MASSACHUSETTS INSTITUTE OF TECHNOLOGY LINCOLN LABORATORY MILLSTONE HILL THClMSON SCATTER RESULTS FOR 1965 J. V. EVANS Group 21 TECHNICAL REPORT 474 8 DECEMBER 1969 This document has been approved for public release and sale; its distribution is unlimited. LEXINGTON MASSACHUSETTS The work reported a center in this for research wi[h the support IAF 19(628)-5167. ducument operated O( the Department This report may be reproduced E Perm]ss\on was performed by Massachusetts of to satisfj’ lsgiven the needs todestroy at Lincoln Institute Air Force Labomtory. of Tec’hnolo%b under of U. S. Government Contract agmi,s. this document ii —,.—. i: .. . ,... ABSTRACT This report presents F-region electron observed during the year 1965 at the densities, and electron and ion tempermures Millstone Hill Radar Observatory by the UHF Thomson (incoherent) scatter radar. The measurements (42. 6°N. 71. 5“W) were made over 48-hour periods that were LIsually scheduled to include the International Geophysical World days near the center of each month. occurred Geomagnetic storm sudden commencements during m,. of tbe observing periods, but do not appear to have given rise to On the other band, measuremems marked variation of the densities or mmperat.rcs. made during the progress of a major storm (15-19 pared with tbc behavior obscrv.d magnetically served quiet, in 1964. in the following 1965) exbihit large .banses c.m - month. The remaining muntbs were and the density and temperature behavior were similar to those ob- Millstone is a midlatitucfe station exhibiting a characteristic and ‘csummer” dayTime densify variation. rapidly June around equinox, thermal structure. The transition befween these tio iypes occurs without any corresponding seasonal change in the F-region Current ideas concerning tbc mechanisms responsible othe,r features of diurnal and seasonal variations arc discussed. Accepted for the Air Force Franklin C. Hudson Chief, Lincoln Laboratory Office “winter” for this and CONTENTS 111 Abstract I, II, 1 INTRODUCTION 2 EQUIPMENT 2 A, General B. The l+~cording C. Spectrum 3 System Analyzer 5 System 8 111. CIIMERVATIONS IV. DATA A. . V. F-Region Critical C. Electron and [on Temperatures (PriOrto.ltllY) D. Electron and Ion Temperatures (July and I.ater) RESULTS C, VIII. 13 Density Diurnal FOR DENSITY Plots and Nighttime N(h) Profiles AND A. Mean Daytime B. Discussion MAGNETIC Au*st B. June C. Discussion SUMMARY ION TEMPERATURES and Nighttime DISTURBANCE A. 13 i5 i6 16 ‘. 16 16 Discussion ELECTRON References PTOfile ELECTRON 13. Mean Daytime VII, 9 Frequency 13. Electron A. VI. 9 RED[JCTION Te and Ti Profiles EFFECTS and September 30 30 42 43 43 43 AND 47 CONCLUSIONS 47 49 MILLSTONE L HILL THOMSON SCATTER RESULTS FOR 1965 INTRODUCTION This report the Thomson 71,50 W). presents Earlier f963i in relation to of special *0 call,v clisturb ccl conditions. presented the results The large reports volume publication that many users period Vurther, transcribing ing toward the results Section tables of changes hcncc, of these reports, the presenting h;Iw for pub- betw of these meas- of the I.-region. or have program is a continuing precludes In this report, althmgh graphically, However, a single about 100 entries, to haw the graphs problem The remedy for analysis. figure and thereby the major in computer-usable *e we .O,1- reeognize corcring a 24-hour Pro,ride an in~mense a“ailable, even if tal,lcs encounterecl, pscsumably form, its ncwd for technical our results would prefer users there may be made available. to the instrumentation and quantity of the results, IV is a brief review were sults for electron density part of the sunspot cycle, could be averaged and ionospheric to produce i.e., that of lies in being able and we are currently The rqmrtz both the L’IIF and L-band made during the activity v?ork- that have influenced respectively. in Sec.s.11 and 111. and \,l present .Altbough the year taken during disturbed in 1964 included a summary results systems. The results o[tbe L-band Ii been reported, and will not be included here. 1 —----- re1965 low that most of the data Sec. VII1 provides obtained radar while Sees.b” was sufficiently Some data were Finally, ccmtaining the results 1965 have already procecl”res, temperatures, in Sec. VII. (steerable) procedure, made during f965 ancf are described monthly- means. and this is discussed conclusions, and observing of the data- redaction was on the rising however, and the means by which and have presented the s,ynoptic me8surenlcnt would not solve to particular versions 1965 using (426” N, this end. A number the quality each having the data into a computer to transmit during journals; most users Finally, system and temper atur’cs, for the behavior the,,, i. tal~ular fOrm. 24 tables, the radar for the: year Massachusetts s such as those made during an eclipse 9 or ma~neti- of presenting would prefer could be included. theories in which all the results begun earlier Westforcl, “ne summer and one equinoct ial month, 5-8 have discussed the significance papers of data gathered can represent sawing in space. densities observstio. in the scientific sLIcii as this, tinue the practice have described accepted currently of the “F-region Ilill, and 1964.2 Shorter for one winter, 3,4 .jourmds. Other in scientific studies at Millstone ionospheric only results urements complete of synoptic radar in this series to determine ohtiained in the years each year lished scatter reports it can be employed results the results [incoherent) periods, and some obtained using observations II. EQfJfPMENT A. General The radar summarizes “d equipment cmplo>-cci in these observations the parameters significantly the recording different of the \-ertic ally directed in f 965 than from of signals at 11,’on magnetic P.rrnittcda cycle of measurements sigm+ls reflected approximately present from 90 minutes. in Sc’c. 13below While valuable an observing effort, 2} weeks to process, A second mit both halves not recognized major ,1 major syst?m proved change made in i 965 was To explore has’ not previously recorded that could be g:lthcred ciuring a single to he analj%ed. 4&hour run usually tedious Polarization Circular of the spectrum extract Transmitter pulse lengths Pulse repetition Rece!ver fixed parabola 1600 mz power frequency bandwidths 2.5 MW O. 1-, 0.5-, and l. O-msec pulses used 50 Hz 25 kHz for O. I-msec pulses 16 kHz for 0.5- ond 1. O-msec pulses 40 k Hz for recording System temperature Post-detector integration ‘300” K (monitored -20 dB analyzer the parameters ~f! MHZ Peak transmitter required Z to to handle. PARAMETERS OF THE UHF VERTICALLY POINTING THOMSON SCATTER RADAR 70-m-diameter of it was possible This was done in a manner to subsequently Antenna cillring of a considerable TABLE I Frequency !, C! compnncnls I..ortunately, improperly. but only at the expense was also extremely macle it difficult aperture of the in a way that was not immediately change during 4965 was the reconstruction Effect;ve “[his scheme the spectra been described, to be one of the lCSS reliable processing, so that the scheme at the time, I without making use 0[ this would have taken the amount of information s~stcm spectrum Table these were of its functions. by computer of the signal radar; pln3-back ancl analysis. 1966-i 968) it malfunctioned The tapes previously.$ scatter in 30 minutes, amounts of data to he analyzed the information amount of extra outline (during large of interest increasing the rccorcii”g I.atcr obvio~] .?, causing to retrieve a brief years, tape for later Since the rccorclimg in greatly period, the equipment. previous to he completed all the altitudes has been described L-IIF Thomson system continuous y) which. to Perthough of electron and ion temperatures ing on the results, from we describe The Recording B. The receiver in the Millstone vided a bandwidth Figure (MINCOM of 250 kHz, on a iO~-inch standard coincident had seven channels of the recording was employed separate the required channels, the band, purposes, pulse, reference signals constructed tone, (3) WWV timing were 40-kHz provided ther-eafte r. reference pro- two complete This filter and (5) start/ 8top signals. IF and i 00-kHz some redundancy signals. sine-wave a !. 