 4.00 4.80 4.39 3.60 3.60 8.65 250 2.90 3.15 5.23 5.00 3.65 3.80 9.10 260 3.25 3.32 6.20 5.57 3.83 4.00 10.10 270 3.95 3.45 6.82 6.50 4.02 4.22 12.30 280 3.97 3.95 7.66 7.08 4.20 4.50 13.20 290 4.45 4.23 8.30 7.85 4.39 4.73 15.00 300 5.00 4.75 4.40 4.68 15.60 310 4.80 5.10 4.50 4.40 15.80 320 5.00 4.87 4.59 4.40 15.80 330 5.50 5.20 4.30 4.40 15.60 340 4.20 4.13 15.00 350 4.20 4.00 14.40 360 4.25 3.95 13.80 370 4.20 3.98 380 4.05 4.10 390 3.90 4.18 400 3.80 3.84 410 3.84 3.72

170 SPENCER, BRACE, AND CARIGNAN - - - - __.26 0 -__ __ __ _ J __ __ __ __ _ NASA 6.02 - - - 279? REPRESENTATIVE EXPANDED VIEWS I% OF V-A CURVE T[ AND AVERAGED TE 270 - KEY: *-CURVE T: -— 2772. * 77 * --- -~o-AYERAGE OF 12 ADJACENT ~''CURVE TE VALUES 276 300 275'. —'_ 2-90 274 ----- - 280 ---— 2743 — 4-/ * X ~ 1 V IE<EXPANDED VIEW 270 _ 272-.260 271-... 25 270 -- -- 2400 2600 2800 3000 3200 240 -___! 230 — o t 220C - mine -n -0 km —---- -- -aFt_ _u il fo wi a S 210C - - - 66 — - - -- - - ---- t- / -- - -- _ - -- -r 200 -r/ r16 7EXPANDED VIEW 160 --- -- -- - 163 150 1-.- --- 82 140 -- -- 181 130 - -- IO 1600 800 2000 2200 2400 2600 120 —-- 0 400 800 1200 1600 2000 2400 2800 3200 ELECTRON TEMPERATURE (DEG. K) Fig. 14. An expanded curve for NASA 6.02. typically 8 minutes and 200 km horizontal sepa- single case at Fort Churchill for which compararation for the E-region data. tive data were available. (d) There is good agreement between electron (e) A distinct temperature inversion is apdensity data (as derived from both beacon and parent above 250 km in the Wallops Island ionosonde equipment) and ion density data (Fig. data; see Figures 12 and 13. 12) for Wallops Island but not as good for the Discussion. The fact that the quiet iono

NONTHERMAL EQUILIBRIUM IN THE IONOSPHERE 171.'.'**^''' -^*:*'' I' IASA 6.01.: been found to be in very good agreement with L:.. Np P.Or<- rs/' G 320 Km., ] predictions. To provide a consistency check of the f Ac~, e'..4~-.^'-I- o' Ltechnique, a typical volt-ampere curve as obNp:: " i.. served in flight was reduced to electron tempera-'(.,:' ture and ion density. These data were then used |1 0>0^Q Hi. I'^; /to compute the volt-ampere curve that a dumb2 bell should exhibit under the observed condiI i i;: ~i' ^* tions. Figure 15 shows the results of the compu-'I ~... (........ \;.7 _- i..tations drawn on a photograph of the portion of I - ~: - \?'/~-'-:.-.. the telemetry record from which the data were 2.47 -1.85 -1.235.614 0.;' deduced. The computed volt-ampere curves VOLTS i, VOLTAE' 01i1qE'~:i4: ~ L:..:iS 9?,t^g uETWEE o s:!; shown correspond to two values of ion density: ___.; _______'_:______:....N, = 7 X 10' ions/cc, and N, = 6 X 10": j;;"::':.l".".::~.: Ik:..'.': ".:ions/cc, the latter value being the density originally deduced from the record. The value of N, Fig. 15. A comparison of the predicted and ob- that satisfies the experimental curve is 6.4 X 10' served volt-ampere characteristic. ions/cc, indicating that, in this instance (chosen at random), the observed N, was less than 10 per cent high. This is known to be due to the sphere electron temperatures, and particularly effect of probe motion which increases the ion the disturbed ionosphere electron temperatures, current and was not considered in the N, reducare clearly greater than generally accepted neu- tion. The consistency demonstrated between extral particle temperatures (Fig. 13) seems to the perimental and theoretical curves appears to the authors to demonstrate that thermal equilibrium authors to substantiate the general validity of does not exist, at least in this region. The curve the theoretical approach and the experimental determined for quiet conditions has a negative technique. gradient at the highest altitude attained, thus A second, essentially unrelated consideration, suggesting that the neutral particle temperature concerns the electrons sampled by the dumbbell is being approached above 350 km. On the other hand, those temperatures observed under disturbed conditions show no tendency to approach the expected isothermal temperature of the neu- tral particles and, presumably, of the ions. Thus, to the authors, these data suggest that possibly one should not expect equilibrium at even higher Since these data are not in agreement with.; <.ii- i other recently published data, and the conclu- sion reached is not in accord with generally held' concepts relative to atmospheric thermal equi- librium at 350 km and higher altitudes [e.g.,:.... Jackson and Bauer, 1961], the authors have believed it particularly important to establish the validity of the technique and the data pre- sented, and, therefore, have given considerable _b i thought and effort to checking the work and providing other means of establishing confidence in the results. Figure 15 illustrates one approach taken in this regard. As noted previously, the theoretical aspects of _ the dumbbell technique have been studied exten- Fig. 16. A photograph of the NASA 6.04 dumbsively, and the general experimental results have bell showing also the small cylindrical probe.

