Effects of Solar Activity on the Ionosphere and Radio Communications

Effects of Solar Activity on the Ionosphere andRadio Communications

H. W. WELLSt, MEMBER, I.R.E.

Summary-The relationships of solar activity on the ionosphereand radio communications may be roughly classified as follows:

(1) There are occasional solar flares or outbursts of ultravioletlight which instantaneously produce radio fade-outs of short dura-tion.

(2) Occasionally solar streams of particles sweep across theearth's orbit producing magnetic storms and auroral displays. Theassociated ionospheric disturbances may seriously affect radio com-munication for several days although the effects are more pro-nounced in polar regions.

(3) The general change of solar ionizing wave radiation in thecourse of the sunspot cycle governs the average intensity of ioniza-tion in the ionosphere. This trend is an important factor governingselection of operating frequencies for radio communication.

GENERALT HE SUN affects the ionosphere and the iono-

sphere controls radio communications. The sud-den radio fade-outs, the occasional complete

disruption of radio and wire circuits, the changes ofradio-communication conditions from day to night orwith season, and the trends from year to year-allof these may be traced back to the sun, directly orindirectly.The sudden fade-outs are very spectacular. One

moment, communications are normal and the nextinstant there will be no signals. There is an eruption onthe sun, and the fade-out occurs instantaneously. Thedisappearance of signals applies only to sky-wavecommunications since line-of-sight transmissions andground-wave coverage are relatively unaffected. Thefade-outs of this type seldom last more than an hourbut they have been responsible for many a good radioset being torn apart in futile efforts to locate the trouble.

However, the more severe interruptions to radiocommunications are associated with magnetic stormsand ionospheric disturbances. On such occasionssignals may be blanketed out or severely interruptedfor several days especially when the wave paths ap-proach polar regions. In addition to the effect on sky-wave communications, strong electrical currents arefrequently generated in the surface of the earth whichrender inoperative many telegraph and other wirecircuits. These radio storms are likewise traced backto the sun although the connections, as will be shownlater, are not so direct as in the case of the radio fade-out.

Characteristics of the ionosphere are normally re-corded by the radio reflection technique developed byBreit and Tuve1 of the Department of Terrestrial

* Decimal classification: R113.5. Original manuscript receivedby the Institute, August 5, 1942. Presented, Summer Convention,Cleveland, Ohio, June 30, 1942.

t Department of Terrestrial Magnetism, Carnegie Institutionof Washington, Washington, D. C.

1 G. Breit and M. A. Tuve, "A radio method of estimating theheight of the conducting layer," Nature, vol. 116, p. 357; September,5, 1925.

Magnetism, Carnegie Institution of Washington, in1925. A pulse of short length is sent out from the trans-mitter and the echo of this pulse is picked up at thereceiver. The time difference between the direct(ground) wave and the reflected signal (echo) measuresthe height of the region of the ionosphere which has re-turned the signal. Early recordings were on fixed wavefrequencies showing the manner with which the heightof the so-called "Kennelly-Heaviside layer" changedduring the day.

Electromagnetic theory shows that a wave of fre-quencyf, radiated vertically into a medium such as theionosphere, will penetrate until it encounters sufficientconcentration of electrons N to bend the signal aroundand return it to earth. If N is not adequate, the signalis lost in space. The simplified formula for refractionof the ordinary wave at vertical incidence is

N = 1.24f2 X 10-4 (1)

where N is the electron density in electrons per cubiccentimeter and f is the frequency of the reflected wavein megacycles per second.

It will be recalled that the existence of this electri-fied region surrounding the earth was not publicly ac-cepted until after Marconi's successful communicationby radio in 1901 over the curved surface of the earthfrom England to Newfoundland. It was reasoned in-dependently by Kennelly and Heaviside that radiowaves, like light, travel in straight lines, hence signalscould travel long distances beyond range of the groundwave only through reflection back to earth froman electrified or conducting region in the outer at-mosphere.