5-MHz this recorder data (i. e., and a rapid falloff three (2) the spectrum we employed A carefully the recorded (1) a 100-kHz At this point, at 30 ips, an hour’s edge of each transmitter for identification at IF, arrangement. to band-limit we also recorded with the leading to provide of +-inch superheterodyne and integration, operated tape (3 M-98i ). flat to within AI dB over announcements the signals When us to record a response to the wanted IF signals, to record reel 7-pole pulses has some bear- is a multiple at 200 kHz. sampling and permitted 1 is a block diagram filter radari to be centered type CMi 07). bandwidth, voice and since the difficulty ionospheric for (+) computer In order and (3) the recorder. Hill in frequency made available bandwidth tape recorder cycles) this reason this change here also. are converted outputs were analyzer, For SYStem employed in which tbe signals separate the spectra. In addition (2) sync signals, (4) As the recorder each recorded on two in the event of a poor recording. ~ E-MHz (F AMPLIFIER 2 MHz ., ? ~ “c””’ &---Bzl \ i 1 Q RECTIFIER ,0 COMPUTER SAMPLING Iwwv START? STOP VOICE ANNOUNCEMENTS Fig. 1. Block diagram of receiver output arrangements for UHF Thomson scatter radar recording system. IF signals were recorded on magnet;c tape for later spectral analysis (carried out with switch in position A). 3 -. -—-a.,,, The tape recorder commenced ber of minutes each [usually run and were feature was started on a minute mark), tbe recorder 7), used on pla,yba.k could be activated easy to play over nals from different lyzer. generator a given This timing not utilizecl in our work because, in the recorder local oscillator frequc (1.0) frequency the IF signals. the spectrum signal Thus, analyzer Two difficulties the *00-kHz LO, tone MINCOM CM !07 TAPE Instead, from reference encountered this problem, a narrow-band pulses the 10O-kHz centered at 2 Mlfz signal seriously) A second was obtained 2- P[Hz lF frequencies. 7’ and then was encountered 2-MHz REJECT CRYSTAL FILTER 0 1.8-MHz MONITOR [ ,. a- M Hz &M PL(FIER 1 ,00-kHr AMPLIFIER ii 2-..= LoCAL OSCILLATOR X2 , ,;oE0~:6E Fig. 2. Block diagram of system employed to recover IF signals from magnetic-tape recordings that eliminated undesirable frequency fluctuations introduced by changes in tape speed. 4 _——,, _,.” __ ... .__ —m. -.s,- of that of the 1.8-MHz 1“ 1 to was that ‘)dropout’; amplitude-limited, problem whose was then applied RECORDER HARD L, M,TER a in sign to those in !00 kHz JxK was the short-period spuriom as well as (less was severely This feature The recovered which rejected The first signal tone. in tape speed introduced A in Fig. 2. 100- kllz tuned amplifier. a ana- tone to synthesize but opposite of Fig. 2. with the scheme 100-kHz sufficiently This Z-MHZ ‘100-kHz wide filter used to control tone. reference Changes equal in magnitude the switch to position by a 7-pole, it of the sig- by the spectrum the recorded it did not reduce in tape speed. the spectra that could bold the tape speed con- a 17-kHz the two, a new [l? signal This “search” thus making the tapes were we used the 400-klIz that were caused the loss of timing TO minimize used to drive were by monitoring of fluctuations by throwing was band-limited governor num- thus separated Of a run. being examined pulses derived a prescribed of the tape, to analyze from at 1.8 MITT. as shown in I’ig. 2. by mixing was independent in order recovered for our purposes, speed. ?Y ci,at>+?s in this LO signal frequency times a servo-controlled stant for the full length of a recording fluctuations After tone marks directions of the timebase counted i O-vsec included The 60-Hz and reverse run several the portion generator of a run (which always this interval. to halt the tape at the beginning Tbe sync pulses altitudes. that governed The tape recorcler to the beginning was Stoppecl. for botb forward relatively timing about 20 sec prior and a 60-HZ tone marked in a manner in obtaining adequate are present put. This 1.8 -MHz filter spectral at the input to the signal will signal produce of the +.8-MHz balanced C. solved the problem, systematic mixers deteriorated at the obviated Analyzer System The spectrum analyzer employed integrators center grators spectrum. in 1964i arranged switches summed Si to S24 were and stored consisted the voltages (typically to be stored 6 minutes), ages at the integrator at the out- filtering to aging of the 1.8-MHz erroneous to cover outputs of these filters the IF signals were The resulting results crystal purity for the echo replaced The filters either in ‘1968 with were crystal were filters 500 Hz in bandwidth the upper o r lower connected This sideband to one of two inte- outputs were to the filters voltages (together integrators were automatically equal pulse with noise ) at the filter to the filters At the end of the integration to stop tbe integration measured The switches to the gated amPlifier of noise to be applied integrators. removed for a period the tm.nsmitted $ through 47, pulse of equal length applied caused a sample in the even-numbered the gate pulses applied at a delay following signal in the odd-numbered oE@- and recorded process, on punched-paper j.,,, f, FILTER ! ‘*W ‘“r= ,EC,, F,ER I d-= I % a&rED DRIVER AMP~kER .. L~ GATE PuLSES .. .. I I 1 Block diagram of spectrum analyzer employed 5 — .,—. from January 3 /,,,! at a and period and the volt- INTEGRATOR Fig. 3. of the the spectral of a hank of 24 single-pole as shown in Fig, 3. then thrown and a second point near the end of the sweep. appear Adequate shown in Fig. 2 were pulse (O. 5 m- 1,0 msec) the height of interest. outputs were at Z MHz will S1 to S24. On each sweep of the timebaee, to the length of the transmitted that defined 200 kHz or 2 MHz this difficulty. The rectified via high-speed If either spectrum. * 966, introducing mixers and spanned an 1i -kHz window that could be moved of the signal of the but owing either during The single Spectrum and 48 DC (Miller) LO signal. shift in the mean speed of the tape recorder, frequency. that Iargely i. 8 -MHz (Fig. 2), a CW signal level LO signal at the center mixer a spurious initially or to a gradual power purity for the derived 1 1963 through June 1965. tape. The punched-paper measured voltages (7 P -.s= Pn where eliminated sPectrum by recording averaged employed the system of the spectrum, made early frequencies. origin, the search to be measured simultaneously. we did not consider it desirable to be examined because, the predominant tbe available scheme source 48 integrators we developed In the modified rectifiers The the second above, since the lower half the two halves. abou< their that the effect and average the tapes of the IF signals, that would permit across the full 22-kHz instrumental between tbe filter polarity. transformer (Fig. 3) were differences scheme analyzer only the difference analyzer, to drive replaced polarized outputs were These to employ the signal effects 48 filters both windOw remained and use and the nOise rectifiers. from relays The In this manner, these rectifiers were (noise ) pulse was applied we performed to the filters treated applied shown in Fig. 4 is not immune from these, by a pair of full-wave were made nearlY the noise voltage errors as DC drifts, introduced separate on each sweep. 6 produced to the filters identical The original (S1 tO S48) that switched the signal- plus -noise with two noise pulses and, to remove rectified rectifiers both sets of diode rectifiers. by mercury subtracted between was now performed gain differences, and, believing wo decided of opposite these oppositely Systematic between of the filters the magnetic the Z4 filters outlined in the receiver), of the spectrum Accordingly, voltages ond pulse was automatically urement to analyze by the synchronous~y to examine difference to the spectrum spectrum the same speed switches amplifier Tests both halves the precautions to store to obtain a set of for this is shown in Fig. 4. providing by employing between despite as first frequency. of the response to distribute of inaccuracy. integra- only the upper half of the signal this was not certain required for a modification spectrum. and this was done by using the computer a systematic we began to measure This change doubled the time and prompted (emplOyed by the asymmetry howeve~, runs and a running signal each for a time equal to the normal here to examine about the radar At the time, was of instrumental These measurements. in 1965, disclosed this was caused determined to correct we thought it would be removed amplifier to be folded in the gains Of the necessary. in the integrators measurements described present, This pro- only noise would be present, ‘rbe computer to drift to all the other were runs were in order for DC drifts the integrators beam parametric this caLwed the signals We now believe the in the gains of the even and odd so that calibration all such drift to be applied or differences pulse out to a delay where to correct If an asymmetry p~lmped electron- halves not removed, and