172 SPENCER, BRACE, AND CARIGNAN''S.....~.. AMP: WSECa D! 2 07 KC~ll'OE SEC ", 2'..,'"."':r-: NASA 6.04.... /'. 190 Km. /. / / DMBEL I?.'-. / /:,,,.,.'',:"'..'.',' / ".-SYSTEM CALIBRATION ii.', -A!,_<,-.:.'~ /:', " " p -----;. —; /;:L... T O......-T.S: /:.:,::,-,..... \-...... l^,.-'.: -—:'; -"... —'-':f/:. " V X"; -,/ ^ -' i t 0 -,. i,..~ a d * - -, " ".T 4; e1 Fig. 17. A photograph of a p)ortion of the telemetry record from NASA 6.04, showing the signal from the small cylindrical probe. probe and their relation to the entire population small cylinder as one electrode and the whole of thermal electrons. Figures 2 and 3 show the dumbbell as a second electrode, to provide a individual volt-ampere curves of the hemi- probe system with a large area ratio. Thus, in spheres before and after consolidation. Figure 2 this case, the entire dumbbell was employed as is particularly useful in understanding how an the reference electrode with a typical sawtooth applied potential SV divides between the hemi- voltage applied between it and the small cylinspheres. 8V causes the hemispheres to assume drical electrode. The configuration is illustrated potentials V1 and V2, respectively, and results in in Figure 16, which shows a portion of the a net current, I.. It is clear that the negative- dumbbell used for flight NASA 6.04. The three going hemisphere'accepts' most of the applied parts of the cylindrical probe are: a spring, a potential, i.e., (V1 - V.) > (V,0 - V2) where short guard cylinder and a long collector cylinV = V1 - V2. The effect of this 8V division is der. that neither probe potential can rise much above Figure 17 shows some of the telemetry record V, so that only those electrons having energies for this cylindrical probe (volt-ampere curves), exceeding V. are sampled. The question thus during one of the many 4-second intervals when naturally arises whether the temperature as in- the cylindrical probe was in use and the dumbdicated by these electrons is equivalent to the bell circuits were being calibrated (dumbbell temperature that would result from sampling measurements were made during alternate 4-secthe entire population. ond periods). The volt-ampere characteristic of To verify the assumption, and the authors' the cylindrical probe (see Fig. 17) shows a ratio belief, that the measured electrons are repre- of electron current (B-C portion) to the ion sentative of the whole population, an independ- current (A-B portion) which is much greater ent probe experiment was carried out using a than the same ratio for the dumbbell probe (see

