Geophyisics and Astronomy in Space Exploration

8/7/2019 Geophyisics and Astronomy in Space Exploration

1/52

NASA SP-13

GEOPHYSICS,oASTRONOMY,, SPACEEXPLORATION

COPyNATIONAL AERONAUTICS AND SPACE ADMINISTRATION Washington, D.C.December 1962 Office of Scientific and Technical Information


2/52


3/52

ForewordThe NASA-University Conference on the Science and Technology of

Space Exploration, conducted in Chicago on November 1-3_ 1962_ was held "toprovide an authoritative and up-to-date review," of aeronautical and spacescience tec-hnolog_." - " .....

The scientific papers delivered at the conference were grouped for presenta-tion by topics and are, in effect, 1962 state-of-the-art summaries. Accordingly,NASA has published under separate covers sixteen groups of conferencepapers to make them conveniently available to those interested in specific fields.This series (NASA SP-13 to NASA SP-28) is listed by title and price onthe back cover.

All papers presented at the conference have also been published in a two-volume Proceedings (NASA SP-11) available from tile Superintendent ofDocuments for $2.50 and $3.00_ respectively. : Those papers presented hereinoriginally appeared on pages 107 to 151 of Volume 1 of NASA SP-11.


4/52

For sale by the Superln_ende_t of Documents, U.S. Government Print ing OE_ceWmJ,hing_on 25, D.C.. - Price 35 cen_


5/52

ContentsPage

THE SOUNDING ROCKET AS A TOOL FOR COLLEGE AND UNIVERSITYRESEARCH ............................................................. 1

ELEANOR C. PRESSLYSPACE FLIGHT STUDIES OF THE IONOSPHERE ........................... 7

ROBERT E. BOURD_AUPARTICLES AND FIELDS RESEARCH IN SPACE .......................... 21

GEOROE H. LUDWIOASTRONOMICAL RESEARCH IN SPACE ..................................... 31

JAMES E. KUPPERIAN, JR.AERONOMY RESF, ARCH WITH ROCKETS A_D SATELLITES ............... 39

N_S0N W. SPzN_

,,,111


6/52


7/52

SoundingocketsaToolorCollegendUniversityesearchBy Eleanor C. Pressly

3_SS ELEANOR C. PriESSL i,_ Head o/ the Vehicle Section, SouT_ding RocketsBranch o/ NASA',_ Goddard _paee Flight Ce_ter. She ha,_ been with God-daJ'd virtually since it,s' ineel)tion and t_ot(_ ha8 al_ importa_tt part in eonduetingNASA"s sound_;ng roeket programr. She holds an A.B. degree from E_',_ldneCollege, and an M.A. degree from Duke University. A former mathematicsteacher, she became a n_athematic;an at the Radio Research La5orato_7/ at Ilar-yard University. In 19_5 she joined the Naval Research Laboratory as amathen_,atieian and later beeame an aeronautical research engineer in theRoeket Sonde Branch. She has been closely associated with the Navy'sAe_vbee-Hi rocket and has followed the development o/ thi,_ vehicle to it,spresent configuration, now in u,e by NASA. 3li,s,,_ Pressly i._ a mere, her o/the American Roeket Society.

The sounding rocket is, indeed, an extremelyuseful tool for research in the space sciences.Since the end of World War II, it has providedthe means of obtaining information about ourupper atmosphere that up to that time had beenunattainable. Under the NASA soundingrocket program, it is used by colleges and uni-versities for scientific investigations in the vari-ous disciplines of geophysics and astronomy.The program now includes more than 10 schoolsin this country who are already conducting ex-periments and there are some cooperative proj-ects with foreign countries. I hope to showthe advantages of the sounding rocket and themethod by which a college or university mayparticipate in such research. Let us look at afew examples of what is being accomplished inthe NASA program. These particular experi-ments have been chosen in order to representdifferent rocket techniques.In the ionospheric research program, which

was dLseussed in the Geopltysics and Astronomysession yesterday, there was a series of threeflights to measure electron temperature andcharge density as a function of altitude usingthe Langmuir probe and the R.F. Resonance

Probe techniques. Thus there were two inde-pendent means of taking the same measurementssimultaneously. Two of the flights were madeat midday and one at night in order to studythe temporal changes. Data were collected be-tween 95 and 125 kilometers on each flight. Anartist's conception of the payload is shown infigure 10-1. The forward payload housing wasof a clam shell desiN1 which was ejected beforethe data collection altitude. This opened up

RESONANCEPROBESENSOR

SHELL

RESONANCEPROBE

"TELEMETRY ANTENNA

FIGURE 10--1. D-region ionosphere experiment.


8/52

SOUNDING

the Langnluir Probe sensor, which was mountedinside the nose tip, and permitted tile R.F. re-sonance probe antennas to be erected. Allthree flights were successful, and immediatepost-flight results confirmed agTeement betweentile two techniques. On tile night flight bothinstnlments recorded a sporadic layer 3 kilo-meters thick, beginlfing at 102 kilometers onboth up and down legs of the trajectory.Yesterday morning the Particles and Fields

program was discussed. Two projects in thatprogram are mentioned below.During the fall of 1960 a series of 14 rockets

was flown from Fort Churchill, Manitoba, Can-ada, to study the beams of high energy nucle-ons, which balloon experiments had shown toarrive at the earth following many large solarflares. Both nuclear emulsions and counterswere used, and recovery was required for tlleemulsions. These experiments were preparedand held in a standby condition awaiting solaractivity. Nine flights were made during thesolar flare period 11 to 18 November. Figure10-2 shows some of the results obtained froman emulsion flown on November 12. In all, tenof the nuclear emulsion packages were recov-ered and all except one of the flights producedgood telemetered data. This project was in-creased after the field operation began, and itwas accomplished with only nine payloads.The payloads that were recovered can be flownagain if the need arises.Another series of three experiments was made

to determine the altitude and intensity of elec-tric current systems over Wallops Island, Vir-

ROCKETS

ginia, using a proton precession magnetometer.Care was taken to make the rockets as noncon-ducting as possible, a special fiber glass payloadhousing having been designed especially for thisexperiment. The 50 pound payloads were car-ried to o00 kilometers altitude. On one of theflights, which took place during a magneticstorm, electric currents were found at 130 kilo-meters altitude. On another, fired during amagnetically quiet period, no currents werefound over Wallops Island. One of the pur-poses of these flights was to prove out aninstrument to be used in investigation of theequatorial ele


9/52

SOUNDING ROCKET AS A TOOL FOR RESEARCH

PA. PHILABALTIMOREMD.

CAMPAP

VA.LANGLEYIELD

!iiiiiii!!_i!iiiiii_i_!

FIGVRE 10-3.--Sodium vapor eh)ud.

firings will be undertaken. Camera stationsrecord the cloud produced by released sodiumvapor. As shown in figure 10-4, the camerasare located on the ground along a circle of ra-dius of the order of 100 kilometers from the fir-ing site. These firings are made at twilight o1"dawn, when the sodium trail is sunlit but thescattered sunlight is low.A companion to the sodium experimenL is one

that measures winds and temperatures between_)5 and 90 kilometers by ejecting and explodinggrenades during rocket ascent. Figure 10-5 isrepresentative of what takes place. The timeand direction of the sound arrivals are recordedby an array of sensitive microphones on theground. From these data_ winds and temper-atures are determined in the altitude layer be-tween each pair of grenades.A third synoptic experiment is the measure-

ment of electron density and electron tempera-ture in the ionosphere. Figure 10-6 shows thistype of payload. It is representative of thepayload packaging technique one uses inresearch with rockets.NASA will fly these three experiments as co-ordinated launches from Wallops Island andFort Churchill. Wind data obtained by thegrenade method and the sodium method will

CHERRYPOINTFIOURE lO--t.--Camera site locations for so(lium vaportests.be compared. The ionosphere experiment willuse these wind data for comparison with theelectron density profile to look for a possiblecorrelation between the strong wind shears ob-served and sporadic E. Other experiments inthis synoptic program measure atmosphereicsructure and composition.Several groups are preparing rocket astron-

omy payloads designed for spectroscopy andstellar and nebular photometry. These werepresented yesterday morning. Such payloadsare relatively heavy (about 200 pounds) andpeak altitude is on the order of 200 kilometers.

.... )

ii

3


10/52

SOUNDING ROCKETS

' i

i

FIGURE 10-6.--Payhmd for electron density measure-ment in the ionosphere.

All altitude control system has been developedprimarily for these experiments. An artist'ssketch appears in figure 10-7. This system iscapable of pointing the entire vehicle to a starwith an accuracy of -+2 and holding that posi-tion to within -% It can, for example, pointto as many as five different distinct targets inone flight with approximately 50 seconds of ob-serving time at each position. A fine altitudecontrol system is under development; it is tohave a pointing accuracy of -+30". Theseflights will also be testing the performance ofvarious components, both mechanical and elec-trical, that are to be used in the Orbiting As-tronomical Observatory.In another experiment a mass spectrometer

was carried to 960 kilometers to detect thepresence and possibly the relative abundance of

FIGURE 10-7.--Aerobee for rocket astronomy.

helium and hydrogen ions and to measure elec-tron temperature and ion density.The highest altitude sounding rocket flight

to date carried a radio astronomy experiment tomeasure the intensity of radio frequency energythat originates in the halo of our galaxy. Thepayload reached 1700 kilometers and had a datacollection time of more than 20 minutes. Thesensing system included a 40 foot t ip-to-tip elec-tric dipole deployed from the payload comp'lrt-ment after nose cone separation. The threeradiometers operated simultaneously from thedipole antenna. From all indications thisflight was a complete success.There are many questions in all of the scien-

tific disciplines that have not been answered.New groups are encouraged to participate inthese investigations. NASA Iteadquarters willbe receptive to proposals for experiments to beflo_m on sounding rockets. There are groupswithin NASA who are equipped to We guid-ance in all areas that may be new. Also, NASAis prepared to supply all support instrumenta-tion-that is, telemetry, power supply, perform-ance gages, aspect gages_ calibrator, commuta-tor, and all rocket wiring external to the ex-periment package. NASA can supply me-chanical hardware such as payload racks, nosecone ejection systems, and special windows andantenna mounts. In other words, the experi-menter is responsible for his own scientific in-strumentation, and everything else is providedas government-furnished equipment. Some ex-perimenters prefer to build the entire payloadpackage. In this event NASA can provide con-sultation and advice.NASA makes available all vehicles, perform-

ance analyses, and project coordination. In-eluded with the vehicles are standard nose cones,payload housing extensions and, for certain ofour vehicles, such accessory equipments as anattitude control system, recovery packages,special types of nose cones, and de-spin systems.The particular characteristics and the predictedperformance data of all of the rocket systemsare available. Special performance calculationsand heating and stress analyses are made on anindividual case basis as required. Contact ismaintained between the scientific experimenterand others who are providing support during


11/52

SOUNDING ROCKET AS

the payload buildup period, and arrangementsat the firing range are made.NASA's family of rockets ,tt the present time

includes seven vehicle systems. Figure 10-8is representative of these vehicles. An attempthas been made to arrive at a group with pay-load-altitude capabilities suited to experi-menters needs, and the same model is fired overand over again for reliability. These sevensystems cover the altitude regime between 100

.......