allowing We later corrections squared as in the shape of the measured differences in both sets of integrators. the signals results. systematic of a number of such measurements We normally center However, of the receiver, noise voltages thought it necessary spwctmun. response introduced the first We also tion time. which pulse and ~n that due to the second. errors out by moving average voltage systematic bandpass weighted removing computer at each frequency [1) due to tbe first the undesirable in each pair were carried ratio S+IJ7” , combinations. integrators using a digital (vn)z bY the overall filter/rectifier analyzed the signal-, to-noise ~~+n is the mean voltage cedure were tape was later and obtained due to the sec- by the first i.e., high- each integrator pulse. the drift meas- per sweep. by channel-to-channel calibration Again, runs in which only ,a running weighted w \ I . . . . . . M m fig. 4. average of these ment, calibration Instrumental because from observations, system noise and no further subtraction probably eliminated changes from July by the computer were to the modified major employed to correct lower analyzer 1968. each signal measure- than those of the original the need fOr twO identical spectrum were 1965 through June system integrators. was made prior The to tbe July i965 made until July 1968 when an entireIy new into operation. disadvantage is the difficulty of the modified of deciding Where system, tbe relationship quired power spectrum. signal voltage ~~ will be much larger P # analyzer in this scheme the original was brought A major of spectrum runs was obtained errors the. automatic changeover made, Mock diogr.m not recognized between the signal-to-noise ratio than the noise at the time the change was the measured is strong [ (P# voltages and the re- Pn) >> *1, the measured so that [2) = (T# n On the other hand, when the signal-to-noise ratio is small [ (P~/Pn) << 1] (3) implying that for small the transition results between are discussed signals the rectifier these two regimes. in Sec. IV-C. behaves as a square-law The consequences device. of this behavior Figure 5 shows on processing the L,NELQ RECTIFIER RECTIFIER -3 Fig. 5. 5q.are signal-to-noise III. ,.-, ,.-, UNITS (~,, n- ~n]’ ARBITRARY ,, m? of voltage (obscksa) measured by spectrum analyzer shown in Fig. 4 .s input ratio. For weak signals, stored voltage wasdire. tly proportional to signal power. OBSERVATIONS During calendar i ‘?63, observations month.4 A complete in 19642 and, as a result, recording density the IF signals profile our desire the time resolution, erally reduced to include near the middle urements made, tbe absence of major activity over a single the Priority and permitting 48-hour or Quarterly Table periods, Indeed, This was reduced The impetus through dawn and dusk. period magnetic the average to I hour U. temperature and in each month. International in We gen- Geophysical World the days on which meas- index Kp during these periods. to he averaged to obtain the mean monthly of the data it was evident of magnetic disturbances that only on the results TABLE II PROGRAM was of this increase in Sec. VII. OBSERVING By Kp index was less than 20, indicating an examination The effects per, for these c~anges As a result 111summarizes the results from in the case of June was this not permissible. are discussed 90 minutes. four times made per month as shown in Table and the mean of the planetary with some confidence. approximately tbe time to obtain a complete of each month. For the msjorit y of the observing behavior required periods in 1965 to 30 minutes. we chose to observe chose these periods 30-hour were processing, the behavior Days occurring were cycle observations for later follow made for observing fewer was further to better were AS A FUNCTION OF YEAR Year Length of Each Observing Period (hours) No. of Observing Periods per Mcmth Time Token to Measure One Profile (hours) No. of Profiles Obt.ai ned per Month 1963 30 4 1.5 80 1964 30 2 1.0 60 1965 48 1 0.5 ?6 8 -- . ... . TABLE ill INCOHERENT SCATTER OBSERVATIONS – 1965 Mean K End (EST) Begin (EST) — 12 January D 16 February 14 January 1000 18 February 2+ Somewhat disturbed q 0700 l– Quiet 1300 8 April Q 1100 1+ Quiet 13 April 0900 15 April ‘3 0900 1 Quiet 18 May 0800 20 May 0200 l– Quiet 1100 16 June 2100 5+ 20 July 0800 22 July 08Q0 10 Quiet 17 August 08$J0 19 A“gwt D 0800 3- Somewhat disturbed 1100 16 September D 1100 3 Somewhat disturbed 0800 28 October D 0800 2- 0800 I D 14 September Q 26 October D 16 November Q 0900 18 November 21 December >., Q 1300 23 December DATA A. Critic81 N scatter results max = of the critical frequency made at Fort Belvoir, i.24x As a first step, curves different behavior electron density i04(foF2)2 observed values only the shape are established by assign- (4) cm an ionosonde. good records Previous Hill C-4 ionosonde 77” W) have indicated we chose to employ were the Fort not always Belvoir obtained values August], since no obvious the at Millstone; of foF2. differences very weli, On a number of days but we did not z-egard the could be found in the shape of the In the case of June [Fig. 6(f)], on tbe i5th and i6th that we made no attempt 9 between of local time for all the hours of observation, foF2 did not repeat or the temperatures. examinations with measurements good correspondence drawn through the points as shown in Figs, 6(a-1), as serious, profiles density to derive electt’ons/cm3 we plotted foF2 as a function were are employed The absolute using the Millstone [38° N, malfunctions, Nmax, at Millstone frequency measured 20-21 .Tuly and f7-48 so different Quiet 0 —— of critical Virginia Owing to equipment in computing obtained height profile. foF2 is the F-layer (e. g., Quiet 0 Frequency ing a value to the peak density and smooth 0 REDUCTION F-Region hence, Very disturbed 1 — of the elect ron-density-vs- where 0 I200 q — The incoherent two. 0700 6 April 15 J“.. IV. 1100 P to construct the behavior an average 24-hour was curve. .IAN )2 )’365 . ,3 0 ,,, h ,, .,0 -14~ o : $ .. 3 ... . . 4.0 — ..0 . d“ ;:...”. so z !,65 . 00 . . ‘a m. , cc .. 1 I I \ I 1 I I I A’ .0 ,.ZO .48,, EST EST (b) (a) 8.0 TEEIII x I 0 I , I ..,2 , t.,. ( EST (c) Fig. 6(a-1). F-region .ritical freq.encyvs local time forperiods ofobservafionin 1965, . . . ..( %noao .I ,“. 60 $ ~. .. 4.0 I I I ( 1 I I , I \ 1 I . ,6 !, 0 A . . . ““ !965 . 2, 0 . . .. . o@ I , Iz 16202’ EST (f) (e) 20 . . . . “ 0%30 EST JUL ~. co” co I ( ! . . . . . 00 . . . . . . 04e 162024 O+SIZ 15 1 .. ,,65 . . . . 6.0 -22A . ~. . A : .. . . 4.0 2 I I AA z.. 20 t [ ( 0.,,. ,620*’ o.a,, t , ,6Z0’ EST EST (h) (9) Fig. 6(cJ-1). Continued. 14 , I I I , [“” ,,, . .,0 Ooco TImiEL SW 1965 ,4 . ,s o A A cm ““ . . ... “’ OOwm co ~.. m 0 0 $ .. . Go -. 4.0 % o A 2. . 27 -i2mcL = ~ o Cmm 26 m Mu M “... . . . . . 00 0 ~ 0 I ( .8,, ( , , I t ! I 3620” ,s1 EST (i) (i) ,.0 t I ..a~, ,6’”’ EST (1) Fig. 6(o-1). Continued. Electron B. DenSiky PrOfile Echo power yield vs delay observations an echo power profile I 0,5-, using O.i -, P~ vs height and 1. O-msec pulses were combined to h from T,& I Ps(h) (5) = h’ where T ~ is the equivalent echo temperature Next, these were power profiles to-ion-temperature observed corrected rat io T~/T i obtained at the completion for the effect from on the scattered the analysis of the signal of the sweep integration.