NONTHERMAL EQUILIBRIUM IN TIllE IONOSPHERE 173 500 -- represent new information, now readily obtainable because of advances in instrument technology. Moreover, the authors believe that the data presented make necessary a revision of the con400 - cept of thermal equilibrium in the daytime ionosphere between electrons and neutral particles, _ HANSON AND )N ASA 6.04 at least up to 420 km and particularly in a WALLOPS ISL.-MARCH 1961 c PJOHNSON PREDICTEDW S disturbed ionosphere. The variation of electron Wj TEMP PROFILE ecto I,,~ ~300!~ \~ / Itemperature observed shows good correlation o /a-T >^ ( w with the state of the ionosphere, higher and T/'.. / more locally variable temperatures being ob0 /20 - b/>served for a more disturbed ionosphere. It is 200 - - / /-T/ /- observed also that the Wallops Ihland data fI / > ~NS RC6LL-JUNE 1960 taken under quiet conditions show reasonably CHURCHILL-JUNE 1960 good agreement qualitatively with the theoreti00o - \ cal data of Hanson and Johnson [1961], see Figure 18, which were computed for conditions corresponding, approximately, to a quiet ionosphere. 0 I I I I t I I I Theoretical studies leading to the temperature 0 800 1600 2400 3200 * *, LECTRON TEMPERATURE (~K) equilibrium concept generally lead one to conFig. 18. Theoretical and experimental electron ciude that the collision frequency between electemperature data. trons and neutral particles and ions is sufficiently high throughout the ionosphere for the excess energy which an electron acquires in the Fig. 6). This is a consequence of the small cylin- ionization process to be very rapidly distributed der being driven far more positive than is possi- among the other species. It is clear, however, ble for a dumbbell hemisphere. A preliminary that the process of energy transfer between the data analysis has been accomplished and the electrons, and between ions and/or neutral parelectron temperature results of the small cylin- tides, must be much more efficient than that der probe are plotted with the dumbbell tem- between electrons and ions or neutral particles. peratures on Figure 12. This analysis is to be This is true because of the very large mass ratio considered preliminary since this was the first in the latter event. Thus, one should expect the flight using the cylinder configuration; however, electrons to be in equilibrium among themselves the authors have sufficient confidence in these and the ions and neutral particles to be in equidata to conclude that the characteristic tempera- librium with each other because of the low mass ture of electrons as measured by the dumbbell ratio. is equivalent to that of the entire thermal elec- The phenomenon of two kinds of particles tron population as measured by the cylinder. existing in equilibrium at different temperatures Although both approaches employ the elec- is observed commonly in the laboratory for trostatic probe technique it should be noted that gases at relative high density, and hence at very the current'collection' mechanism of the small high collision frequency, as compared with the cylinder probe is quite different from that of ionosphere. Thus, by analogy, one should not the dumbbell. In the dumbbell case the ion cur- expect to find complete thermal equilibrium rent is described as being'sheath area limited,' in the ionosphere. It is required only that there and for the cylindrical probe,'orbital limited.' be a source of energy that affects electrons selecConcluding remarks. This paper has pre- tively. Solar radiation does this, providing adesented a brief review of a particular approach quate energy to ionize a modest fraction of the Fto making direct measurements of electron tem- region oxygen atoms and thus to raise electrons perature and ion density in the earth's iono- to many volts of energy. sphere. Although the basic technique is not of It thus seems to the authors that the neutral recent origin, the authors believe that the data particle and the electron temperatures should

174 SPENCER, BRACE, AND CARIGNAN be expected to be different in the daytime iono- at this time, however, to support such a hysphere and that the major question is one of the pothesis. magnitude of the difference. The general factor- Further experiments are being prepared to of-2 difference observed in the data corresponds provide additional data. For example, it is conto an electron energy excess of only about 0.15 sidered very important to conduct similar measev, and since electrons resulting from He II urements in the nighttime ionosphere since equiionization of oxygen initially have 28 ev of librium should be observed at that time. Thus, energy, the excess energy observed seems entirely the next launching, planned for the fall of 1961, credible. The authors thus conclude that the is scheduled for midnight. Both small cylinder temperature results presented are reasonable and and dumbbell measurements are contemplated correctly represent the conditions which held at for this and subsequent experiments. Additional the time the measurements were made. experiments are underway that will permit Accuracy. The temperature derived from simultaneous measurements of the temperature each volt-ampere curve has a standard deviation of both neutral particles and electrons to 250 of approximately 10 per cent of the mean. The km, and that will measure electron temperature profiles obtained through the averaging tech- for altitudes up to the order of 800 km. nique have an accuracy of somewhat better than 5 per cent above 130 km. Below this level, the Acknowledgments. The early flights mentioned per cent above 130 km. Below this level, the in the paper, and on which the present series is error may be greater because the sheath condi- based, were supported by the Geophysics Research tions become more complex as the ion and elec- Directorate of the Air Force, Cambridge Research tron mean free paths become comparable to the Laboratories, and the Ballistic Research Laboraprobe dimensions. tory of the Aberdeen Proving Ground. The work and flights reported herein were supported by the The general agreement observed between the National Aeronautics and Space Administration. electron densities as obtained by the two-fre- We wish to acknowledge especially the contribuquency beacon system and the ionosondes, and tions of John Maurer, Walter Hoegy, James Findthe ion densities observed by the probe, is ex- lay, Hugh DiGiulio, and Lyle Slider, a few of the pected,*.sn it wol.se t many individuals of the Space Physics Research pected, since it would seem to be difficult to Laboratory whose efforts in theoretical work, cornexplain a difference except on a very local basis puter programming, data analysis, and equipment and under highly transient conditions because development and construction made possible the of the electric field implications of any pre- work reported. The contribution of launch support dominance of charge of one sign over the other. personnel at the Fort Churchill and Wallops Island dominance of charge of one sign over the other. ^ gratefully acknowledged. sites is also gratefully acknowledged. The much poorer agreement between electron density as determined from the ionosondes at Fort Churchill and the dumbbell ion densitiesAPPENDIX taken simultaneously (Fig. 10) has not been ex- PERTURBATIONS IN THE MEASUREMENTS plained. A similar departure observed at Wallops Island (Figs. 11, 12) above 300 km likewise 1. RF effects. Two questions have been has not been explained.' It should be noted, how- raised frequently in discussion of this work. The ever, that the ion density result is based upon an first questi on is concerned with the effect of the assumption of only atomic oxygen ions at the telemetry RF field on the local electrons. One 300- to 400-km level. If there were, for example, can compute that, at the 230-Mc/s telemetry fremore hydrogen and/or helium ions than ex- quency, totally insignificant energy is imparted pected, the high current would be due to the to electrons. To establish this experimentally, a higher ion mobility rather than a greater nmber dumbbell with a data storage system was flown of ions, and the computed density values would during the equipment development stage. Elecbe too high. There is no experimental evidence tron temperatures measured in the complete absence of any field were judged to be within 5 per cent or better of the electron temperatures 4 A possible explanation is that initial velocities measured in the presence of the field. This figof ions at the sheath edge were not consi u detered largely by the low resolution the calculations. This effect [e.g., Schulz and Brown, 1960], although relatively small, would instrumentation in use at that time. In a more tend to reduce the indicated densities somewhat. recent flight, NASA 6.04, the telemetry power