Technical literature reveals that as early as 1878-long before the time of Marconi's "wireless"-otherscientists had likewise suggested an ionosphere sur-rounding the earth. Balfour Stewart and ArthurSchuster, however, were studying the earth's magnet-ism. They observed systematic variations of theearth's magnetism which, they reasoned could be pro-duced by electrical currents circulating in regions ofour outer atmosphere.

After Marconi's historic experiment, radio com-munication over long distances became widely applied.Much practical use was made of this unseen and un-identified "Kennelly-Heaviside" layer. The trends tohigher frequencies in the 1920's revealed peculiar de-velopments involving skip distances and phenomenalranges on low power, even in the daytime.

Concurrently, researches by Breit and Tuve, whichwere related to studies of the earth's magnetism,

Proceedings of the I.R.E.April, 1943 147

148 Proceedings of the I.R.E. April

lT __ 1 1 1 1information which is usefully applied'Y PEW

Ae"rrrFY^^"r/N>sW¢t to studies of radio wave propagationas well as magnetic variations.

In general, the ionosphere is strati-X fied into at least three separate re-

gions during the daytime. The Eregion at heights of about 60 miles is

2 the lowest one regularly observed.The F1 region exists only in the day-

w _ IC l&rgroF<,<,~FrNr¢[5tGtAto_ time and is found at heights of abput"V |-X9UUga|J&.-UGtON140miles. The F2 region is the highestaG==-_-tsAcr UN rr and is found at heights from 200 to

KXOCYCLE,SP.R iONOMIEOCtECTRlw ' 400 miles above the earth. At nightCUBIC CENrAME7SWR

the F1 and F2 regions merge leavingFig. 1-Exploration of the ionosphere. only the E and the merged F regions.(A) Typical record of heights of ion densities obtained with automatic variable- A brief summary of exploration offrequency equipment at Kensington, Maryland, 3 P.M., May 15, 1936. (B) Distribution

of ions deduced from (A). (C) Density of ionization deduced from (B) showing paths of the ionosphere is given in Fig. 1. Awaves of various frequencies. Diagram (C) shows transmitter and receiver separated for typical daytime multifrequency re-simplication, although in actual work the radio transmitter and receiver are at samestation and the wave paths are vertical. cording is illustrated. In the figure,

frequencies up to 3500 kilocycles persucceeded in directly observing reflections from the second were reflected from the E region; frequenciesouter atmosphere. As mentioned before, the radio from 3500 to 5200 kilocycles per second were reflectedaecho' technique developed for this purpose has now from the F1 region; and frequencies from 5200 to 7800

-4

I7°05.OE-

29720r-

k 296607-a' .

'U -

296407-

p 0/ 0. IL /A 'A 'I

E LT Ir

N

A-REFLECrION OF RADIO WAVES FROM IONOSPHERE,//007- Ht/ANCAYO MAGNErIC OBSCRVArORY

X _ _ riT _ L /A yr ( BS-RECORD OF EARTN'S MAcNTr/C CONDIOTON, HuANCAYOw ~~~~~~~~~~~~~~~~MACNET/COBSERVATORY

'. /140y C-SUN IN RED HYDROGEN LIGHT, 14h421,, mT. wILsoN4~~~~~~~~~~~~ ~~~~~oBsERvAroRYh I j6h

Fig. 2-Changes in earth's magnetism and cessation of radio reflections from ionosphere accompanyingbright eruption in solar chromosphere, August 28, 1937 (75 degrees west meridian times).

been widely adopted by investigators in these fields.In England, Appleton and Barnett2 also succeeded inidentifying the ionosphere at about the same timewith somewhat different technique. Improved meansof exploration of the ionosphere now provide detailed

'E. V. Appleton and M. A. F. Barnett, 'On some direct evi-dence for downward atmospheric reflection of electric rays,' Proc.Roy. Soc., ser. A, vol. 109, pp. 621-641; December, 1925.

kilocycles per second were returned from the F2 region.Above 7800 kilocycles per second the vertical-incidencesignals passed through into space. For oblique-incidencetransmissions where transmitter and receiver are sepa-rated by real distances, signals on frequencies above7800 kilocycles per second would be reflected from theionosphere. For example, the ionospheric conditions

_ _

'C44f_

Wells: Solar Effect on Radio

illustrated would support communication overa distance of 1000 miles on a wave frequencyof 14 megacycles per second.