NIKE--CAIUN AEROBEE50A JAVELIN JOURNEYMAN

I_'IGURE 10-8.--Family of sounding rockets.

and 2500 kilometers. Figure 10-9 shows theperformance capability of each of these vehicles.Calculations are for sea level htunch at thestandard firing elevation angle for each rocket.Net payload is defined as all additions to thebasic vehicle, and it does not include the stand-ard nose cone housing. Recovery systems havebeen developed for the Nike Cajun, NikeApache, Aerobee 150A, and Journeyman. Asthe need for other capability arises and asproven vehicles become available to meet thaineed, such vehicles will be added. Most of thesevehicles have been used in several differentscientific disciplines.The weight and vohune characteristics of

payloads that can be flown on rockets availableto us now are shown in figure 10-'10. Noticethat except for the 500-700 kilometer region,the range to 2500 km is relatively well coveredby some payloads.The time required to prepare a payload for

firing varies with the specific experiment andthe amount of NASA support equipment neces-

A TOOL FOR RESEARCH

iooc_2 L . .. . _ten -._-'- -ALTITUDE _

NIKE "/'_"

_=_o - 50 1oo 20o 300.... NETPAY[.OADPDUNDS ................

Fm_raE 10-9.--Sounding rocket peak altitude in termsof payloads.

sary. The mininmm time period is about sixweeks, but such tinting presupposes that the ex-perimenter will be responsible for the wholepayload package and that rocket vehicles andtime on the firing range are available. Formalnotification to all firing ranges is required atleast 30 days before the expected launch date.A more normal and comfortable time intervalis six months; and for large projects geared tospecific time periods, such as the InternationalQuiet Solar Year, planning a year or so in ad-vance is appropriate.NASA's primary firing site for sounding

rockets is Wallops Island, Virginia. All of thevehicles may t)8 fired from that range. Otherranges from which NASA may launch certainvehicles are : Fort Churchill, White San(ls Mis-sile Range, Pacific Missile Range, and AtlanticMissile Range. In addition through inter-gov-ernment agreements, ranges are being or have==*__=" :_ _ MAXIMUM ..... ALT]ToDE ....

' _GWEIGHT DIAMETER USEFULOLUME RANGE__ ...... [CUBICEET} [KILOM[TE:RSI_20 - IN W I 230 - 110_i_ : _00 12 4.8 260- 180

-::_0::: i_ _o.o 2eo _6o8 : 2 0.9 500- 340

_--_ 30:252_ 18.25 37 [ 1250- 68030_-_:_: 1825 4,9 2500- 1400

FIOC'nE 10-10.--Sounding rocket l_lyluad statistics.

672058 0-63--2


12/52

SOUNDING

been set up in India, Norway, Sweden, andItaly. At the present time these latter rangescan accommodate the Nike boosted vehiclesonly.Let us look at the procedure one uses to obtain

NASA approval to do research by means ofsounding rockets.If you are interested and have an experiment

you would like to conduct, you should discuss itwith NASA representatives who are workingin the same scientific discipline as your experi-ment. They will probably suggest that youtalk with the rocket support group at GoddardSpace Flight Center as to the support avail-able. You will be asked to prepare a proposalfor presentation to NASA Headquarters. Yourproposal should include what you hope to ac-complish by your experiment, the method bywhich you obtain your data, the time period youwill require, and the number of firings you wishto make. It should also include a statement ofthe support you will need from NASA. It isnot necessary, that you specify a particularrocket type; however, you will probably havedetermined the best rocket vehicle. Your pro-posal will be reviewed for its scientific valueand for our ability to provide the support yourequire.After your project receives approval and the

funding arrangements have been made, we arein a position to make detailed plans. A con-ference that includes all groups who will beproviding support to the operation is arranged.The areas where support is needed are deter-mined, and a time schedule is set up. Theresult of this conference is that each group willknow their responsibility and what they mayexpect from each other group. A completeinterference check of the entire payload is re-quired at Goddard before the payload is takento the firing site. During this period environ-mental tests will be conducted and the payloaddynamically balanced if these are considerednecessary. The payload is mated to the rocketat the firing site. Preparation time therevaries, depending on the number and complex-ity of the various tests that must be run. W]_enthe rocket has been fired, all telemetry andtracking records are provided to theexperimenter.

ROCKETS

There are many reasons why we use thesounding rocket both as a complement to andpreparation for satellite experiments. It pro-vides a vertical cross section of the atmosphere,and it reaches altitudes between balloon peakaltitude and satellite perigee. It can be flownfrom selected places and at selected times; thusspecial events can be covered. Preparationiime is short and the program is sufficiently flex-il)le that discoveries can be followed up quickly.An experiment can be flown many times on asounding rocket to once in a satellite. This en-ables the scientist to modify his experimentafter a few firings and so broadens the totalamount of information obtained. Similarly itis good for synoptic work in which the sameexperiment is repeated over a period of time.Usually one experimental group instrumentsthe whole scientific payload and can use theentire flight time to best advantage, thus avoid-ing problems of compatibility between experi-ments. The sounding rocket is relatively in-expensive, and the logistics of firing it are rela-tively simple.A considerable number of sounding rockets

have been used as proving grounds for satelliteexperiments. Since one is doing basic research,it is appropriate to use the sounding rocket as ameans of determining the best experimental ap-proach to the satellite experiment, the areas ofgreatest interest, and the proper gains to set sothat all dat,_ return from the satellite will bemeaningful.Most of these advantages of the sounding

rocket contribute to making this program agood training for those who are beginning to doresearch using rockets and satellites.During the several years since the beginning

of NASA, 191 sounding rockets have beenflown. Of these, 20--or 10%--have been collegeor university prime experiments. Another 31flights, 16_f of the total, have had college oruniversity support. During 1962 there havebeen 18 flights in which colleges and universitiesparticipated, and 17 are planned for the rest ofthe year. If these flights are all made, thecollege and university participation will be 43_of the total program during 1962.


13/52

SpacelighttudiesftheonosphereBy Robert E. Bourdeau

ROBERT E. BOI_T.DEAUi8 [lead of the Planetary Ionospheres Branch at theNASA Goddard Space Flight Center. In addition to these duties, he wasProject Manager/or Explorer VIII and Project Scientist for the first inter-national satellite, Ariel, and for S-66, the ionosphere beacon satellite. Mr.Bourdeau was formerly with the Naval Research Laboratory, assisting inupper atmosphere research with the V2, Aerobee, and Viking rockets, andlater in atmospheric electricity research using aircraft. He received the B.S.degree in physics in 19/e3 from the University o/M;#ssachusetts, and is a mem-ber o/ the American Geophysical Union.

SUMMARYResults from rocket and satellite experiments desig-

nated to study the characteristics of charged particleswith thermal energies which comprise the ionosphereare discussed. Data obtained since 1959 are comparedwith ground-based observations and with spaceflightdata on the ionizing flux. Thls comparison then isused to update models of the ionospheric D, E, and Fregions. In addition, model concepts of the upperionosphere up to 2000 km including the helium ionlayer and the lower protonosphere are presented. Theresearch areas needing the greatest attention areoutlined.

INTRODUCTIONIonospheric physics is the study of the charac-

teristics of charged particles with thermal ener-gies, a group which exerts the greatest influenceon the propagation of radio waves. The studynecessarily includes the causal relationshipsbetween the charged particle characteristics, theionizing sources, and the ionizable constituents.Previous to spaceflight, knowledge regarding

upper atmosphere ionization was obtained fromionosondes, which are radar-like methods basedupon the reflection of radio waves by the ionizedregions. Early observers using this techniquewere convinced that the daytime ionization wasconcentrated in three separate and distinct lay-ers which they called E, F1, and F_. Thisincorrect conclusion was primarily due to the

limitation imposed upon the method by the factthat echoes cannot be reflected from a givenaltitude unless the density exceeds that found atlower altitudes. The early results of Seddonand Jackson, who used a rocket-borne two-fre-quency radio propagation experiment, showedthat the ionization is not concentrated in layersbut rather is significant at all altitudes.Because of their work, then, it is more appro-priate to use the term "region" rather than "lay-er" in discussing subdivisions of the ionosphere.The work of Seddon and Jackson has beenreviewed adequately by Ratcliffe (ref. 1) andwill not be detailed here.With the advent of the satellite age, new in-formation has been made available for altitudes

well above those previously studied. There-fore one can now discuss seriously the char-acteristics of low energy particles out to theinterplanetary plasma, rather than limit thescope to the original classical altitude region(up to approximately 300 kin) imposed by thelimitations of the ionosonde technique. It al-ready is appropriate to consider the upper Fregion, the helium ion region, and the protono-sphere as subdivisions of the ionosphere to beadded to the classical D, E, FI, and F'2 regions.The principal purpose of this report is to

compare theories of formation for each iono-7


14/52

GEOPHYSICS AND ASTRONOMY

spheric region with existing spaceflight obser-vations. When this is accomplished, it will beseen that one is left with inadequacies that jus-tifiably require continuing consideration of theconflicting theoretical models which have beenproposed. These inadequacies also invite thegeneration of new theoretical models, and area measure of the vast program that remains tobe accomplished, not only in the form of addi-tional spaceflight observations but also in theform of supporting theoretical and laboratoryresearch.