% power of the electronspectra2 i + (T~/Ti)h N(h) = ps These obtained corrected Although To do this, profiles results aII the power density the value of the peak electron electron profiles June, the accuracy (Secn,IV-D). Electron probably density Nmax decreased i 20-km altitude, i.e., The ratio T~/Ti to July, These curves (between above, temperatures temperature were charts were of Nmax. in the values of to yield frequency the width for fixed on the assumption everywhere is f OMHz. and the values that This obtained T:. and plotted in the CIRA profiles from of these quantities prepared’ used and smooth next hand- scaled being underestimated, assumed corrected on charts, then determined by the symbol these plots was employed These values the same procedures were frequency into hourly groups to true values The peak the power at the center showing the variation and are represented from introduced machine-plotted smooth and ratio These the fixed temperature scaled and its shape was (Sec. C below). we employed were and that the plasma divided with re- This mean pro- For the data taken since errors next drawn through the points and extrapolated as outlined electron (Ref. i ). of Ti and T~ were were altitudes shape. averages. to July) and ion temperatures leads to the electron as fictitious The values sity profiles (Prior using charts ion present of T~/Ti the scatter only the hourly at given previously,z owing to systematic made prior The electron by interpolation of Ti and the ratio T~/Ti Smooth curves scaled 30 minutes, was taken to be the mean of the individual (to half peak power) 0+ is the predominant are described every to compute with the average mean values the 24 points in each spectrum and that in the wing). last assumption obliged has been discussed the measurements of w~ values was obtained in each hour were drawn through these by eye. the parameters and ratio profiles and Ion Temperatures in,1 964.2 That is, were obtained of this procedure T:/Ti In analyzing profile that we were to Eq. (6) using the hourly for the se mean hourly C. density was so large in height to place the peak at the mean of the peak heights, according The accuracy titious then assigned and a mean taken to obtain a profile was adjusted corrected values were an independent spect to the peak, curves (6) m ax 1 via Eq. (4). of the temperature file [ i + (T~/Ti)h as a function of altitude. down to a value of 355-K at i 965 model atmospheres. in Eq. (6) to obtain the corrected were then employed to convert den- the fic- viaz (7) lm3m NO YE* T“(RO FIT ,E~ TRIAL? PARABOL. P = bo+ b,f eb,+’ COMPUTE YES cOMPUTE ~ ccTRY 8 } G-J:,HIH’D !S PEAK WITHIN I“Hz m, c NO HAVE 2 POINTS YE, 0’CENTERYES“EN‘EM0vED7 OFMbTRIX? c~G:TE OUT OF PEN w R NO > ANY 4X MEAN t’ Fig. 7. L.gi c fl This section was w diagram employed of part of computer to fit parabola to wing program empl eyed to of spectrum in way process signal spectra that would eliminate to obtain temperature spurious points. measurements. This expression higher values is believed probably only at altitudes D. Spectra first T; These (10 or more Ti and T:/Ti. to at least graphical method: spurious points. ,,, expressions i.e., deviations, points. to the remaining substantially for drifts and channel-to- In the $ourse in time taken toubtain of developing the nonlinear and the echo Wwer the spectra hence, “oltage always as the square tion would be correct, of the signal-to-noise be zchieved nately, log, be greater and a proper ratio would be to adopt some this could not readily coded on the punched-paper tape, be done. The effect T~/Ti of interpreting to be overestimated, the two effects these assumptions in em-or. and, again, result~ pulses, and after signals in error. will be chiefly before relation he interassump- some knowledge altitudes. Unfortu- tape but in a separate 1.0 msec) was en- the O.5-msec measure- or pulses. Eq. (2) will be to cause will be least interpreted Therefore, in T& since “sing Eq. (3), Ti the effects to cause parts of the Ti curve July, Con- the best that could with 1, O-msec The error [Eq. (3)]. neither would require Eq. (2) to interpret altitudes analyzer tbm noise, voltages altitudes, law at these middle Fur strong we rec- the signal-to-noise on the punched-paper using the incorrect results obtained of the echo power As tbe length of the pulse (O. 5 one another. provided we knew that weaker that the measured T& should not be greatly on the temperature By comparing power made at greater weak signals and an entirely analyzer, invariably of the spectrum and Ti to be underestimated. tend to compensate will be overestimated were At intermediate we chose to employ ments and Eq, (3) for the measurements as cmtl<iied. in this program Since this was not available, intermediate sdopted tbe temperatures, with the earlier requiring was not recorded was then by the new spectrum using 0, 5-msec interpretation a paraboM ~~~ than a cer- The computer.. program in complexity, measured pulses) [Eq. (2)]. at each frequency. the height of the spectrum hence, than unity, root of tbe echo power by more the edges of the .Spectrum, the delay, to determine voltage (225 km), ways. .@f rejecting values. upon experience examined using the earlier gain differences grew of in this way was The parabola ,.,alue. could be taken as the best estimate altitude values Zm.m the final procedure. Despite program (using 1-msec deviated rejected, chanw:l 968. the signal Based at high altitudes at the lowest would nearly preted the computer between (Sec. H-C), the measured versely, ratio relation obtained to temperature To locate the program temperature polynomial or in the wing of the spectrum, the half-peak until mid-i expressions fo? various was in devising 7 shows i“ schematic was net obtained ognized reduction the computer high-order than could he achieved and all. such points were Figure of all these .requirements, savings important The philosophy Thus, these into analytic obkained by fitting better with machine See; :li:: As a result ersion by the computer. and width parameters satisfa~toryv immense were to three ..poirhs nearest to correct but at gathered. hand analysis, the heig!it of one of the wings in the spscirum; line was fitted was also required this is probably were test was then made to see if zmy points tain number of standard x t O-i “K cm3, of ratio and width computed Thus, to find the height in the center refitted a straight analyzed and inserted to the values encountered to the points.:A to determine ,. In practice, completely and ~, of the ratio i percent, The rna,jor. d.iffimdty was fitted analytic The conversion accurate were of l~h terms] (T&/N) <2 few useful measurements t~’e same as in the earlier the parameters and Ti. expressions provided (July and Later) in July and later for this was basically estimated to yield and Ion Temperatures obtained accurate Te to be overestimated. above about 700 km where Electron employed to be reasonably causes it appears that the error is of to he largest in the region 300 to 500 km where sate for this when constructing used to correct the electron as the data taken prior The accuracy The largest the spectrum numbers less V. of measurements. were RESULTS A. Diurnal over mean profiles were was the random associated range the systematic 300 to 500 km; hence, reported herein. used to correct we place Unfortunately, the electron with by averaging taken in July and later, in the altitude curves error has been these density profiles. DENSITY Plots electron the height interval density profiles obtained intbe of constant electron manner density i50 to 650km and O to 24 hours EST. month of June [Fig. 8(f)] where, to plot the 48-hour owing to the markedly period loss of much of the last Iz-hours’ constructed to compen- in the same manner and this could be reduced than on any others in Figs. 8(a-1) as contours necessary These corrected to July to obtain temperatures For the measurements the random errors ELECTRON The mean hourly presented prior in the measurements pm sent also in the T&/Ti FOR curves. themselves width and shape parameters, on these measurements uncertainties temperature and were employed uncertainty (in Ti) exceeded reliance profiles We attempted to July. discussed.2 large the mean hourly density of the procedures determining errors Ti has been underestimated, temperature in full. profiles averaged over behavior equipment data during this period, in See, IV are plasma frequency) The only except ion to this is the different Unfortunately, outlined (actually, from day-to-day, malfunctions it was caused the To obtain this plot [Fig. 8(f)], intervals of 2 hours in order to reduce we the uncer- tainty in the measurements. i B. Me8n Daytime For some purposes, use than tbe complete density curves Figs. 9(a-k) and Nighttime the electron time There . . density provided 1000to These in which the latter is a noticeable increase discussed earlier smaller C. and less some period were intbe themselves curvature The effect constructed computed, average are presented from i.e., in the hourly aver- by averaging the of hmax and Nmax. for the region We believe through Eq. (6), were are of greater we computed above bmax intbe this tO be caused by the errOrs cause the density is not present at night, daytime tithe ratiO to be progressively overesti- since then T~/Ti tends to be Discussion A single of the winter daytime tween i200 and 1300 hours. at sunset which is probably summer. over Accordingly, altitude-dependent The main features of 1964. which, altitude. averaged averages these the mea.n values for July onward [Figs .9( f-k)]. mated with increasing profiles by Figs. 8(a-1). i500 and 2400 to 0300 EST and these respectively. shapes and assigning profiles T:/T, history for the periods andi O(a-k), ages in the same manner profile N(h) Profiles maximum Above enon to the rapid decrease rise.5 however, counterpart observed in Te occurring This suggestion . in <965 differ density there of a second .