Long-distance communications may be pos-sible by reflections from any of the ionosphericlayers, although the F or F2 regions are normallymost effective in supporting high-frequency sig-nals. Existing methods of translating ionosphericmeasurements made at vertical incidence intoterms of communication frequencies over vari-ous distances have been adequately described.3Any communications outside the ground rangeand not in the line of sight depend upon propaga-tion through the ionosphere. Consequentlysignificant changes in ionospheric conditions Fare reflected as changes in communicationconditions. These changes may be for better orfor worse. It happens that most of the sudden chaniof the ionosphere are associated with poorer commu:cations.

SHORT-PERIOD RADIO DISTURBANCESThe sudden but very temporary ionospheric d

turbance generally known as the radio fade-outvery spectacular in its development and effect upcommunications, as mentioned above.4 There isoutburst or eruption on the surface of the sun whiis immediately associated with a fade-out of radsignals on the daylight side of the earth only. Coimunications on the night side of the earth are entireunaffected. These temporary disturbances are fquently associated with a small pulse or change in tearth's magnetic condition. One very unusual examsof this relationship between the sun, the ionospheand the earth's magnetism was recorded at tHuancayo Magnetic Observatory on August 28, 19:The bright eruption illustrated in Fig. 2 occurred ncthe eastern limb of the sun, and was photographedabout maximum brilliance at 14 hours 42 minutes. Trecord of radio reflections from the ionosphere onfixed frequency showed development of a sudden dturbance or fade-out at 14 hours 25 minutes EasteStandard Time which was followed by complete asorption of all signals until about 15 hours 00 minutThe recording of horizontal intensity of the eart]magnetic field showed a quiet and normal downwatrend until the fade-out at 14 hours 25 minuteswhich time the magnetic intensity was suddenly icreased. The maximum deviation of magnetic",tensity from normal trend was recorded at about ttime the solar eruption was photographed which lilwise corresponded to the mid-point of the fade-oiAfter the fade-out the magnetic pulse disappeared athe normal trend of magnetic intensity was continu4

a N. Smith, "The relation of radio sky-wave transmissionionosphere measurements," PROC. I.R.E., vol. 27, pp. 332-3May, 1939.

4 J. H. Dellinger, "A new cosmic phenomenon,' Science, vol.p. 351; October 11, 1935.

/rfla2lI-NORMALWs 3-RECOVERYNIR i/930O ,#As

'(6I0

i2_.

,4h27.^j0S .-/DE-OtUI

%k600-

O-1k2

4-NORMAL. ,h osid S

SCALE OF F;REQENCY AN tC/StC

ig. 3-Radio fade-out recorded at Kensington, Maryland, February 26,1941, illustrating characteristic sudden development and gradualrecovery, 75 degrees west meridian time (E.S.T.) is indicated.

res Sudden ionospheric disturbances of this type occurni- in various degrees of intensity; some weak-some

strong.6 The highest intensities are usually observedwhen the radiation from the solar eruptions is mostdirect. Analyses of a number of these effects show that

is. about 60 per cent of the fade-outs endure less than 15is minutes, while an occurrence of more than an hour ison very unusual.an Without special facilities, it may be difficult to recog-ch nize a radio fade-out until after the disturbance haslio passed. It is characterized by an abrupt beginning andm- a somewhat more gradual recovery. The associatedely solar flare may only be identified by special observingre- apparatus. It cannot be seen with the naked eye. Thishe effect is contrasted with the disturbance associated)le with 'radio storms" to be discussed later which de-re, velop gradually and last much longer. Sudden fade-he outs occur at random intervals, sometimes several in37. the same day and occasionally none at all in a month.ear In general, their frequency of occurrence follows theat sunspot cycle. When sunspot numbers are high, fade-he outs are more numerous, and when sunspot numbersa are low, fade-outs are less frequent.