THE D REGION UNDER QUIET SOLARCONDITIONS

The lowest altitude region where a significantnumber of free electrons are found lies between50 and 85 kin. Because of the relatively highgas densities, this D region is characterizedalso by high collision frequencies between theelectrons and the neighboring neutral constit-uents. In these collisions, electromagneticenergy is transferred irretrievably to the neu-tral gas so that the region acts as a strong ab-sorber of radio waves. Measurements of Dregion electron concentration are very difficultto make with ground-based techniques. Thesehave, however, provided estimates for daytimeconditions of 103 electrons/cm 3 near 80kilometers.The most favored theory of the formation

of the daytime D region during conditions oflow solar activity is that proposed by Nicoletand Aikin (ref. 2). The end result of their

ALTITUDEfKM] ....................................................

FIGUI:E ]l-l.--The daytime D region under quiet solarconditions.

hypothesis is illustrated in figure 11-1 in theform of the altitude dependence of electron (n_)and positive ion (n+) concentrations. At al-titudes below 70 km (dashed portion of thecurves), cosmic rays are the principal radia-tions causing ionization and the diatomic ionsof nitrogen and oxygen probably exist in great-est abundance. Between 70 and 85 km (broadportion of the curves) Lyman a radiation(1215.7X) is the dominant ionizing agent eventhough it can only act on a minor constituent,nitric oxide. In the base of the E region (85-100 km), X-radiation is an important ioniza-tion source with O:, and NO + believed to be themajor ionic constituents. There is a high prob-ability that newly formed electrons can attachto neutral particles to form negative ions onlyin the lower part of the D region. The pre-dicted negatire ion abundance is the differencebetween the theoretical n_ and fl curves in figure11 1.The model of Nicolet and Aikin was calcu-

lated.from an expansion of the following con-ventional equation of ionization:

---- anJ, (1)where q is an equivalent production of elec-trons and a is an effective loss process coefficient.The expansion includes the differing effects ofall possible ionizing sources and several lossprocesses. For a single source and a singleionizable constituent an equation of the fol-lowing form results :q- (2)where a,_denotes the loss coefficient for recombi-nation of positive molecular ions with electrons,a_ denotes the loss coefficient for recombinationof positive with negative ions, and h is the ratioof negative ions to electrons. The value of _. iscomputed from the probability of electron lossin lhe formation of negative ions and the effectof solar radiation in photo-detaching electronsfrom the ions. The charged particle densitiesare computed from spaceflight observations ofcosmic rays and solar radiation, laborato_-measurements of the loss processes, and an as-sumed model of the number density of ionizableconstituents.

Principally because of the low ratio of('harKed particle to neutral gas density, electrondensities in the D region are difficult to measure


15/52

SPACE FLIGHT STUDIES OF THE IONOSPHERE

from a rocket, although refinements of existingexperiments should soon provide valid resMts.Some rocket data on the positive ion character-istics are available and these are consistentqualitatively with the Nicolet-Aikin model.Bourdeau, Whipple, and Clark first measuredthe positive ion conductivity (_r+) with a rocket-borne Gerdien condenser and confirmed cosmicradiation as the ionizing agent up to 60kin (ref.3). Positive ion densities can be estimatedfrom the observed conductivity according to

a+--n+e]+,(3)

_qmre e is the elementary charge and k+ is theion mobility, whose altitude dependence ismeasurable in the laboratory.More recently, Smith (ref. 4) obtained a posi-

tive ion conductivity profile by use of a deprobe analogous in principle to the Gerdien con-denser. The positive ion densities which hecomputed from the observed conductivity areplotted in figure 11-1. At an altitude of ap-proximately 90 km the theory of the experimentrelevant to positive ion characteristics breaksdown. Above this altitude the same device actsa Langmuir probe, thus permitting a measure-ment of electron density. Converse to the posi-tive ion determination, the Langmuir probetheory becomes inadequate below about 85 kin.For these reasons, the election and positive iondensity values in the region of overlap (85-90kin) are bracketed to indicate the possibility ofsecond-order errors. Except for the lowestaltitude point, the shape of the experimentalcharged particle density profile is similar to thatof the Nicolet-Aikin model. It favors the con-cept of three separate ionizing sources in the 50-100 km region. The individual accuracies ofthe electron and positive ion densities are be-lieved high enough in the 85-90 km region toconclude that negative ions become unimportantin the daytime ionosphere above 85 kin. Thegeneral excess of the experimental positive iondensities over the theoretical estimates may beinterpreted as an altitude invariant error eitherin the ion mobility used for the experimentalresult or in one of the assumed theoreticalparameters.Popoff and Whitten (ref. 5) have challenged

the Nicolet-Aikin model by proposing X-rays

as the principal ionizing agents above 70 km.Nicolet and Aikin did consider X-rays usingpublished data from Friedman and others.Popoff and Whitten used a flux value, integratedover the responsible wavelength region (2-8_), of 10 -3 ergs/cm_-sec. Aikin (ref. 6) con-siders this flux value either as typical ofconditions of a disturbed sun or as an upperlimit for a quiet sun during the solar maximum.Recent data from the Orbiting Solar Observa-tory I (W. White, private communication) andthe Ariel I satellite (ref. 7) are consistent withAikin's conclusions.It is believed, but not generally accepted, that

the Lyman , flux is relatively constant withtim solar condition and throughout a solar cycle.The X-ray flux, on the other hand, varies by afactor of 1000 (ref. 8). The question of therelative influence of these mmzmg agents onthe normal daytime D region will not be re-solved with certainty until valid charged par-ticle density profiles of both signs are measuredsimultaneously with the appropriate cosmicray, Lyman _, and X-ray fluxes, as well as thenumber density of the ionizable constituents.It as possible that such correlative data willshow that the two source candidates for the70-85 km region will alternate in relative in-fluence depending on the exact solar condition.It is also possible that the normal D region isextremely more complicated than that describedabove. For example, Whipple (ref. 9) hascalculated ion densities from the original con-ductivity profile and attributes a resulting largedecrease at altitudes between 70 and 80 km toionic diffusion to dust which may be trappedtemporarily in this region of temperature in-version shortly after meteor showers. Thissort of loss mechanism has not been consideredin the previously described theoretical models.When one considers that the theoretical and

experimental data presented in figure ll-1 arenot totally consistent and furthermore that theyare representative of a particular latitude, aspecific season and time of day, and a uniquep'trt of the solar cycle, it is realized quite easilythat the study of the O region has just begun.The geographic and temporal coverage must beextended by ,_ vast nmnber of rocket flights,domestically and on an international scale.


16/52


Before this can be meaningful, however_ newexperiments are needed. Valid methods ofmeasuring electron density below 80 km stillescape us primarily because of the low ratio ofcharged to neutral particle density. There areno existing spectrometers for measuring eitherthe ion or the neutral composition in this regionof relatively high pressure. For example, thevery existence of the minor constituent, nitricoxide, which the Nicolet-Aikin theory requireshas not been verified experimentally. This lackforces the theoretician to make unverified as-sumptions in estimating both the productionand the loss rate of electrons. There remaininconsistencies in the absolute values and timevariations of the ionizing radiation. As for theloss coefficients which are measurable only in thelaboratory, we are faced with orders of magni-tude disagreement between investigators forsome important reactions. On the theoreticalside, additional loss mechanisms, particularlythat suggested by Whipple, bear further investi-gation.

THE D REGION UNDER DISTURBED SOLARCONDITIONS

During periods of high solar activity, the Dregion is characterized by enhanced ionization,with associated strong attenuation of electro-magnetic waves causing radio blackouts. Thereare many phenomena associated with solarflares which increase the normal D region elec-tron densities by up to two orders of magni-tude. A comparison of the amount of enhancedionization for disturbed solar conditions ismade with tile quiet mm model in figure 11-2.Simultaneously with the appearance of a

flare, radio absorption is observed in the Dregion on the sunlit side of the earth for periodslasting up to approximately one hour. Thisparticular event is called a Sudden IonosphericDisturbance (SID). Friedman, et al. (ref. 10)made rocket flights into such an event andobserved abnormally high X-ray fluxes pene-trating as low as 30 kin. It has been ob-served that not only does the overall intensityincrease but also that the spectrum ill the_.-SX region "hardens" in that the flux per unitwavelen_h is much broader than the quiet sunspectrum. Nicolet and Aikin (ref. o,9.) have

JJITUDE IKILOM_ERS] ........

| f21o_ENONOSf'._RIC.-'" " /{3] POLARCAPABSORPTION / _ --[4] POLARCAPABSORPTION f / *". -g$ "*_'-1 //fJ'f'P_

j" ROCKETMEASUREMNT ,,"

_ lo2 lo3 i0 _O_----_ELt,.CT_OHENSITY_CM_

FIGURE ll-2.--Comparison of the D region under quietand disturbed solar conditions.