— from those usually be- peak occurring evening increase observed in 40 we attribute this phenom- eclipse, at sunset and the redistribution has produced little is found to occur is evidence of the marked during a solar 46 —- observed in the peak electron hmax, the winter Based upon the behavior which this gives diurnal behavior a spirited discussion of the ionisation to in the literature that ,,, ,LA,MA ,,66 FREQuENCY (MHz) 0, ,00 - 600 — % , : ,Ca . : . ?00 ,0 0 048!2 ,6~M ,*T (a) *LAW. FREQuENCY (MHZ) ., ,00 - z. : .00 : ,Co ?co I I 046!, ,.20*” EST (.) (b) Fig. 8(o-1). Contour ofconstant formeos.rements made in 1965. plosmafrquency observed as functions of height and local time 17 , .,..,m,,%, I I.48,. \62024 EST E,, (.) (d) ... ,, !5-,7 WM ,969 PLASMA FREQuX.CYIMHZI K EST (f) Fig. 8(cI-1). Continued 48 ml ,.0 w Jk 600 — ,.0 *.* ,.5 ,$ ,,. s.. ,,0 !5 \ 3.5 g : ,00 ii . 4.0 4.5 ,.0 300 t..” . 5.5 .--=. 3.0 ~ J-- I, \ ,Cc / _/’ \ @ WAS.. I 0.81z < / I IS J“L !,.5 F. EQUENCY (MHZ) I 1 202 ,s1 ,s7 (h) (9) I .484. f.,.,, ,,1 (i) (i) Fig. 8(a-1). Continued. <9 (k) I I oaf, ,,2024 EST ‘oom ,.0 ,0, ([) ,0. E,T Fig. 8(wl). Continued. 20 .00 TEEi3 14EAN MYT4ME PROFILE (4000-$500 UT) JANUARY ,965 600 I 5 > 400 % r \ Fired,. 4.6 x 405 ELECTRW/Cm3 ,0, (0) lmm I MEAN DAYTIME ,Immt (1000-1,00 EST) 8-0 APR)L 1965 \ t “,1 2 s 7,.2.50 RELATIVE 1 DENSITY mr [PMCM (c) (b) Fig, 9(a-k). Mean daytime de:si~ profiles; these plots obtained by averaging all measurements between 1000 and 1500 EST. Nmox is mean peek density during this time period. Zi “t — \ 200 I Fimo, 3,35 x d cLEcmo Ns/m3 .,.. ‘( o / 1 \ I 1 1 1 1111 ,71.=5. {, I 1 I I II! ref., [wc,”ll RCLAT,VE DENsITY (9) loco ~ OAYT!ME PROFILE [(000-1500 EST) ,E,TwBER 1965 6m.z : ,3 h.,” . 3,2 x ++,.,,.,0.s/.”’3 / 1 , 1 1 .,!.., ! 1 1 I II ,,,. (”, 1 205. 1 1 I . @ Ewmowc.’ ,00 /“ t 1 7. L “1 1 z 1 1 ,,,.,, C’GISIT” (Percent] 1 II 1 I 7,02.50 ”, ,,.,1,, (i) (h) Fig. 9(a-k). Continued. 23 1 ,,.,!,.!, 1 1 1 11 701c \ (i) L i.,, 5.2 x ,05 m,cmo.v+ / I ,..-. ,2 (k) L t Fig. 9(a-k). Continued. 24 \ 800 ~ .s., NIGHTTIME PROFILE (2$00-0500 EST) ,* N”An” ,,65 \ (a) MEAM NIGHTTIME PROFILE ,2!00 -0300 EST) .-e APRIL !s65 \ “f , I I , I ,7,.2.5. I?EMTIVE II I ! , L I L \ I m,, ,,..7 DENSITY lw~~”tl = 1.! x DE I+S!TY {9,,.”11 I I I Ill ,,m-~ ,”, @CL ECTRONS/Cm3 \ (.) (b) Mean nighttime@wity Fig. 10(a-k), betwee. 2100 a.d0300 EST. N i.,” profiles; Inox these plots obtained by averaging is mean peak density during this time period. 25 all measurements I I Ill TO,, . .,,, ii.., . \ 200 NIGHTTIME PROFILE (?lO. -0300 ,s, ! ,,-!5 iPRIL 1965 1 ,,2 x ,0,EL KTRrJNs/cm3 ; t 1 1 o I ,2 .,,,7 1 I \ II , !0 )., DEmsll” I m 1 I 1 (1 m 70 +., IWm.tl (d) !00 - ME,, N,OH, T>ME PROFILE (2100 -0s00 ml) , ,“LY (,s5 .0. ~ . .0 0 ~ ‘1 “! ‘$ - fima, . ,.6 x ,Os ELW,0N5/Cm3 \ Fmo, / t 1 a 1 I I ,7! 1 I II .,,.,,,, 1 020 1 1 1 1 m I L !. O, NSIT” (Owe”!) (e) 26 ,.36 x,05C,ECTRONS/mB 0-” .00 : ~ ... : Y .,., l\ N,G”,TIME PROFILE ,,,00 -0300 EST) MJGU,T ,965 Nm.x= ,,7xm~mcTRoN5/m3 I / 1 1 o {z t ! I 1 Ill ,74. ms. RELUT, VE DENS(TY [PWWI1 1 I 1 I I Ill m4, (9) \ I I 1 1 ,7 )11 ! I \ !...,, RELATIVE Oc.sl.Y 1 I 1 ~. W,,”,, (i) (h) Fig. 10(.a-k). Continued. 27 .mr .,,. H,G”TT,ME PROFILE 121.0-0300 EST) ,OVEMBER 1,65 (i) \ C,”ot . ,,6 xm. CL EC,.OWG., /’” “, I * I I I 57,. R, L.TIVE 1 Ill C.c.ml” 1 20s0 1 I 1 I IL m!. .,., -1K {,erGe.t) .“. 1 Lt!&!EJ NIWTTIME PROFILE (2!,0 -0300 Em> C’ECEMOER 1965 ~ ;’”” (k) ire.,. ‘,5.!0$ ELECTROWCF /-” m. } Fig, Io(a-k). Continued. 28 is too extensive attempt to review here, However, to model the phenomenon ity equation we may using a time-dependent has been performed by Thomas electron and ion temperatures based produce the evening but a later increase, the Millstone model points as follows. An solution to the electron density continu12 employing a simple model for the and Venables upon the salient summarize observations. model This in which the electron failed temperature to re- was allowed to govern the loss rate according to the laboratory results of Schmeltekopf, et al., ‘13 did produce -— i4 maximum. i5 Kohl and King, and Kohl, ~ al.,~6 have argued that neutral winds, which serve to drive an evening ionization sities, down the field lines during the daytime are set up in the upper atmosphere. the wind direction, recombination which causes rate. The evening variation depress maximum Strobeli7 treatment of the problem seems 18 $t_Q., who solved the time-dependent and Sterling, of winds, the original and the variation explanation of ionization. height of the layer s ity.6 peak to be raised This feature 8(1 ), seems during winter from the c~njugate scatter of the plasma 19, by the F-region. ionization of Millstone, for the F-region thought to reach density understood, During summer, winter to summer behavior Figs. 8(c) and 8(d). March, and summer Reasons day electron (i) density lines mnlit drifts). throughout in den- excite is tbe source complicated at sunset. quite rapidly of signals c?n act as a sOurce Of the downward sunset, of the ionization temporal flux is and thereafter decrease seen in these morning hours is not and spatial variation of peak usually reaches Tbe transition from a maximum before the characteristic around equinox as may be seen by comparing behavior is encountered from mid-October through through September. maximum have been discussed composition blow away from of higher by means of incoherent in the early is thought to be caused by either hemisphere Hill in winter. at the layer mid-April that (Direct at Millstone in the spectra At equinox, about 1 hour after of ionization the night. arriving that the $fltonOsphere for their occurrence winter diffusion This was accomplished much of the night. density from into regions that for the when it was recognized of photoelectrons night. change in the neutral winds which in the summer the a redistribution it is necessary of downward lines that the photoelectrons evening in summer in terms remains detection maximum occurs behavior for the summer a summer-to-winter down the field to Millstone from about by the more At Millstone, to indicate is the predawn increase sunset,6 was discarded it is believed the electron including appear for it in these data is found in Figs. 8 (k) and tube that must occur noon, and a second much larger out by equation electrodynamics diurnal behavior If the protono sphere of the field this idea. in f965 than in + 964, and is thought to disappear the reason but may be brought the temperature as that observed, in of lower pronounced throughout increases, supports continuity den- a reversal to have been carried Both studies (or possibly a peak of 1040 electrons/cm2 or less monotonically. nocturnal electron Best evidence point during the winter For the latitude more conjugate observations scattered by winds conjugate of this has been obtained from in that the rapid fall in Te initiates an increase for this phenomenon, following the region temperature. of the winter far less offered the pi-otonosphere evidence feature at sunspot maximum. An explanation from com-ect to obtain as large A second anomalous completely of electron is partially However, results of hmax seen in Figs. 8(a-1) The most comprehensive effects the midday to be driven up the field lines to regions the ionization The general and thereby above. The depressed or both the following of the atmosphere,z f‘23 mid- effects: or (2) nentral the equator at noon and drive ionization 15 rates. Experimental evidence favoring recombination 29 . .. . .