is- The generally accepted explanation of these phe-irn nomena which completely absorb radio waves is quiteLb- simple. Ultraviolet light of the solar flare penetrateses. into the lower part of the ionosphere. It is not ab-h's sorbed extensively in passing down through the normalxrd F2, F1, and E regions. However, at some level below theat normal E region the intense radiation is absorbed andin- produces a high degree of ionization. This ionizationin- is in a region of relatively high molecular density,he which results in complete absorption and the dissipa-ce- tion of energy of high-frequency radio waves. It is in-it. teresting to note that fade-outs occur and disappearnd without producing any noticeable effect on the othered. ionospheric regions. In spite of the intense ionizationto just below the E region, there is no appreciable change

§47;6 L. V. Berkner and H. W. Wells, 'Study of radio fade-outs,"

82, Terr. Mag., vol. 42, pp. 183-194; June, 1937; and pp. 301-309;September, 1937.

1943 149

1,hZ.3S

150 Proceedings of the I.R.E. April

/6 2075 wcsr MERID/AN MEAN HOURS

400 i _____F2 -REGIONMINIMUM V/RrUTAL HEIGHrS Ax -. - o o-o °

I $ ,\ - ,4 m B s ~~~~.0' 10. t A,

-5LAX, 0FRGO 4N0B-; F-REGION

-200 Eo -v

_ 0~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 1-_. Ap-_O-.o>O>off+- _ __< o+ o6/,s f i ' - ° 4

_~--"o..o'.--o~ ~ 0. -0. E-REGION

F2 CRIrICAL FREOUENCY

| )-cRIrICAL FREOUENCIES l ,0oa0 o

a2-A

-80-0.~I -8 o -1--- t------ oo~~~~ - 00

NO REFLECrIONS -d(PENcErRATroN) , -

Z' ~~~~F2 -REGIONA | ao (FOR MINIMUM HEIGHTS SEE ® ABOVE)

0~t CRIrTCAL FREOUENCY

% 5-REGION\. F-REGION- E CRIrICAL FREQUENCY |.

-4 -/% A

B.

E-REGIONV

- ~ WHICH REFLECTrIoNs oBsERvED -=.- --

NO REFLECTrIONS(ABsorPTION)

Fig. 4-The ionosphere showing radio fade-out July 31, 1937, determined from automatic multifrequency registrations, KensingtonMaryland, (39001' North, 77°05' West).

---o---o---o- observed values for o-wave component; weekly mean of values for o-wave component (July 26 to 31, 1937);lowest frequency on which reflections observed; 1////1// fade-out at Huancayo Magnetic Observatory (120 South, 750 West)

determined from automatic fixed-frequency registration (4.8 megacycles per second).

in the degree of ionization in the normal E, F1, or F2layers when it is possible to observe them again follow-ing the radio fade-out. Automatic multifrequency re-cordings of ionospheric characteristics during radiofade-outs have definitely established this fact which islikewise illustrated in Fig. 4. Similarly heights of theseveral ionospheric regions appear to remain relativelyunchanged.Normal multifrequency recordings of ionospheric

characteristics include measurement of the lowest fre-quency from which reflections at vertical incidence arerecorded. In Fig. 4 the lower limit was about 1.0 mega-cycle per second. Signals below 1.0 megacyle per secondin this case were highly attenuated since no reflectionswere recorded. The radio fade-out, therefore, extends

this lowest frequency of reflections up to very highvalues. When this absorption limit is boosted up highenough it cuts off all reflections from the higher regions.Following the disturbance the recovery to normal issomewhat gradual as illustrated in the figure.