used these observations to infer an electrondensity profile for the SID condition (curve 2of fig. 1i -2) in a manner similar to the develop-ment of the quiet sun model. Here, the in-fluence of cosmic ray and Lyman (_ radiationbecomes minor, the secondary layer in the 65-80km region disappears, and the profile ischaracterized by overall enhancement and amonatonically increasing electron density.A second type of ionospheric storm is asso-

ciated with active auroral and magnetic dis-turbances and predominates at night. Theionizing agents are believed to be energeticparticles comprising corpuscular emission fromthe sun. This belief is founded on the observa-tion that D region absorption occurs some 21hours after the appearance of a flare, an inter-val corresponding to the sun-earth transit timefor these particles. During the storm, D regionelectron densities increase to values highenough that echoes are observed on ionosondes.Layers appear at 90 km during weak geomag-netic activity and as low as 70 km for the moreactive events.A third type of disturbance occurs at auroral

latitudes. These phenomena, called Polar CapAbsorption (PCA) events, are produced byenergetic protons emitted from the sin1 duringcertain solar flares. Here echoes are ob_r_-edfrom ionosondes at altitudes as low as 60 km.The phenomenon has been the subject of con-siderable study during the past few years.M'ost recently Maehlmn and O'Brien (ref. 11)have proposed a semi-quantitative time history

10


17/52


of the D region electron density profile duringa PCA by inserting proton fluxes observed onthe Injun satellite as the source function in anexpansion of the conventional equation of ioni-zation. One of these profiles (curve 4 of fig.11-2) coincides with the peak of the 27.6 mccosmic radio noise absorption (17 db) observedat College, Alaska. Kane and Jackson (ref.12) performed a rocket measurement of the Dregion electron density profiles (curve 3 of fig.11-2) during a less active phase of a PCAevent (3 db absorption at 30 mc). This com-pares favorably in shape but is generally lessenhanced than that proposed by Maehlum andO'Brien during an interval when the observedradio absorption also was 3 db.The status of our knowledge of the altitude

distribution of charged particles for the multi-plicity of disturbed conditions is exemplifiedby the fact that the dashed curves in figure 11-_are speculative. We have some information onthe ionizing radiation. The speculation startswith the absence of observational electron dens-ity information during special events and nec-essarily includes a model of the ionizable con-stituents which have not been measured. Ad-ditionally, a new set of loss coefficients, depend-ing on the reaction of the ionizable constituentswith energetic particles rather than ultra_doletand X-ray radiation, need to be measured in thelaboratory.THE ION CONTENT OF THE LOWER IONOSPHERETo a high degree of probability, the predomi-

nant ionizing sources of the E and F regionfor quiet solar conditions at mid-latitudes aresolar ultraviolet and X-ray radiations. If oneknows the altitude dependence of the numberof incident photons at each wavelength, of thedensities of the individual neutral constituents,and of the absorption and photoionization crosssections of these constituents at each wave-length, it is possible to compute the rate at whichdifferent ion species are formed. Few of theseparameters are known. There are availablesome altitude profiles of the photon fluxes asa function of wavelength but only for limitedlatitude and temporal conditions. Results fromrocket-borne neutral gas spectrometers reportedto date are extremely controversial, principally

because recombination effects within the instru-ments distort the gas under study from its am-bient condition. Thus, our understanding of thecomposition of tile neutral gas is somewhat spec-ulative. Another serious problem is that someof the absorption and photoionization cross sec-tions are not known to an order of magultudeand others have not yet been investigated.

ne[CM _i =

FIGURE ll-3.--Productton and actual abundance of ionsin the lower ionosphere.

The most recent estimate of the altitude de-pendence of the rate at which various ions areproduced was made by Watanabe and Hinter-egger (ref. 13) and is illustrated in the left-handside of figure 11-3. They hasten to point outthat in view of the uncertainty in our knowl-edge of the neutral atmospheric compositionand of the pertinent cross sections, the curvesare suggestive rather than quantitative. Evenif the production rates were known, the prob-lem of explaining the detailed altitude distribu-tion of the electron density, also illustrated inthe left-hand side of figure 11-3, is extremelycomplex.The chemical complexity of the ionosphere

can be demonstrated by comparing a rocketmeasurement (ref. 14) of the actual ion abun-dance (right-hand side of figure 11-3) withthe inferred rates of production. It is seen, forexample, that even though the diatomic nitro-gen ions (NF) are expected to be produced ingreat quantities, the spectrometer results showthat they exist as an extremely minor chargedconstituent. The most likely reason is that ,_rFions dissociatively recombine almost as quickly

11


18/52


as they are produced. As a second example, itis seen that the spectrometer observations showNO + to be one of the predominant ion speciesbelow 200 km even though they are not formedas a direct result: of solar radiation. The mostlikely process leading to the NO abundance isthe reaction of monatomic oxygen ions (0 _)which are directly produced with molecularnit rogen.Considering only the two reactions indicatedin figure 11-3 greatly oversimplifies the mechan-isms by which the _rious ionospheric regionsare formed. There are many reactions thatneed to be studied in the laboratory, includingion-atom interchange, charge exchange and dis-sociative recombination, to name a few.Exhaustive lists of the important reactions havebeen prepared (ref. 15) but there is no assur-ance that any of these lists are complete.It is important to emphasize the need for a

long range program of simultaneous vertical-sounding rocket observations of the solar radia-tion, neutral gas parameters, and chargedparticle characteristics. It is just as importantto underline the need for supporting theoreticaland laboratory research. When such a longrange program has led us to universallyaccepted sol.utions for the continuity equation,it should 1)e possible to l)redict the state of theionosphere below an orbiting observatory whichmonitors only ionizing radiations.

THE DIURNAL VARIATION OF THE E REGIONThe altitude region arbitrarily assigned to

the E region lies between 85 and 140 kin. Mosttextbooks favor a theory of E region formationbased on general ionization of air by softX-rays, yet IIintere_er and Watanabe (ref.16) have recently disputed this hypothesis onthe basis of more refined measurements of thesolar ultraviolet radiation. The rocket, ionspectrometer work of Johnson, Meadows, andHolmes (ref. 17) first showed that the majorionic constituents are 0: _ and NO +. The pio-neering effort of Seddon and Jackson with arocket-borne radio-propagation experimentshowed that tim electron density in the daytimebehave with altitude in a manner similar tothat illustrated in figure 1I-4. The m'ljor dis-covery of this work was that, in terms of the

ele_.tron density distribution, the F region is acontinuation of the E region. Considering thedaytime electron density profile shown in fig-ure 11-4, we can see that a ground-based lone-sonde would show reflections only up to analtitude of about 105 km and |hus is unable todefine the electron density distribution betweenthat altitude and the base of the F region.-__ j ]

NASAROCKET10.5127 OCTOBER9610435 EST

INASAROCKET10258 0EC1960I152 EST

SPORADIC-E/ROCKET10.5217 AUGUST19Sl2206 EST _-"r _

I I 1015lo_ io3 104 io_ELECTRONOENSITY[CM :_i

FI_t :RE l l- 4. --Diurna l variation of the E region.

It is important to pause here and dissolve apopular misconception that spaceflight experi-ments are difficult to conceive and implement.The experimental method used to obtain thesets of data illustrated in figure 11-4 consistedof telemetering the current to the nose tip of arocket as a function of a variable potentialapplied between that electrode and the mainrocket body. In obtaining the end result, it isa more challenging task to apply kinetic theoryand derive the desired charged particle param-eler from the measured current than to imple-ment the experiment.Until very recently, there have been no

observations of nighttime E region ionizationbecause neither the ground-based ionosonde northe rocket-borne propagation experiments aresensitive enough. All the rocket data illus-trated in figure 11-4 were ol)tained t>y use of ade probe (ref. 4 and 18).Two nighttime electron density profiles are

illustrated in figure 11-4, one obtained in theearly evening hours and the other close to localsunrise. The general decay in the average elec-tron density of about two orders of magnitudefrom the daytime condition is what would be

1,2


19/52


expected from a limited knowledge of therecombination coefficients. Both of these pro-files show deep valleys of ionization above 110km. The resulting ledge of ionization at alti-tudes of about 100-110 km are suggestive of anighttime source of ionization. These valleyshave been observed also by Japanese investiga-tors (ref. 19), who used a rocket-borne radio-frequency oscillator based on the principle thata unique resonance is observed at the plasmafrequency.One anomaly which is observed quite fre-

quently in both the daytime and nighttime Eregion is the sporadic-E layer. These areshown as dashed perturbations on the twonighttime profiles in figalre 11--4. Rocket ex-periments have defined this condition, which toan ionosonde constitutes reflection at a constantvirtual height, as enhanced ionization confinedto a narrow altitude region of less than one kilo-meter. Consequently,_tle term "layer" is appro-priate in defining this anomaly. There perhapsare many mechanisms producing sporadic-E.One of the likely causes is mechanical redistri-bution of the electrons from the regions aboveand below induced by wind shear. Anextremely interesting exercise soon to take placewill consist of the simultaneous launchings oftwo rockets, one instrumented for ionosphericstudies and the other to measure the wind vec-tor by the sodium release method. Should thesporadic-E layer and a pronounced wind shearcoincide precisely in altitude at the same time,this hypothesis will have been demonstrated at]east from an empirical point of view.

THE FORMATION OF THE F2 PEAKThe altitude region generally assigned to the

lower F region lies between 140 and about 300km. As illustrated in figure 11-3, there is aninflection point at about 140 km in the other-wise monotonically increasing electron density.For this reason the F1 and F9 subdivisions havebeen invoked. As a result of ground-based ion-osonde investigations, it has long been estab-lished that the maximum electron densityoccurs at about 300 km. Until the satellite age,the region above this altitude was open tospeculation.