— the first I of these carried explanations has been obtained in the L-band incoherent scatter observations 41,24 These appear to indicate that, relative to atomic oxygen (0+), out at Millstone. mole cular ions (02+ and NO+) are present the 10Ss of ionization largely proceeds to greater altitudes via the charge in summer exchange than in winter. Since reaction 0+ + N2 -NO++O NO++ this suggests is lower N+ e- O+hu (8) that the abundance of the ionizable in summer 1).ncan” species recently reviewed all tbe explanations concluded that a variation in the neutral composition suggested that the change in composition is brought mer to winter hernispbere, If this is the case, and that a rapid changeover sphere causes Changes lence, in the relative and becomes neutral better collision frequency forms Continuation are probably transport as the sun moves and of O from 13e the sum- (i) and (2) are related, into the northern could also be bmugbt ceases to be mixed its own hydrostatic for such changes from at ‘t15 km, anomally, hemi- incoherent of these studies, needed to resolve equilibrium scatter studies together the relati”e about by due to turbudistribution. of the ion- with more importance and of the effects. scatter measurements seasonal the seasonal anonmlly. ELECTRON A. AND ION t? the density and local results is found, it seems inadequate to account for TEMPERATURES we attempted of Figs. 8(a-1), Te and Ti Profiles to present i,e., our temperature as isothermal results contours in a manner drawn a. functions analogous of height This work has not been altogether very rewarding since, apart from somewhat 4,6 we tend to find that the temperatures remain fairly constant bebavior in winter, complicated the daytime two regimes urements. and again during the is sufficiently in presenting to a single the results In addition, true at dawn, altitude in order where and nighttime tbe transition it is properly it has been necessary to obtain adequate time for i 965, we have plotted only tbe average (4000 to 4500 EST) and i2(a-k), nighttime. rapid that it is doubtful whether This is especially ments restricted (Zi 00 to 0300 EST). These explored to carry resolution.8 temperature are given between tbe in these measout measureAccordingly, PrOfiles fOr the in Figs. Ii(a-k) respectively. When drawing (indicated peratures velocity winds has been provided by 1%-ench inco27 IIowever, alnear the F-region peak. time. throughout daytime drift neutral of the wind velocity and Nighttime years, involving of vertical variation 28 Mean Daytime In previous ments about by global then it may be that effects abundance of O to NZ in the Fi-region measurements, seasonal tbe two types of behavior. one in which each species though some VI. (02 and N2) is the only one that will fit all the facts. wind pattern between Support for the second explanation herent to molecular for tbe F-region (near 110 km) at which the atmosphere has found evidence rocket various in the global the abrupt transition a change in the altitude WeJdteufe12’ (0) relative than in winter. these curves, by the vertical are essentially we have been guided by the height resolution bars) as well as the belief that electron, the same below about <30 km (Refs. 30 of the measure- ion, and neutral 14, 29, and 30). Therefore, temboth (a) Me.. daytime electronand ion-temperature profiles. These plots obtained by averaging Fig. 1 I(a-k), all measurements belween 1000 and 1500 EST. Bars denote height of exploring pulse. From July onwards, ion temperature between 300 and 500 km has been underestimated (see Sec. IV-D). 3i , ) 0 1 ... 1 1 ! sw 1 f200 1 1 1 ,6.0 TEMPERA,”,, 1 ‘am 1 1 1 ‘w “m [w) (c) !3-$5 Am,, t ma !OOO-! 500 1965 EST I 1 I { o tllllllll .0. 800 120. le.. TEMPERATURE 1°Kl (d) Fig, 1I (a-k). 32 Continued. z.~ 24m 2“00 1 0 I 1 1 4CU 1 1 8.. 1 ! 1 1 I t2.. l-o TEMPFRAT”RC {W I =.. 1 1 ,4W 1 I ( ,4~ I 1 p,,. 1 (.) ,., !.., JULY $965 moo -1500 E3T I ~ . so. – : . ‘ ..0 - xc – , I . .Ca , em 1 , , , , !200 IWa TEMP6RAT”RE PK1 I ~ (f) Fig. I l(a-k). Continued. 33 ., z... 3- AUGUST ,000 -t50C 1965 EST ,00 — : 6“0 ‘i $ Y i 40. — 200 — I \ ..0 0 I 1 8.0 ) 1 120. i I I I .. . . 1 ~ t 1 1 ‘*” ‘am I ,4~ z... TEMPERATURE {-K) (9) ( ,F,,EM8ER ,,65 moo -,500 ,s7 am — I : 60” : ; .m — 200 — 1 “ , .0. 1 I *CO 1 1 1 1*OO I !600 1 TEMPERATURE VK] (h) Fig. 11(a-k). Continued. 34 1 moo 1 em T/ t J J%’” . “t t 0 I 400 1 800 I 1 1600 TEMPERATURE 1 \ 1 1 I \ I I I f,., mm ‘“w ‘m” (T1 (i) ‘“””r ‘ -iIEmzl 4 ‘00”-”00 ‘“ - , 1 I , m \ 1200 ,sOO K1.!FCRhT”,, ?0 (i) Fig. Continued. 11 (a-k). 35 I ‘“m 1 I 400 1 ,800 !.00 mzml %=3’% t 1 we 1 o , I em 1 I !.W I TEMPERATURE l~K1 I , moo I I I 2.W ,400 2,M (k) Fig. 1 I (.-k). Continued. !00. T@!mL JANUARY 2!00-0300 1965 EST m. — / ~ ‘OO k g I : 400 — .0. — 1 400 1 “ I 1 SW 1 1 !200 1 ! 1 I ,600 1 ,~ 1 z~oo 1 1 1800 TEMPERATURE W -.1 (a) Fig. 12(a-k). cdl meawreme”ts Mean nighttime behveen 2100 electronand 0300 and E5T. ion-temperature Bars denote 36 profiles. height of These plots obt~ined exploring pulse. by averaging - F,,WA,Y 2400-0300 ,,.5 EST . ‘“””E ~ , /“ / / 1 a I I 1 4,, 1 I s00 I 1 !?00 I 1 !600 TEMPER AT” RE (°K1 (c) Fig. 12(a-k). Continued 37 I 1 *W { ,4w 1 I ‘mo — !3-!5 APRIL 2100-0300 !965 E*T ,00 — . : 6C0 — : * ‘ +00 — 200 — ) mm I !600 ,,0. I I 2400 ( 1 mm TEMPERATURE lPK1 (d) . f .00 1 I ma 1 I !m. I ,.. 1 !..0 Pm ATui7E (P.) (e) Fig. 12(a-k). Continued 38 I I 2... I I 24.. I 1 .~o ) . ! I 800 1 ,.. ) I zoo ! i I 1600 1 2000 1 1 1 I ’400 I 2’00 TEMPERATURE lPK1 (f) ,06. ~ r---”--”””””’” AUGUST 2,00-.300 800 ~~~ 0 I 4.. 1965 EST 1 1 80. I j... ! !20C TEMPERATURE 1.K1 (9) Fig, Iz(a-k). Continued. 39 1 ,@o 1 1 2“00 1 “0” ,.00 Tmm ,,,,,.,., *,,5 2100-0300 800 EST [ I ,,00 1 I !600 T.M,CRAT”R 1 ! 2400 I 2000 1 1 2*OO E [r. ) (h) OCTOBER 800 o I $965 2,00-0300 EST [ 1 1 400 1 1 1 eca 1 I !200 1 Mm I TEMPERATURE PK] (i) Fig. 12(a-k). Continued. 40 1 mm I 1 24~ I 1 ,~. I ,“”” = 1 40. . 1 ,.0 I 1 1 1600 I I2.. TEMPERATURE I .000 I 1 \ z... I 2*OO I (w} (i) DCCEM8ER ,965 2100-0300 EST [ .0. t 0 1 1 .OO 1 1 S@ 1 ,200 1 1 ,..0 TEMPERATURE ?K] (k) Fig. 12(a-k). Continued. 1 I ZOO. I 1 MO. 1 1 2,.0 ~ Te and Ti have been extrapolated downward at 420 km in the CIRA atmospheres. , For the daytime 1965 model results obtained after to the value of 355-K employed June [Figs. ii (f -k)], the curve for Ti has been drawn through the point at 225 km, and then taken higher than the plotted values underestimate “cry in Ti (Sec. IV-D), However, it seems as the temperature to compensate doubtful that this portion for the of the curve is accurate as the exospheric temperature during these months is thought to have been of the Ii of 800a K or higher. As noted earlier, the electron temperature should be substantially order independent of the error Discussion B. The temperature the photoelectrons energy escaping behavior created to escape photoelectrons ters in Ty the local ionosphere the conjugate electrons. ionosphere. in the electron ducted into the ionosphere that the escaping of the escape energy of these?5’37 included. theory some of its energy on the order the effects via Coulomb temperature gave rise behavior escape sophistication However, This encoun- directly gradients to seen to a heat flux con- (Ref?,, 7 and 31), and assumed Recent thermal ionosphere. to be given theoretical estimates 32-35 support to this conclusion. of photoelectron and the measurements. daytime of 5 X i 09 eV/cm2/sec of the F-region the enormous is presumed that this effect has been made with the Millstone Despite photo ionization postulated m upward flux of to the conjugate in the protonospbere of the energy Some of have sufficient during the daytime, to account for the large we earlier flux have given general comparison Thus, previou.sly.4 in the atmosphere the protonosphere flux was even larger. A number of studies direct completely. The balance In order temperature, has been discussed of ion production that traverses flux deposits with the ambient at Millstone in the process is established photoelectron observed of the magnitude have been carried out in which a 35-37 In the mgst recent measurements. and fine-structure of these models, these may simply cooling of oxygen were remain between differences result from uncertainties both the in the cross sections, collision cross sections, and other parameters that were em35 ployed in the calculations. For example, it was found that the results of the two studies differ 35 more widely between themselves than with the measurements. One phenomenon existing models ticles. These that has recently is the importance encounters they cause a portion b’dckscatteredi’’ local serve come to light and is not adequately of elastic to limit the escape of the flux of photoelectrons The albedo of the atmosphere many of the photoelectrons to make several their into the F- region the ene?gy or penetrating of the escaping and suggests line?’ served at high altitudes The photoelectron served is still is deposited flux, from the conjugate photoelectrons will deposit is possibly traversals accurately hemisphere their to be energy in the as high as 0.5 (Ref. 41] causing of the protonoiphere before losing This must cause a larger at one end. to appear in any of the and neutral par34 In addition, as pointed out by Nisbet. in the form flux can he determined than was previously from of heat conducted the temperature all fraction of along the field gradient ob- thought, escape sunlit and supplies directly the protonosphere more included photoelectrons flux is believed to be responsible for the high temperatures ob4,6,7 during the winter nights because at this time the region conjugate to at Millstone Millstone photoelectrons that the escape between