There are certain indications and reports which ap-pear to be substantiated by actual observation thatcommunications on low radio frequencies, say in therange 100 to 200 kilocycles per second or less, may beimproved rather than absorbed during the suddenfade-out.6 The process by which these waves are re-turned from the ionosphere is more nearly equivalent to

6 J. E. Best, J. A. Ratcliffe, and M. V. Wilkes, "Experimentalinvestigations of very long waves reflected from the ionosphere,"Proc. Roy. Soc., ser. A, vol. 156, pp. 614-633; September, 1936.

Fig. 5-Geomagnetic activity (u) and relative sunspot numbers (R), annual means, 1835-1941.


that of metallic reflection than to a refractive or bend-ing process. The increased ionization and the improvedconductivity in the lower ionosphere resulting from thesolar flare could reasonably explain an improvement ofsignals on the low frequencies.

LONG-PERIOD RADIO DISTURBANCES

On other occasions, the sun appears to shoot outstreams of electrified particles which reach the earthabout one to four days later, causing mag-negic storms, auroral displays, ionosphericdisturbances, and interruptions of radio com-munications. It will be recalled that the"fade-out" illustrates a direct or instantaneousrelationship between the sun and the earth.Now we have the delayed effect which may bevery severe and which can disturb communi-cations over the entire world. The intensestorms may occur only once or twice a yearalthough moderate disturbances are morefrequent. In general these disturbances followthe 11-year sunspot cycle with magnetic andradio disturbances of all types more fre-quent when the sunspot numbers are high.The comparison of geomagnetic activity withsunspot numbers since 1835 (Fig. 5) empha-sizes, however, that there is not a direct, one-to-one relationship between sunspot numbersand geomagnetic activity. Occasions of highmagnetic activity and low sunspot numbers,or of low magnetic activity and high sunspotnumbers, are frequently recorded. There have Photo-even been occurrences of magnetic storms graphwhen no sunspots were seen. The general (A)trends, however, are very definitely estab- (B)lished. (C)The intense disturbance of September 18, (D)

1941, is still vividly recalled by many becauseof the brilliant aurora and the associatedeffect upon communications. This storm resulted froma very large and active sunspot group. The develop-ment of this sunspot group as photographed on eightsuccessive days is illustrated in Fig. 6. It will be notedthat on September 12 the group of sunspots was firstseen on the eastern limb. The normal rotation of thesun which amounts to one revolution in about 27 daysplaced the group on the sun's central meridian, point-ing at the earth on September 16. The storm, however,did not develop until September 18, two days after theactive solar region passed the central meridian. Pre-sumably' the solar particles were shot out at the earthon September 16, but did not reach us until two dayslater.The North Atlantic circuit disturbance ratings pre-

pared by the Radio Corporation of America, show thatcommunications over the North Atlantic were affectedfor the best part of three days, namely, September 18,19, and 20. At the Cheltenham Magnetic Observatory

of the United States Coast and Geodetic Survey, themagnetic ranges were the greatest in the history of theObservatory. Auroral displays directly overhead orsouth of the zenith were seen throughout the UnitedStates, and the aurora was reported from points inequatorial regions.The extent to which radio circuits are interrupted

during periods of magnetic activity depends upon thereaction in the ionosphere. Some of the general rela-

Courtesy of Uniled States Naval Obsenatory,Washington, D. C.

Fig. Photographs of sunspots associatedwith magnetic storm and auroral dis-play of September 18, 1941.