The most favored theory of the formation ofthe F2 peak is illustrated by comparing a meas-ured altitude dependence of electron density(ref. 20) with the rate of electron productioninferred by a Watanabe and Hinteregger. Thiscomparison is shown in figure 11-5. It is

F'IOURE lI-5.--The formation of the F region.

obvious immediately that the altitude of maxi-mum electron density lies considerably abovethat at which the production rate is highest.This perhaps is best explained by consideringthe following simplified form of the continuityequation: COlTesponding to quasi-equilibrium.

q= _Sn,-_d(n.WD) (4)dz Jwhere q is the production rate. The firstterm on the right-hand side is representative ofan attachment-like loss mechanism, whereas thesecond term defines change in electron densityby a diffusion process. In the region up to theF2 peak the loss by an attachment-like proc-ess predominates. Because this loss rate de-creases more rapidly with altitude than does theproduction rate, the electron concentration in-creases up to its maximum value. At the F2peak the diffusion term begins to predominateand the electron density begins to decrease. Thediffusion mechanism is caused by gravitationalforces acting upon the ions which, by coulombattraction, cau_ the electrons to diffuse down-ward. At altitudes of about 400 km and above,both the production and loss processes becomeminor factors. As a result of diffusion, the

672058 0-63--31.3


20/52


electron density distribution in this region cor-responds to hydrostatic distribution as follows:

d(ln n,) m+gdz _k(T.T TO' (5)

where m* is the ionic mass, te and t/ the elec-tron and ion temperature respectively z is thealtitude, g the acceleration of gravity, and t_ isBoltzmann's constant.It should be en-_phasizetl that what has been

presented is a favored theory for the formationof the F:?,) peak which is not univergally ac-cepted. Some investigators haw propo_dmodels without resorting to the diffusion pro-cess, by invoking loss mechanisms associatedwith minor ionic constituents. Such trace con-stituents have been observed on spaeeeraftscarrying ion spectrometers. The exact solutionof the continuity equation for the region nearthe F2 peak has been the subject of considerableinvestigation and yet is not resolved satisfac-torily. Furthermore, all theoretical models in-set't: only solar ultraviolet flux into the produc-tion function. Yet, as a result of satellitesdesigned to study the Van Allen belts, largefluxes of quasi-energetic particles which couldprovide an additional ionization source havebeen observed at some geographic locations inthe upper ionosphere.THE HELIUM ION LAYER AND THE BASE OF THE

PROTONOSPHEREAs shown in Equation 5, the electron den-

sity at. altitudes of about 400 km and above isgoverned by the ionic composition and theatmosphere temperature. Rocket-borne spec-trometer data, typically represented in figureI I-3, showed that 0 + ions begin to predominateat an altitude of about .o00 kin. It was laterestablished by the Sputnik III (ref. 21) andExplorer VIII satellites (ref. 22) that this par-tieular constituent was dominant up to 1000 kmfor a daytime ionosphere during the middle ofthe solar cycle. It generally was hypothesizedthat a direct transition from oxygen to hydro-gen ions would take place just about 1000 kmuntil Nicolet (ref. 28) suggested that helium isan important neutral constituent of the upperatmosphere. He based this suggestion on ob-set_-ations of drag on the Echo I satellite.

There is now direct and indirect evidence thationized helium plays an important role in con-trolling the electron density in the upperionosphere.

-- L0

0.8

__ 0.G- 04

0.2

O

T = 1750 _'KI _ \ EXPERIMENTAL0

I N _ EXPERIMENTAL

2 4 G IORETARDINGPOTENTIALVOLTSI

SATELLITESXIN

',i M+Vs:*

FIGURE ll-6.--Direct detection of helium ions from theExplorer VIII Satellite.

The first direct evidence was provided by aretarding potential experiment flown on theExplorer VIII satellite (ref. 24). This ex-perinaent is illustrated schematically in figure11-6 together with the experimental observa-tions. The sensor consists of three electrodesconcentrically arranged in planar geometry.As a result of the negative bias applied to theinner _'id, the collector reports to the telemetrysystem only that current which is due to posi-tive ions flowing from the medium under studyinto the sensor. Because the satellite velocityexceeds the thermal velocity of the ions, theirkinetic energy relative to the spacecraft is gov-erned by their mass (M.) and the known satel-lite velocity (Vs). This kinetic energy can berelated to the potential energy, due to a knownretarding potential (,I,c) applied to the collec-tor, by

M+ V_2 =_b,e. (6)

Because the thermal velocity of the ions can-not be neglected completely, the instrumentfunctions as a "poor man's ion spectrometer"in that the ion resolution is coarse.The interpretation of the experimental volt-

ampere curves from the Explorer VIII satel-lite is summarized in the left-hand side figure11-6. The predominant constituent can be

14


21/52


identified either from the potential required toretard half of the ions or from t)m shape of thecurve, in accordance with the theoretical workof Whipple (ref. 9.5). For daytime conditionsduring the active life of the satellite (Novem-ber, 1960), it was observed that O* ions pre-dominated up to 1200 km but that helium ionswere most abundant from 1"200 km up to maxi-mum observational altitude (1600 km). Theaverage daytime electron temperature observedon Explorer VIII during this interval was1800K (ref. 26).The first indirect evidence for a helium ion

layer was reported by Hanson (ref. 27) usingan ion density profile obtained by Hale (ref.28). From the same data Hanson also reportedthe detection of the base of the protonosphere atan altitude of 3400 kin. The atmospheric tem-perature deduced from the scale height of theelectron-ion gas was 1600K. Hanson alsosuggested that the thickness of the helium ionlayer should be relatively invariant withdiurnal time. On the other hand, Bauer (ref._9) theoretically demonstrated that there is adiurnal variat ion.

tOs 104ELECTRONE_[CM ":_] -

L_EOMETRICALTITUOEKm_

NASAST7 @-21 jWALLOPSSLANO,VA. _19 OCTOBER96! _20001238EST 4

|OO0

105 I0 6

FZGURF, ll-7.--Indirect detection of helium ions fromvertically launched Scout rocket,

The indirect method of determining the ioniccomposition is illustrated in figure 11-7. Thesolid and dashed lines are theoretical predic-tions of the electron density distribution.They were computed by an expansion of Equa-tion 5 which included more than one ionicconstituent, and with the assumption that theatmospheric temperature is constant with alti-

tude. The circles represent electron densitiesobtained experimentally by Bauer and Jackson(ref. 30) using a radio-l)rol)agation technique.The right-hand ordinate is true altitude andthe left-hand ordinate is a reduced altitudescale which takes into account the change in theacceleration due to gravity. It is seen that theexperimental data are consistent with a theo-retical curve for a binary mixture of oxygenand helium.

LEGEND /

1BOURDEAUt al {EXPLORERYlllJ_-2 HANSON(NASAST2) ]v3 BAUER& JACKSON /- INASAST-7} /

4 WILLMOREet al (ARIEL]/_........ -5 DONLEYNASAST-9/ -]

- HYDROGEN -_

OXYGEN_600 800 1000 1200 1400 1600 180_

_K]FIGURE ll-8.--Ion Lransttion altitude as a function of

atmospheric temperature,

Most observational data on the ionic composi-tion support Bauer's suggestion of a strongdependence of the thickness of the helium ionlayer on the atmospheric temperature, particu-1,'trly his most recent, theoretical model (ref.31). This predicted thickness as a function ofthe atmosl)heric temperature is given by theseparation of the solid lines in figure 11-8. Thecircles represent experimental observations ofthe altitude at which the helium and oxygenions exist in equal concentrations. Tile squares,also experimental, are the altitudes at which thehelium and hydrogen ions have the same hum-

15


22/52


ber densities. The data sets 1, '2, and 3 were dis-cussed above. A significant contribution hasresulted from the Ariel I satellite (ref. 3_9).The experimental approach used on Ariel I is amuch improved version of the retarding poten-tial experiment. During the time of these ob-servati.ons (May, 1962), helium was detectedas low as 600 km for ,_ nighttime condition andat about 900 km during the daytime. For thelatter condition, protons were not observed upto the apogee altitude (1200 km). However,they were detected at this altitude during night-time passages. The last set of observations areby Donley (to be published) who used the directand indirect methods simultaneously on a Scoutrocket launched into the nighttime ionosphere.The ionic composition of the upper ionospherecould be more variable than indicated by figure11-8. It should be recog'nized that the obser-vational data was obtained only at low latitudes.Secondly, some rocket flights have been made atnight into this region where the detection ofhelium ions has not been reported (ref. 33).This could either be due to a larger temperaturedependence than indicated in figure 11-8 or toexperimental inadequacies. A wider latitudi-nal coverage is required. More importantly,the altitude region above 4000 km has yet to beexplored for thermally charged particles byspaceflight experiments. The altitude regionup to 10,000 km is being investigated by ground-based radar backscatter apparatus (ref. 34).

IONOSPHERIC TEMPERATURESThe temperature of the neutral gas (T) has

been t!_e subject of considerable study. It isgenerally accepted that this temperature be-comes quite independent of altitude above 300kin. It will be the purpose of this section tocompare obsmwations of charged particle tent-peratures with reference neutral atmospheres tointroduce additional questions in the ah'eadyconflicting theories of heat..rod ionizationsources in the upper atmosphere. This can bestbe accomplished by limiting the discussion toaltitudes above 450 km where the electrontemperature (7',) is nearly equivalent to thatof the neutral ,,'as, (ref. 35), although this isnot universally accepted.Two types of observations are available.