arriving 38-40 so that not all the arriving ionosphere. energy encounters in the local is on the order heat to the protono sphere.8 ionosphere by arriving of half the daytime 42 It is not certain photoelectrons, value.7 Observations how much heat but heat conducted of the plasma from lines in the spectrum conclusively lines of the that are observed deposit During temperature the F-region hydrostatic pressure MAGNETIC during the winter nights shou, numbers. $9 However, pksma energy the photoelectrons off” (convection) below. with electron the daytime to a to warm the F-region temperature as it as Te 512 , the lowering the proton osphere of to continue altitudes of Millstone, hut quite important DISTURBANCE between 20,42 Ackiitimv+l heat is brought into the P-region 20 under the influence of the reduced sunset. to lower to persist of the protono sphere. is heated during continues the heat 10Ss, permitting For the latitude conduction, appears by the cooling the protonosphere for many hours after than thermal A. difference and at nighttime varies “chokes of electrons I’-region that whatever temperature and Rowhill,42 Since the heat conductivity via transport Vfl. a small above that of the F-region, the F- region effective the in significant high altitudes. nights, out hy Geisler temperature to warm from Millstone hmax F2, indicating at rather the summe~ signals do reach and ions and this is thought to be maintained As pointed cools. scatter only above is introduced electrons Thomson photoelectrons this process 20 is thought to be less nevertheless. EFFECTS August and September The average magnetic I<p figures In August and September, were followed a direct by disturbed comparison recomputed the first quiet and disturbed periods. 111are for the entire observations ‘rhe results were months therefore quiet and disturbed and nighttime days, in Table These of consecutive the daytime given 24-haursl temperature afford periods. averages are presented period relatively of observation quiet periods the opportunity In order to do this, fOr these mOnths, in Figs. i 3(a-b) tmd for making we have separating tl?e and ‘14(a-b) for the two months. The August storm ., sudden commencement this, comparison appears to have been one of the principal of Fig, ‘13(a) shows little difference f 7 and i8 August, The slightly lower may have been caused by the slightly ~t $imuld seem that the observations commencement averaged for significant over the periods in Fig. 53(b). increase There in electron The September electron higher electron conducted appears to be a distinct temperature storm ohsemed densities to have been too short for between increase as encountered encountered distribution, although the ion temperature [Fig. ‘14(a)]. A comparison changes dots of the nighttime observed on day [Fig. ‘13(a)] on that day [Fig. 6[h)]. 9 August is compared but no dramatic in September 1964 (Rrf and lasted about 4 days. of the storm and the daytime to have been produced appear behavior and i8-i in the ion temperature at nighttime the commencement significant on the storm with a Despite too soon after the sudden 44 to have taken place. The behavior began at 0954 f?ST on ‘f 5 September the time interval 5 days. on +8 August followed heating of the atmosphere seems B. of i965 (Ref. 43), between the daytime temperatures temperature 2100 to 0300 EST for the nights of ‘t7-i8 again, also shows little storms (SC) at 0840 EST on i8 August and lasting for the following to he higher 10). 43 Ilcrc obser”atmns in the temperature on the disturbed for quiet and disturbed periods day [Fig. 14(1>)] difference. June As already noted, A sudden commencement the observations occurred made in June were xt -0600 the most disturbed EST on 15 June, hut the magnetic 43 for the entire indices year’ Ap and K p I J---J. t ,:00 t ,:00$ TEMPERATURE (a) o I I ,00 I 1800 TEMPERATURE ~) ~g. 13(a-b). Kp is mean Comparison value of Kp of index mean over temperature these time ~ ,;oor I Daytime. I ,200 , ,;OO ~ [°K1 J I 3000 I ,400 I leK) Nighttime. profiles intervals. 44 for quiet and disturbed periods in August. I sEPTEMBER ,000-,500 $965 EST M SEPTEMBER ?, = 2- ,5 ij :} SEPTEMBER = 5- :} 2.0 [ I 1 6.0 0 1 I 1zoo ,,00 TEMPER&TURE [°K1 I ,400 I ,000 (Q) Daytime. I ,00 0 (b) Fig. 14(a-b). ~p is me.. Comparison of mean temperature I ,*OO I )m. TEMPERATURE 1°K1 I ,.., ,000 Nighttime. profiles value of Kp index over these time intervals. 45 for quiet and disturbed periods in September. reached erably their highest depressed sunset. suggesting Millstone. and July. At nighttime, on ib June. Figures Here, the electron 15(a-b) the storm The electron density with that on the 15th [Figs, density that the trough of low ionization had been increased increased. values compared compare effects to anomalously observed the average are clearly and the electron [The June values fell for the period around for the time of year. 45 Alouette had moved over and nighttime temperatures During the daytime, reduced. of Ti are not plotted on the 16th was consid- except low values by the satellite daytime evident. temperature observed 6(f) and 8(f)] At nighttime, in Fig. i 5(b) because for .1une the ion temp~rature Te and Ti were both they fall on the July Te curve. ] )0.0 JUNE/JULY 4000!500 1965 EST .) I 0 600 I I ,,00 !,.0 TEMPERATuRE [-K I r ?.00 ,00. I 2.00 I I 300, ) (a) Daytime. ,000 wNE(JULY 1965 2100-0300 EST 0 I em I ,2,, I ,,00 TEMPERATURE [~K1 (b) Nighttime. fig. 15(a-b). ComFmrison of mean temperature profi Ies for June Kp is mean value of Kp index over these time intervals. (disturbed days) and July (quiet cloys). 46 ---- –—-———i ~.-..-....--...-,*- C. Discussion In agreement of disturbed conditions temperature, (i. e., is to raise at altitudes below nce with conclusions to disappear 10 earlier, the exospheric This seems since at this very that the principal temperature, temperstur.es clear altitude It appears neutral where the ion and neutral about 400 km). seems reached from are substantially Fig. 15(a). the electr’on daytime and consequemly Above temperature effect the iml the same -600 km, this differe- begins to govern the ion temperature, A reduction did occur to have in June t 965. formed been a result, electron temperature chving to the expansion higher the electron than in July, ceeds in the daytime was not observed of the neutral in JmIe than in July [compare density N ia actually higher This would lead to increased at a rate proportional local to N2(Te – Ti)/T~/z, maximum observed in September atmosphere, hut i 964, the F-1ayer appears Figs. 8(f) and (g) or 9(d) and (e)]. at altitudes cooling in the range of tbe electrons As 300 to 500 km in June by the ions, which pro- and may account for the decrease in electron temperature. The very large evening that at sunset on both days the layer tral wind velocity The very proposed or, as seems likely, large, either as a result the action of an east-west of increased electric neu- ‘ field. low peak electron density observed on i 6 June prior to sunset supports the view 25 46 Norton, and others that at midlatitudes the storm time behavior z-esem - then often encountered. the seasonal anomally, transition These high electron because authors namely, The June 1965 nighttime August height was very more to the fact by Ouncan, bles the winter-to-summer very on $5 and fb June must be attributed attribute a decrease behavior temperatures abnormally the storm in the ratio closely were low F-region densities to the same basic that observed in September are cause as of O to Nz (and 02) at F-region resembles encountered. behavior electron heights ‘1964 in that On the other hand, this was not observed in [Fig. ?i3(b)] although the average Kp “alue was 50 in both cases. We speculate that in June 45 which appears to mark the boundary between the regular midla.titude iono 47 and the irregular polar region, moved southward over Millstone. Norton and Marovich the ionization sphere trough, have observed that the trough appears higti plasma August temperatures exist to be associated inside the trough, and June nights could be explained with the midlatitude If this is the case, by different locations red arc, and that the difference between of tbe trough with respect the to Millstone. VIff. SUMMARY The Millstone AND CONCLUSIONS Hill UHF Thomson 1965 to obtain F-region were usually made for 48-hour month, Magnetic (August and September) type behavior ing months, previous density the magnetic scatter densities periods radar evident prior was operated routinely and ion temperatures. on the Geophysical sudden commencements and immediately system and electron occurred World Days near the middle during two of the observing to the observations in June. only in the case of the June measurements. activity was low and the results throughout Measurements differ little from of each periods However, storm- During the remain- those observed the (sunspot minimum), is a midlatitude beha”ior, of electron storm is clearly year Millstone electron density station that exhibits with a rapid transition is not accompanied between by marked a characteristic winter the two near equinox. variations in either and summer This seasonal the electron electron variation or the ion 47 ..