Date 750 West Photo- Date 75° WestSeptember Meridian Time graph September Meridian Time

h m s h m s12 11 31 20 (E) 16 10 53 2013 10 40 00 (F) 17 13 13 1514 11 38 36 (G) 18 15 12 3015 12 40 00 (H) 19 11 24 30

(I) Enlarged print of sunspot group, September 17, 1941.

tionships between magnetic activity and radio com-munications, especially over the North Atlantic cir-cuits, have been very adequately treated by Hallborg.7Such investigations have established the importantevidence that circuits over various parts of the worldare reacted upon differently by magnetic disturbances.Under the effect of such disturbances the ionosphereexhibits unusual absorption and turbulence. Either ofthese effects can interrupt normal communications,and a combination of the two is doubly severe.

Absorption during magnetic storms may be equallyas intense as that which causes the sudden radio fade-out. Occasions of complete disappearance of signalsare recorded. There is increasing evidence to indicatethat some, if not all, of the absorption during a mag-netic storm occurs in the same general part of theionosphere where the radio fade-out is produced.

7 H. E. Hallborg, 'Short-wave radio transmission and geo-magnetism," RCA Rev., vol. 5, pp. 395-408; April, 1941.

1943 151

Proceedings of the I.R.E.

Ima

z-e

ISS

l

*SVC-CSOOUtSCtr /MC,Ic

VI *c

1-

L

ITwAS'e-f*Oe -A Me/ .C

AFig. 7

I

. I `St IS . 7 Zt -eq M/e w^- -raQrstg/w.. rA"£C A/..w

B

A-Ionospheric records illustrating changes during magneticstorm of March 24, 1940, and showing disappearance of F, region,Huancayo Magnetic Observatory (12° S, 75 W).

Probably the same general explanation holds for bothtypes of absorption; that is, a high degree of ionizationis temporarily produced in a region of relatively highmolecular density. However, the ionizing mechanismsare distinctive. During magnetic disturbances the ab-sorbing ionization frequently seems to be a downwardextension of sporadic E-region ionization. The transi-tion from sporadic E-region ionization to the absorp-tion effect apparently is very critical and involves an

Courtesy Mount Wilson Observatory. Carnegie Institution of Washington

Fig. 8-Streamers over an active sunspot extending into space manythousands of miles.

B-Ionospheric records illustrating changes during magneticstorm of March 24, 1940, and showing production of new F2 region,Huancayo Magnetic Observatory (120 S, 750 W).

effective lowering of height which may not amount tomore than a few kilometers.8Turbulence in the ionosphere during magnetic

storms is evidenced as abnormally high or low electronconcentrations. Systematic analyses of such variationhave been made.9 Probably the most outstanding effectduring severe disturbances is the apparent "blowingup" of the outer ionosphere as illustrated in Fig. 7. TheF2 region gradually disintegrates and often completelydisappears. Echoes are returned from great heights andthe critical frequencies are low indicating greatly re-

duced electron concentration. Under these circum-stances, normal high-frequency sky-wave communica-tion by means of the F2 layer again is disrupted sincethe electron concentration is not sufficiently dense tobend the waves back to earth. The lower ionosphericregions, however, are only slightly altered during suchdisturbances and propagation occasionally takes placein this manner. Wave propagation by the E or F,regions, however, is relatively limited as to time of day,frequencies usable, and range of coverage.

As additional evidence of these streams which are

shot out from the sun we have actual photographswhich sometimes show great extensions or streamers

' P. 0. Pederson, "The Propagation of Radio Waves Along theSurface of the Earth and in the Atmosphere," Danmarks Natur-videnskabelige, Samfund, Copenhagen, Denmark, 1927.

9 L. V. Berkner and S. L. Seaton, 'Ionospheric changes associ-ated with the magnetic storm of March 24, 1940," Terr. Mag., vol.45, pp. 393-418; December, 1940.

152 April,t0<

R00_-q- nommomw


in particular directions as illustrated in Fig. 8. Thesestreamers occasionally move with high velocity. Someappear to travel right away from the sun while othersfall back to its surface.