The first method involves the use of Langnmirprobes to measure Te directly. The secondmethod is indirect. Referring to Fignre 11-5,we see that if the ionic composition is knownand if one has an accurately measured charged-particle profile it is pdssible to compute the sumof the electron and ion temperatures. This sumis related to the neutral gas temperature by"T: (T_+ T,)/9.It is known that in the isothermal altitu:te

region the neutral gas temperature varies con-siderably with diurnal time and the solar cycle.This variation can be correlated with solar deci-meter flux which is observed at the earth andwhich serves as an index of solar activity.Neutral gas temperatures have been estimatedfrom observations of satellite drag. It shouldbe noted, however, that it is the atmosphericdensity, not temperature, which is directly ob-tained from satellite drag. The temperature iscomputed from the observed density and anassumed neutral gas composition. As em-phasized previously', the latter parameter repre-sents a major observational gap. It would beexpected t]mt if the absorption of solar ultra-violet radiation were the only heating acting onthe atmosphere, then the maximuna temperatureon a given day would occur near sunset. Whatis observed is that the maximum density occursin mid-afternoon. Some investigators implythat the time of maximum density correspondsto the time of maxinmm temperature and thenreason that because of the phase shift fromsunset to mid-afternoon an additional heatsource must be introduced. This same hypoth-esis is used to explain observed seasonal varia-tions of atmospheric density. One heat sourcethat has been suggested (ref. 36 and 37) iscorpuscular radiation associated with the solarwind which should not be latitudinally depend-ent. The charged particle temperature obser-vations are not completely consistent with thesereference atmospheres. It is not known withcertainty whether this is hecause of a lack oftemperature equilibrium or of inadequacieseither in the neutral gas models or the chargedpart i('le temperature measurements.Langmuir probes flown on the Exph)rer VIII

satellite showed a diurnal variation of the elec-tron temperature from 1000K at night up to

16


23/52


1800K during the day, at a time when the solardecimeter flux index was 150 (ref. 26). Thesemeasurements were not highly accurate (ref.38). Furthermore, the lack of a tape recoMerprevented the establishment of an exact diurnalvariation and any observation of latitude de-pendence. More complete data are availablenow from the Langmuir probe experiment usedon the Ariel I satellite which did contain a taperecorder. The results of Willmore, eta]. (ref.32) are reproduced as dashed lines in figure11-9. Plotted are three curves of temperature

| !} 4 8 i2 .... 16 20 24---

The Ariel I observations correspond to a solardecimeter fluz index of 100. The neutral gasmodel of Harris and Priester (ref. 36) for thislevel of solar activity is plotted in figure 11-9for comparison with the Ariel I data. AsHarris and Priester point out, the nature ofsatellite drag observations is such that themodel is applicable only to low latitudes. Theirentire temperature curve was multiplied by 1.3to achieve agreement with the nighttime valuesat the lowest latitude of the electron tempera-ture data. Rather good agreement would beachieved with the low latitude electron tem-perature curve if the atmospheric compositionat high altitudes was assumed to vary withtemperature in a fashion corresponding to thatillustrated for ions in figure 11-8, rather thanaccording to the variation which was used byHarris and Priester.

FIGURE 11-9.--Diurnal and latitudinal variation of ex-ospheric temperature.

as a function of local time, each curve corre-sponding to a different latitude region. Thereare three important observations to be madefrom these data. First, there is a pronouncedsunrise-effect which most probably represents adistinct departure from temperature equilib-rium. As the sun rises, the electrons tempera-ture would be expected to rise considerablyabove the neutral gas temperature. As a con-sequence, the sunrise effect should be ignored inthe comparison with reference atmospheres.The second important observation is that

there is no pronounced peak in the mid-after-noon. On the assumption of temperatureequilibrium during this part of the day, thiswould be inconsistent with the inferred diurnaltemperature variation contained in referenceatmospheres. As Willmore, et al. point out, themost important observation is the pronouncedlatitude effect, which they ascribe to energeticparticle bombardment in the auroral region.

FZOUaE ll-10.--Diurnal variation of exospheric tem-perature at mid-latitudes.

To give further confidence to the Ariel I elec-tron temperature results, data taken from thecurve corresponding to latitudes between _0and 40 degrees are compared, in figure 11-10,with previous observations at these latitudes.The comparison necessitated normalizing alldata to a given solar condition. This was ac-complished by using the temporal variationsuggested by Harris and Priester. It is seenthat for the most part there is approximateagreement between the Ariel I and Explorer%TIII electron telnperature values and betweenthese electron temperatures and values of(T,+T_)/2 obtained by the indirect method.

17


24/52


Furthermore, there also is approximate agree-ment between the charged particle temperatureand values of the neutral gas measured by Bla-mont by the sodium release method (ref. 39 and40). These all are suggestive of electron tem-peratures being approximately equal to the neu-tral gas temperature at altitudes above 450 kmexcept during the sunrise period. However,the suggestion of temperature equilibrium im-plied by this comparison is based on the assump-tion that the normalization procedure for solaractivity is correct.It is important to note that the amplitude of

the diurnal variation corresponding to directlymeasured temperatures is larger than that ofthe model atmosphere which is based primarilyon satellite drag. The most likely reason is thatthe model atmosphere applies only to the near-

equatorial region and thus the difference is lati-tudinal in nature. This is the same conclusionreached by the Ariel I investigators. It is alsopossible that Te/T, is latitude dependent.If there is a latitudinal variation of T, the

hypothesis of corpuscular radiation associatedwith the solar wind as a second source of heat-ing becomes subject to doubt. A likely secondsource is auroral bombardment by energeticparticles, as suggested by Willmore, et al. Ifthis is so, a continuous ionization source at highlatitudes, in addition to the solar ultravioletand X-ray radiations, is also likely. The possi-bility of additional ionizing sources even forquiet solar conditions should expand an alreadylarge program of spaceflight observations,theoretical research, and laboratory investiga-tions.

LIST OF REFERENCES1. RATCL]FFE, J. : Physics Of the Upper Atmosphere. Academic Press, 1960.2. NICOLEa, M., and AIKI,_, A. C. : The Formation of the D Region of the Ionosphere. J.

Geophys. Res., col. 65, May 1960, pp. 1469-1483.3. BOURDEAV, R. E., WHIPPLE, E. C., and CLAR_:, J. F. : Analytic and Experimental Elec-

trical Conductivity Between the Stratosphere a_d the Ionosphere. J. Geophys. Res.,col. 64, October 1959, pp. 1363-1370.

4. SMITrl, L. G. : Electron Density Measureme_ts by the Asymmetric Probe. NASATeeh. Note, to be published.

5. POPOFF, I. G., and WttITTEN, R. C. : D Region Ionization by Solar X-lCays. J . Geophys .Res., col. 67, July ]962, pp. 2986--2989.

6. AIKIN, A. C. : The Formatia_ of the Ionospheric D Region. Proe. of the InternationalSymposium on Equatorial Aeronomy, September 1962, to be published.

7. POUNDS, K, and WILLMORE, A. P. : X-Ray Measurements on the Ariel Satellite. Proe.of the International Conference on the Ionosphere, July 1962, in press.

8. KREPLII_,', a. _V., CHUBB, T. A., and FRIEDMAN, H. : X-Ray and Lyman-Alpha Emission]rora the Sun as Measured 1rom the NRL SR-1 Satellite. J. Geophys. Res., col. 67,June 1962, pp. 2231-2253.

9. WHIPPLE, E. C. JR. : Direct Measurements of Ion De_sity a_d Conductivity in the D-region. XI th International Astron. Congress, 1960, pp. 99-102.

1O. FRIEDMAN, H., CHUBB, T. A., KUPPERIAN, J. E., KREPLIN, R. _V., and LI_,'DSAY, J. C. :X-ray and Ultraviolet Emission o/ Solar Flares. Ann. Geophys., vol. 14, 1958, pp.232-235.

II. MAEHLUM, B., and O'BRIEN, B. J. : Solar Cosmic Rays of July 1961 and Their Iono-spheric Effects. J. Geophys. Res., col. 67, August 1962, pp. 3281-3288.

12. JACKS0_, J. E., and KA_E, J. A.: Measurement of Iot_osphcrie Electron Densities Us-ing an RF Probe Technique. J. Geophys. Res., col, 64, August 1959, pp. 1074-1075.

13. WATANABE, K., and IIINTEREGGER, H. E. : Photoio_lizatio_, Rates in the E and F Re-gions. J. Geophys. Res., col. 67, March 1,_2, pp. 999-1006.

14. TAYI.0R, H. A., and BRINTON, H. C. : Atmospheric Ion Composition Measured AboveWa?lops Island Virginia. J. Geophys. Res., col. 66, August 1961, pp. 2587-2588.

15. AIKI'_, A. C. : Charged Particle Reactions of Importance in the Ionosphere. GoddardSpace Flight Center Rept. X=615-62-132, August 1962.

16. HINTEhEGOER, It. E., and _,VATANABE, K.: Photoionizatiotl Rates Dt the E and F Re-giom% 2, J. Geophys. Res., col. 67, August 1962, pp. 3373-3392.

17. JoH_'So_, C. Y., MEADOWS, E. B., and HOLMES, J. C. : Ion Compositio_ of the AtticIonosphere. J. Geophys. Res., col. 63, 1958, pp. 443_44.

18


25/52

SPACE FLIGHT STUDIfS OF THE tONOSPHEItE18. SMITH, L. G. : Rocket Measurements of Electron De_tsity and Temperature in the Night-

t ime Ionosphere. NASA Tech. Note, to be published.19. Aor_o, Y., HIRAO, K., and MIYAZAKI, S. : Positive Ion Density, Electron Density and

Electron Te_perature in the Ionosphere. J. of the Radio Research Laboratories,col. 8, 1961, pp. 453-465.

20. JACKSON, J. E., and BAUER, S. J. : Rocket Measurement of a Daytime Electron De_sityProfile up to 620 kin. J. Geophys. Res., col. 66, 1961, pp. 3057)--3057.

21. ISTO_fIN, V. G. : An Investigation o] the Ionic Con_positton of the Earth's Atn_osphereUsing Rockets and Earth Satellites. Artificial Earth Satellites, col. 2, 1962.

22. BouRnEAU, R. E. : Ionospheric Results with Sounding Rockets and Satellites. Proc.Second International Space Sciences Symposium, April 1961, pp. 5;)4-573.23. NICOLET, _[. " Helium, and Important Constituent in the Lower Exosphere. J . Geophys .

Res., col. 66, March 1961.24. BOURDEAU, R. E., DONLEY, J. L., _VHIPPLE, E. C., and BAUER, S. J. : Experimental Evi-

dence for the Presence ol Helium Ions Based on Explorer VIII Satellite Data. J.Geophys. Res., col. 67, February 1962, pp. 467--475.