—_.._ . ., . .,,_.__,-_ ._ temperatu~c, The depressed a change in the neutral results is still composition lacking. tYPe of composition electron The change, density encountered of the atmosphere, storm behavior but in extreme observed although is thought to rcs!dt confirnmtim transport of the turbopau. sealtitude have both been advanced as mechanisms from of this from in June is thought to represent Global t’orm, in summer of atomic oxygen which produce Imcket the same and changes this composition change. In contrast behavior to the uncertainties of the electron understood. Although onstrated, employed we feel density, a per?ect both to higher mining to permit met recently observations source theory are capable of the F-region seems magnitude governing the reasonably well has not yet been demof various pzrametcrs error. of a number of improvements (b) improvements and (c) increased by the construction the mechanisms and measurement the absolute of remaining altitudes, (at low altitudes) concernit,g structure concerning at Millstone the ion temperature, ments were the thermal is the chief and lower exist match between that uncertainty in the theory The measurements tensions that still in accuracy time and height resolution. of a new spectrum using short-pulse analyzer, autocc>rrelation including – especially (a) exin deter- Most of these requireand by modifications methods measurements, ACKNOWLEDGMENTS J. Morin, and The invaluable assistance of W. A. Reid, J. H. McNl*lly, at the Millsmne Hill Radar Observatory in maintaining, modifying, and operating the radar equipment for these measurements is gratefully acknowledged. R. Julian and ]. Moriello contri!sdted extensively to the data reduction by writing the computer program employed to analyze the measured signal spectra. J. K. Upham carried out the computer processing of these data, while the data reduction requiring handanalysis was chiefly performed by Misses D. Tourigny, F. Angier, and L, Zak. The constant support of P. B. Sebring is also gratefully acknowledged. others 48 -,.._— of spectrum REFERENCES 1. J, V. Evans, “Ionospheric Backscatter Observations at Millstone Hill,” Technical Report 374, Lincoln Laboratory, M. 1.T. (22 Jm.ary 1965), DDC AD-6166~7. 2. —, “Millstone Hill Thomson Scatter Results fOI 1964,” Technical Report 430, Lincoln Laboratory, M. 1.T. (15 November 1967), DDC AD-668436. 3. —, Planet. Space Sci. ~, DDC AD-616607. 1031 (1965), ‘4, _, Planet. Space Sci. ~, 5. _, J. Geophys. Res. ~, 1175 (1965), DDC AD-61431O. 6. _, J. Geophys. Res. ~, 4331 (1965), DDC AD-623606. 7, _, 8. — 9. _, Planet. Space Sci. ~, 1387 (1967). 1557 (1967). , J. Geophys. Res. u, 3489 (1968), DDC AO-673605. J. Geophys. Res. 7Q, 131 (1965), DDC AD-613886. 10. , J. Geophys. Res. ~, 11. _ and L,P, Cc,.x, J. Geophys. 2726 (1965), DDC AD-621233. 7&, 159 (1970), DDC AD-703492. Res. 12. G. R. Thomas and F.H. Venahles, J. Atmos. Terrest. 13. A. L. Schmeltekopf, F. C. Fehsenfeld, Sci. Q, 401 (1967). 14. G, R. Thomas, J. Atmos. Terrest. 62~ (1967). G. 1. Gilrnan and E. E. Ferguson, Planet. Space Phys. a, 15. H. Kohl and J.W. King, J. Atmos. Terrest. 16. H. Kohl, J. W. King and D. Eccles, phys. ~, 1429 (1968). Phys. ~, 1045 (1967). J. Atmos. Terrest. Phys, ~, 1733 (1968). 17. D. F. Strobel, ,0. the Ionospheric F2-Layer and Its Maintenance at Night by Thermospheric Winds ,“ PhD. Thesis, Department of Applied Physics, Harvard University (1968). 18. D. L. Sterling, W. B. Hamon, R.]. Moffett and R. G. Baxter, Radio Sci. ~, 1005 (1969). 19. J,V. Evans and I.J. Gastmin, J. Geophys. Res. ~, 20. A. F. Nagy, P. Bauer and E. G. Fontheim, 807 (1970). J. Geopl,ys. Res. L?, 6259 (1968). 21. W, Rishbeth and C. S. G. K. Setty, J. Atmos. Terrest. Phys. a. 263 (1961). 22. J. W. Wright, J. GeoPbYs. Res. Q, 4379 (1963). 23. W. Becker, Electron Densify Profiles in tie Ionosphere and EXOSP&S. (North-H.lland Publishing Co., Amsterdam, 1966), p. 218. ‘ 24. J. V. Evans, ]. Geophys. Res. ~, 3343 (1967), DDc AD-658912. 25. R. A. Duncan, J. Atmos. Terrest. Phys. Q, 59 (1969). J. Frihagen, ed. 26. P. Waldteufel, personal communication (1969). 27, G. Vasseur, J. Atmos. Terrest. Phys, ~, 28. _, Phys. (in press, J. Atmos, Te,re$t. 397 (1969). 1970). 29. H. cam-u, M. Petit and P. Waldteufel, J. Atmos. Terrest. 30. R, H. Wand and F. W. Perkins, J. Geophys. Res. ~, 31. J. V. Evans and G.P. Mantas, J. Atmos. Terrest. 32. E.G. Fontheim, A.E. Beutler and A.F. Nagy, Ann. Phys. ~, 6370 (1968). Phyi. 30, 563 (1968). Geophys. ~, 489 (1968). 33. M. L. Duboin, G. Lejeune, M. Petit and G. Weill, J. Atmos. Terrest. 34. J. S. Nisbet, J. Atmos. Terrest. Phys. Q, 351 (1967). Pbys. 3!!, 299 (1968). 1257 (1968). 35. A. F. Nagy, E, G. Fontheim, R. S. Stolarski and A. E. Beutler, J. Geopbys. Res. ~, 4667 (1969), I 36. A. Dalgarno, M. B. McElroy and J. C. G. Walker, 37. A. Dalgarno, hi. B. McElroy, 1371 (1968). Planet. Space Sci. ~, 331 (1967). M.H. Rees and J. C. G. Watker, Planet. Space Sci. @, 38. P.M. Banks and A. F. Nagy, J. Geophys. Res. ~, 49 1902 (1970). 39. R. J. Cicerone 40. B.C. Narasinga 41. R. J. Cicerone, and S. A. Bowhill, Radio Sci. ~, 49 (1970). Rao and E.J. R, Maier, ]. Geophys. Res Phys. ~, J. Geophys, Res. 7J, 975 (1966). VIII, 800 (1968). — 45. D. B. Muldrew, J. Geophys. Res. ~, 2635 (1965). 44. L. G. Jacchia, Space Res. 46, R.B. 47. 1970). private communication (1969). 42. J. E. Geisler and S. A. Bowhill, J. Atmos. Terrest. 43. J, V. Lincoln, (in press, Norton, Proc. IEEE g, 1147 (1969). R. B. Norton and E. Marovich, Proc. IEEE ~, 1 ,.,’ 50 1158 (1969). 119 (1965). . . . . .... . .. . . .... . ... .. 00CUMENT (Sawrl,, 1. OR IGINAT, .l.,sl!l NO <s.!,.” L. CT, VtTY of ,1,1., b.adr (Cormrate.wth .[ ebe,,.., and CONTROL DATA - R&D Indexing Mmofe, ion must b. entered or) 2.. when the ..,0,, overall raw?, ,EC”,, T” i. CIa$elfied) CL ASS, F3CATION Unclassified Lincoln Laboratory, M. 1.T. 2b. 0.0”, None 3. REPORT TITLE Millstone Hill Thomson Scatter Results for 1965 $. IJ.SC.(P7,VE .o’r E5 (Type 01mporlandincl.sfve dales) J Technical Report ,. AU?. O.<S1(L.. * “m*, f,rel “m., i.ll;.W Evans, John V. i 7.. ,, REPORT OATS TOTAL NO. 8 December 1969 9., 3., ,. CONTRACT ,FZOJECT OR NO. GRANT 0, ,.., s 7b. NO. OF NO. QRIGINATOR>S AF 19(628)-5167 REFS 4? 56 REPORT ‘ NUMBER(S) Technical Report 474 649L ,h. OTHER REPORT assjgned . . this .0(S1 {A,!, ofhecoumber. that ma, be ra.oort) ESD-TR+9-3S7 d. !0. .v.l..,, LITY/., ulT10l0. PJO,lcE5 This docwr+ent has been approved for public release and sale; its distribution is unlimited. 1. 5UPPLEME.T&RY NOTES ,2, Ae, MILITARY ACTIVITY Air Force Systems Command, USAF None ,. sPONSOR(NG TRAc, This report presents F-region electron densities, and electron and ion temperatures observed during the year 1965 at the Millstone Hill Radar Observatory (42. 6°N, 71.5° W) by the UHF Thomson (incoherent) scatter radar. The measurements were made over 48-hour periods that were usually scheduled to include tie International Geophysical World days near fbe center of each month. Geomagnetic storm sudden commencements occurred during two of the observing periods, but do not appear to have giVen rise to marked variation of fhe densities or temperatures, On the ofher hand, measurements made during the progress of a major storm (15-19 June 1965) exhibit large changes compared with fhe behavior observed in fhe following month. The remaining months were magmetical.ly quiet, and tie density and temperature behavior were similar to those observed in 1964. Millstone is a midlatitide station exhibiting a characteristic “winter” and “summer” daytime density variation. The transition between tlwse two types occurs rapidly around equinox, without any corresponding seasonal change in the F-region thermal structure. Current ideas concerning the mechanisms responsible for fhis and other features of diurnal and seasonal variations are discussed. 1, KEY ‘//0.0S Millstone radar F -region diurnal variations electron density ionospheric scatter seasonal variations 5i temperature effects spectrum analyzers UNCLASSIFIED Security Classification . --- .—., —-z