NONTHERMAL EQUILIBRIUM IN THE IONOSPHERE 175 was periodically reduced by a factor of 5, but Analysis of photoelectrons from solar extreme no differences were observed in the electron ultraviolet, J. Geophys. Research, 64, 961-969, temperatures. Hoegy, W., and L. H. Braze, Scientific Rept. JS-1, 2. Photoemission effects. A second question University of Michigar ORA Rept. 03g99-6-S, frequently asked concerns the effect on the June 1961. measurements of photoemission from the elec- Hok, G., N. W. Spencer, and W. G. Dow, Dynamic trodes. The measurement of any current im- probe measurements in the ionosphere, J. Geo.. phys. Research, 68, 235-242, 1953. balance such as might be caused by photoemis- H G... Spencer,. G. Dow, and ifHok, G., N. W. Spencer, W. G. Dow, and A. Reifsion is possible during the time when zero voltage man, Dynamic probe measurements in the ionois applied to electrodes (see Fig. 6). Typical val- sphere, University of Michigan ERI Report, ues of current that have been measured are August 1951. generally less than Hinteregger's [1959] values. Jackson, J. E., and J. A. Kane, Measurements of The maximm p urrent measured is les ionosphere electron densities using an RF probe The maximum photocurrent measured is less technique, J. Geophys. Research, 64, 1074-1075, than 10' amperes and is insignificant in compari- 1959. son with the normal ion and electron current Jackson, J., J. A. Kane, and J. C. Seddon, Ionoencountered, which is usually about 4 to 6 micro- sphere electron density measurements with the *at the F. maximum (see Fig. 15). Navy Aerobee-Hi rocket, J. Geophys. Research, amperes at the F maximum (see Fg.51 1956. Jackson, J. E., and S. J. Bauer, Rocket measureREFERENCES ment of a daytime electron density profile up to 620 km, J. Geophys. Research, 66, 30553057, Berning, W. W., A sounding rocket measurement 1961 of electron density to 1500 km, J. Geophys. Re- Johnson, C. Y., E. B. Meadows, and J.. Holmes, search, 65, 2589-2594, 1960. Ion composition of the arctic ionosphere, J. Boggess, R. L., L. H. Brace, and N. W. Spencer, Geophs. Research, 63, 443-444, 1958. Langmuir probe measurements in the iono- Seddon, J. C., Propagation measurements in the sphere, J. Geophys. Research, 64, 1627, 1959. ionosphere with the aid of rockets, J. Geophys. Dow, W. G., and A. F. Reifman, Dynamic probe Research, 68, 32-335, 1953. measurement in the ionosphere, Phys. Rev, Schulz, G. J., and S. C. Brown, Microwave study of 76, 987, 1949. positive ion collection by probes, Phys. Rev., Hanson, W., and F. S. Johnson, Electron tempera- 98 June 1960. tures in the ionosphere, Tenth International Astrophysical Colloquium, Liege, Belgium, 1961. (Manuscript received July 24, 1961; revised Hinteregger, H. E., K. R. Damon, and L. A. Hall, October 13, 1961.)

UNIVERSITY OF MICHIGAN 111111111 illlll l 3 9015 03526 7882