25. WHIPPLE, E. C., JR. : The Ion Trap Results in Exploration o] the Upper Atmospherewith tlve Help of the Third Soviet Sputnik. Proc. IRE, col. 47 December 1959, pp.2023-2024.

26. SERBU, G. P., BOURI)EAU, R. E. and DANLEY, J. L. : Electron Temperature Measure-ments on the Explorer VIII Satellite. J. Geophys. Res., col. 66, December 1._1, pp.4313-4315.

27. H,_r_S0N, W. B. : Upper Atmosphere Helium Ions. J. Geophys. Res., col. 67, January1962, pp. 183--188.

28. HALE, L. C. : Ionospheric Measurements with a Multigrid Retarding Potential Ana-lyzer. J. Geophys. Res., col. 66, pp. 1954.29. BAT_r_, S. J. : On the Structure oI the Topside Ionosphere. J. Atmos. Sel., vol. 19,

May 1962.30. BAUER, S. J., and JACKSOr_, J. E.: Rocket Measureme_t o[ the Electron Density Dis-

tribution tn the Topside Ionosphere. J. Geophys. Res., col. 67, April 1962.31. BAUER, S. J. : On the Thickness of the Helium Ion Layer. NASA Tech. Note 1686 to be

published.32. V_'ILLMORE, A. P., BOYD, R. L. f., and BOWER, P. J. : Some Preliminary Results o! the

Plasma Probe Experiments on the Ariel Satellite. Proem. of the International Con-ference on the Ionosphere, July 1962, in press .

33. ULWrCK, J. C., and PFISTER, W. : Spatial and Temporal Variations o[ Electron Den-sity from an Orbiting Satellite. Third International Space Science Symposium, May1962, in press.

34 . BOWLES, K. L., CellS, G. R., and GREEN', J. L. : On the Absolute Intensity of IncoherentScatter Echoes from the Ionosphere. NBS Journal of Research, Section D, in press,1962.

35. BOUaDEAU, R. E., and BAUER, S. J. : Structure o/ the Upper Atmosphere Deduced [romCharged Particle Measurements on Rockets and the Explorer VIII Satellite. ThirdInternational Space Science Symposium, May 1962, in press.

36. HARltXS, I., and PRIESTER, W. : Theoretical Models for the Solar-Cycle Variation oI theUpper Atmosphere. NASA, Tech. Note D-1444, August 1962.

37. JACCHIA, L. G., and SLOWEY, J.: Accurate Drag Measurements [or Eight ArtificialSatellites; Atmospheric Densities and Temperatures. Smithsonian Institution A_trophysicai Observatory Report, No. 100, July 1962.

38. BOURDEAU, R. E., DONLEY, J. L., and IVHrPPLE, E. C. : The lo_tosphere Direct Measure-ments Satellite. NASA Tech. Note D-414, 1961.

39. BLAMOY, T, J. : La Temperatttre de l' Ionospl_ere. Proc. Second International SpaceSciences Symposium, April 1961.

40. BLA_fOr_T, J. : Private communicat ion.

19


26/52


27/52

ParticlesndFieldsesearchnSpaceBy George H. Ludwig

I)m GEOnC,EII. Iwnwlo has headed the Fields and Particles Branch Instru-"mentation Section of the NASA Goddard Space Flight Center since 1960.Instrumentation fo_' a nunTbev o/spacee_vrft, hwluding the receipt E.vpto_'ers Xand Xl[ was developed here. He is currentl 9 the pro)vet ,_eientLs.t for the900 pound Orbiting Geophysical Observatory to be launched in 1963. Dr.Ludwig received the B.A. and M.N. degree.s, with a major iz_Physics in 1,956and 1959, respectively, and the Ph. D. degree in Elect_'ieal Engineering hi1980, all/ram the State Uni_.er._it!/ o/ Iowa. A.s. a graduate research assistantto Dr. J. A. Van Allen, he developed me.st of the eorpw_('ular radiation inst_'u-mentation /o_" E,vplorers I, ll, H1, IV, T', VII, at_d project S-{6. Explorersl, ll[, IV, and T'// were sueee_s.s.fully hm_whed and led to tile discovery o/the hlyh b_ten._it_# trapped radiation. Dr. Ludu'/y is a member o/Phi BetaKappa, SOma Xi, American Geophy.s.ical Unlo_, I_stitute o/ Radio EngLavers, and A meriean Reelect Society.

INTRODUCTIONThroughout the universe physical processes

occur which result in the formation, ionizatioIt,and acceleration of matter. The motion ofthese charged particles results in the formationof magnetic and electric fields. The furtherinteraction of these moving particles with othermagnetic and electric fields results in their fur-ther acceleration. As a result, there exists awhole continuum of charged particles havingenergies from thermal to at least 10 _ electronvolts , and nuclear structures from that of hydro-gen (single protons) to that of heavier mate-rials (at least iron). And throughout regionsas large as whole galaxies, there exist complexmagnetic and electric fields. Until the comingof high altitude balloons and rockets, and theearth satellites and interplanetary probes, wewere restricted to the study of the particles andfields which reached the earth through itsmodifying atmosphere and strong magneticfield, and to the study of secondary processesoccurring far from the earth which produced

672058 0-63--4

radiations capable of reaching the earth. Twoexamples of such indirect observations are: (1)the study of the magnetic field fluctuations atthe surface of' the earth to deduce the existenceof large electric currents (ring currents) in thehigh atmosphere ; (2) the observations of radiosignals which presumably are produced bysynchrotron radiatiort of electrons contained ina Jovian high intensity radiation belt.The new instruments of the space age allow

us to conduct experiments wit]fin and outsidethe earth's modifying atmosphere and magneto-sphere to learn more about these interestingphenomena.This paper will summarize briefly our present

state of understanding of the energetic chargedparticle populations in space and the character-istics of the magnetic fields which interact withthem. For convenience, we shall begin with adiscussion of the galactic cosmic rays, which arethe largest scale phenomenon, and then proceedto solar cosmic rays, solar plasma, and theearth's geomagnetical]y trapped radiationwhich forms the Van Allen radiation belts.

24


28/52


GALACTIC COSMIC RAYSGalactic cosmic rays are those p.u'tieles which

are accelerated outside our solar system and ,ar-rive with kinetic energies greater than about 10Mev per nucleon. Cosmic rays having energiesas high as about 10TM electron volts have beendetected. It seems di/l_cult to imagine thatthese very high energy particles could be pro-duced wilhin our own galaxy, since their radiusof curvature in the galaxy's magnetic fiehl(which cannot be much stronger on the averagethan about 10.5 gauss) is of the order of thedimensions of the galaxy itself. Thus, it ispresently believed that they arrive from othergalaxies. Lower ener_ _ galactic cosmic rayscertainly must originate from within our owngalaxy. It is probable that at least some ofthem are produced initi_dly by localized erup-tions on the turbulent surfaces of stars, similarto those which produce the solar cosmic rayswhich we will discuss later. It is also probablethat supernovae may be an important source.These particles are then accelerated. Parker

(ref. 1) has shown that within the frameworkof the present understanding of plasma dynam-ics, all particle acceleration mechanisms occur-ring outside of the laboratory are reducible tothe Fermi mechanism (ref. 2, 3, 4) which isbased on random particle collision with mag-_etic inhomogeneities. Insofar as comparisonscan be made, the experimental results seem con-sistent with this Fermi type accelerationmechanism.The features of the "galactic cosmic rayswhich are most amenable to investigation are

the energy and charge spectra, and the modula-lion of their characteristics with time. Thestudy of the directional characteristics to as-

certain their sources is not very fruitful be-cause the directions of motion of the particles'Ire changed by the very complex magneticfields through which they pass, so that the fluxis very nearly isotropic.

Charge SpectrumThe composition (or charge spectrum) of the

primary cosmic radiation striking the top of theearth's atmosphere is reasonably well known atthe present time (ref. 5_ 6, 7). Briefly, 85%of the particles are hydrogen, 1'2% are helium,approximately 1% are in the carbon, nitrogen,and oxygen group, about 0.25% are in the lith-ium, beryllium and boron group, and 0.25%are neon and heavier. In this heavier group,nuclei of all charges up to and including iron(Z='26) have been identified. High energyelectrons constitute I to 3% of the total flux(ref. S, 9). Whether these electrons are ofgalactic or solar origin is not fully understoodat the present time.A detailed summary of the charge spectrumis given in table lO_-I. There are two featuresof this dmrge spectrum which should be men-tioned. First, the flux of lithium, beryllium,and boron is surprisingly high if we assume thatthese elements are extremely rare in the sourceregion, as the table of stellar abundance wouldhave us believe. These nuclei must be producedby fragmentation of heavier nuclei collidingwith interstellar hydrogen nuclei. This, then,gives a measure of the average distance tra-versed by the galactic cosmic radiation beforestriking the earth. If we use presently avail-able fragmentation parameters (ref. 7), thiscorresponds to a value of 4 gins/era _ or im-plies an average age of the particles of the order

TABLE 12-1.--Relative Abundances of Multiply Charged Nuclei

*Galactic Cosmic Rays ...................

tSun ...................................*Solar Cosmic Rays .....................

He

360

?1250

Li, Be, B

11


29/52

PARTICLES AND FIELDS RESEARCH IN SPACE

of 10 7 years. The other interesting feature istile large flux of carbon, nitrogen, and Z> I0.The observed vahles are 10 times greater thanthat expected from the known stellar abun-dauces. At the present time, this is not under-stood but can perhaps be explained in terms ofsupernovae origin (ref. 10).

Energy SpectrumIn tile high ener_o T region (E>3 Bev) it

has been observed that all charged componentshave enerKy spectra of the form

KzJ(E) =_,

where J(E) is the flux in particles/(m:-sec-sr)with kinetic energy >E, _ is a constant inde-pendent of Z, and K(z) is a function of Z.J(E)>0.5 Bev) is typically 1800 pai_cicles/(m2-sec-sr) at solar maximum.In the region 0.3-5 Bey it is observed that allcharge components appear to display similarspectral forms when expressed in terms of rig-idity (R _ p/Z) where p is the momentum andZ is the charge of the particle. The differentialenergy spectrum shown in figure 12-1 displays amaximum in the region of 400 Mev per nucleonand extends down to at least 90 Mev per nu-cleon. In the region 90 to 200 Mev per nucleonMeyer and Vogt have found large increases,which they believe are due to solar cosmic rays(ref. 11). Data from Explorer XII do not show

FIGURF 12-1.--Cosmic ray differential energy sl)e('trum(ref. 11).

this increase; further work will be necessary be-fore this disagreement can be explained.Modulation of the Galactic Cosmic RadiationIt appears reasonable to assume that the flux

of cosmic rays incident on the solar system isconstant. However, in the vicinity of the earthlarge modulations are observed which appear tobe controlled by solar activity. The two mostimportant types of modulation appear to be the11 year variation and the Forbush or storm-type decrease. It was first noted by Forbushthat the cosmic ray intensity varied inverselywith solar activity over an 11 year cycle (reL12). Also, Forbush first observed the rapidworld-wide decreases in cosmic ray intensitywhich are associated with some maguetic storms(ref. 12), generally following large type 3+solar flares. At present the best evidence indi-cates that the modulations are heliocentric andcontrolled by solar activity. The effects on cos-mic rays can be summarized as follows (ref. 14) :

(1) The proton and alpha particles (hydro-gen and helium nuclei) (R>I.0 BV) similarforms of differential rigidity spectrum at solarminimum, and maintain the _me relative formof rigidity spectrum throughout the solar cycle.(2) The total intensity is smaller by a factor

of 2 at solar maximum than at solar minimum,and the lower energy particles are modulatedmost. Iq:owever, the low energy component isnever completely removed.

(3) At low energiestheform of thespectrumchange appears to be the same for the 11 yearcycle as for the Forbush decreases which havebeen observed.

The modulation effectsare interpretedas be-ing due to large scale variations of the solarmaguetic field. Increases of the ma_etic fieldstrength near the earth result in the deflectionof the lower energ'y particles before they pene-trate this field to the position of the earth. Thisfield arises from two sources. A relativelysteady magnetic field is the result of the stream-ing of solar plasma from the sun, and is of theorder of 5 to 10 gammas at the earth's orbit,where one gamma equals 10 -5 gauss. Varia-tions in this field result in the 11 year cycle.The Forbusb, or storm decreases are caused by

23


30/52


FIGURE 12-2.--Sun-earth phenomena.

somewhat more localized fields generated as aresult of large solar flares.

SOLAR COSMIC RAYSTile surface of the sun boils in an active man-

her as the result of the continuous productionof enerp.o" inside the sun. At times the surfacebecomes unusually turbulent and ejects cloudsof charged particles and streams of radio, vis-ual, x-ray, and gamma ray radiation into thespace sm'rounding the sun (fig. 19-2). The_solar eruptions are known as fares and thecharged particles are known as solar cosmicray,q,When the flare is situated in the right positionon the sun's snrface, the clouds of particles areejected in su'h a (lire'lion that they reach theearth and interact with its atmosi)here. There

they produce magnetic storms, radio blackouts,auroral displays, and other phenomena.The streams of solar cosmic rays containing

a large high energ3" content can be detected bysea level ion chambers and neutron monitors(ref. 15, 16). The development of the rio-meter (ref. 17) in 1957 provided an instru-ment which is sensitive to the effects of smallfluxes (10 particles/(em2-sec-sr)) of incomingparticles in the low energy region (10 to 100Mev). There were at least a dozen events peryear during the last solar maximum which weredetectable by riometers (ref. 18).The most detailed information about these

solar cosmic rays has come from particle de-tectors flown on balloons, rockets, and space-craft (ref. 19, 20, .Ol). Enough data have lx,enobtained to give an indication of the sequence

24


31/52

PARTICLES AND FIELDS RESEARCH IN SPACE

FIGURE 12-3.--Idealized solar proton event configura-tion, showing the solar proton cloud, the drawn-outfield, and the effect on galactic cosmic rays.

of events occurring during these disturbances.YVhen a solar flare occurs, a tongue of chargedparticles erupts from the sun's surface at a speedof from 1000 to '2.000 kilometers per second (fig.12-3). It takes about a day for this cloud toreach the earth. The charged particles dragalong the lines of the solar magnetic field whichbecome frozen into the cloud and forced to movewith it according to Maxwell's laws. As theforce lines become distended in this manner theylose their strength. However, the field is stillstrong enough at the distance of the earth tocause a partial screening of the galactic cosmicrays resulting in the Forbush decreases dis-cussed earlier. Solar cosmic rays continue toarrive at the earth for several days after the in-itial encounter, presumably because of thestorage of particles in this magnetic field.The particle cloud advances outward from

the sun at such a high speed that it must be pro-ceeded by a shock wave. This shockfront maybe the reason for the sudden-commencementcharacteristic of many magnetic storms.Parker (ref. 22) believes that this shock, re-sulting in a linked and stren_hened ma_mticfield behind the advancing front, is the mostimportant part of the solar storm process. Theadvancing shock wave and particles result inthe geophysical phenomena mentioned earlier.One of the most striking features of these

solar cosmic ray events has been the variationfrom one event to another. In some events, the

flux of protons above 9,0 Mev has exceeded 103particles/(cm2-sec-sr) for more than a day (ref.20, '23, '24) and the total energy arriving at thetop of the atmosphere of the earth for the wholeevent has been 10' ergs/(cm*-sr), about the sameorder of magnitude as that for cosmic rays foryear. On the other hand, events which aremore than a hundred times smaller than this

have been seen (ref. 25, 9.6), and even smallerones probably occur frequently and are not de-tected. The energy spectra also shows markeddifferences from one event to another. For ex-ample, at comparable times in the November12, 1960, and the July 12, 1961, events (ref. 9.4,97) the integral fluxes above 10 Mev were nearlythe same, but above 100 Merv they differed bymore than a factor of 300. In some events, themaximum low energT intensity occurred asearly as two hours after the flare; whereas inothers, it occurred as late as thirty or fortyhours after the flare. In order to show someof the general characteristics of these events,the time variations of the integral fluxes above20, 100, and 500 Mev are shown for three largeevents (ref..28) in figure 19,-4.On the basis of the detailed study of these

three events, there seems to be good evidencethat the relative abundances of all elementshaving charges greater than one but less thanabout sixteen are nearly independent of energy.they are shown in table 19.-I, along with thenatural relative abundances existing in the sun,cleduced from spectroscopic evidence for thoseelements. It is seen that the two sets of num-bers are the same within uncertainties. Al-though the helium abundance in the sun cannotbe determined by spectroscopic means, the rel-ative abundance given in table 12-I for heliumin solar cosmic rays is within the rather widelimits set by theoretical models of the sun.The proton component has been observed to

have a very different ener_-W spectrum from theother components, probably because of its dif-ferent charge to mass ratio. This has the effectthat the relative abundance of the proton com-poneut is a function of velocity. For example,in one event the proton to helium ratio variedfrom about twenty at 40 Mev/nucleon to aboutthree hundred at 1'20 Mev/nucleon.It can be seen from table 1:2-I that the rela-

25


32/52


PROTONS

. 1 59.NOV.12, 191DE ,,_o _t- ..... :", ,,_ SUDDENDECREASE

1_,,." "'" ,.,'w_&,_'le HIGHENERGYFLUX "_=1o'_,_.... -,--_. -_-,--10_ SEPT. 28,1961! j '"I V-- _oo.."---. ;oo..- _ _......

f l _! " , * t t "-1, i i i J i""1 3 6 9 12 24 36 48 60 72 84 96 108 --:TIMEIN HOURSFROMBEGINNINGOF FLAR_

FI(}URE 12--4. Time variation of the integral flux above20, 100, and 500 Mev for three solar events (ref. 2_).

tive nuclear abundances in the galactic and so-lar cosmic rays are significantly different. Thedifference is primarily a result of the fact thatthe galactic cosmic rays have passed throughseveral grams per square centimeter of materialwhere fragmentation processes have increasedthe relative numbers of the heavier nuclei.

THE SOLAR PLASMAIn addition to the solar cosmic rays produced

by large solar flares, a large number of very lowenergy charged particles are continuously givenoff by the sun. This stream of particles is oftenreferred to as the solar wind. Gringauz, et al.(ref. "29), and Bridge, et al. (ref. 30) indicatethat on quiet days during which there are nomajor solar flares, the velocity of the solar windis of the order of 300 to 500 kilometers per sec-ond, with a density of about 10 protons per cu-bic centimeter at the orbit of the earth. It isbeieved that the particle density varies withsolar activity, and that the mag'netic field ac-

FZGVRE 12-5.--The geomagnetic fielil and the cavityboundary configuration when acted upon by the solarwind (Ness).

companying the particles modulates the arrivalof charged particles from outside the solar sys-tem as discussed earlier. Another effect of thesolar plasma is to distort the shape of the geo-magnetic cavity within which the earth is lo-cated. This boundary is a rather sharplydefined surface separating the region withinwhich the earth's magnetic field exerts primarycontrol over the particle motion (the magneto-sphere) from the interplanetary region. Thisregion would be in the shape of a sphere cen-tered at the earth in the .absence of externalparticle motion or fields. But the solar plasmacauses the magnetosphere to be compressed onthe sunward side of the earth and extended awayfrom the sun as shown in fig_lre 12-5. The flowof the high speed plasma also creates a shockwave ahead of this magnetosphere boundary.Thus, there are three regions w