Particles ejected from the sun with a speed of about1600 kilometers per second would take about 26 hoursto travel to the earth's orbit. Evidence has been col-lected indicating this value to be reasonable for an

current densities over the polar cap with maximum cur-rent flowing in a belt about 23 degrees from the mag-netic axis pole. These theoretical zones of greatest currentintensity nearly coincide with the observed zones of maxi-mum auroral activity which have been carefully estab-lished by international co-operation. Outside the auroralzones the theoretical current densities fall off veryrapidly.

Base Map, Courtesy, U. S. Hydrographic OfficeFig. 9-Relationship of great-circle paths from Washington, D. C., with auroral belts.

average figure although the transit interval variesfrom one to three or four days.

These streams probably are electrically neutral; thatis, they contain equal quantities of positive and nega-tive charges. When they approach within the effect ofthe earth's magnetic field the particles are deflectedinto the polar regions producing magnetic disturbancesand auroral displays.'0 It has been shown that thegeneral form of magnetic disturbances could be ac-counted for by a current system in the earth's outeratmosphere. This theoretical current system has high

10 S. Chapman and J. Bartels, "Geomagnetism," Oxford Uni-versity Press, Oxford, England, 1940 ch. 9.

It is common knowledge that radio-communicationcircuits which traverse the polar regions are erratic inoperation and subject to frequent interruptions. Cir-cuits which only approach the polar regions are subjectto somewhat less frequent interruptions. The degreeof magnetic activity appears to be a factor controllingthe disturbances. The North Atlantic circuit from NewYork to Europe is an excellent example since thisradio circuit approaches but does not cross the polarregions. Studies by Hallborg7 and others of circuitperformance over various paths and magnetic activityhave indicated unquestionable relationships. Theyhave shown that normal circuit performance is

1943 153


associated with quiet magnetic conditions, while erraticor poor circuit performance obtains during periods ofmagnetic disturbance.

It has been mentioned that the normal auroral zonesare roughly concentric around the geomagnetic poles.(These are occasionally known as the magnetic axispoles and represent the location the poles would as-sume if the earth's magnetic field were uniformly dis-

central point to any other point are straight lines. Forexample, a circuit from Washington to London passesclose to the edge of the auroral zone as illustrated inFig. 9. For magnetically quiet periods this circuitwould be expected to operate normally. However,magnetic and auroral disturbances have the effect ofexpanding the zone radially outward from the geo-magnetic poles. Results of magnetic observations in

l-~~~~~~o"Base Map, Courtesy U. S. Hydrographic Office

Fig. 10-Relationship of great-circle paths from San Francisco with auroral belts.

tributed.) Furthermore, the auroral activity is an in-dication of abnormal ionization at levels whichinclude the ionospheric regions. The high currentdensities required to explain magnetic observationslikewise indicate a high conductivity in the outer at-mosphere which may be attained through abnormalionization. One effect of high ionization in the lowerionosphere has already been discussed as the probablecause of the radio fade-out. It seems reasonable to as-sume that radio-absorbing regions which limit com-munications over polar areas are very closely identi-fied with the zones of high magnetic and auroralactivity. The relationship between auroral zones andcommunication paths is illustrated by great-circlemaps with auroral zones added. Wave paths from the

polar regions show that this zone appears to broadenand move toward lower latitudes during periods ofintense disturbance so that a station which is ordi-narily on the equatorial side of the zone may, during amagnetic storm, be under, or on the polar side of thezone.10 In this case a very small expansion would sweepthe absorbing regions across the North Atlantic circuitpath and interfere with operation of the circuit.

Other circuits from Washington (or the east coast)to the Far East, including China, Philippines, DutchIndies, etc., cut across the absorbing zones. Com-mercial as well as amateur experience has establishedthe generally poor nature of communications over thesepaths. It is also seen that circuits from Washington (orthe east coast) to Africa, South America, or Australia

April154


are free from disturbances related to the auroral zonesexcept possibly during severe magnetic storms.

It is interesting to compare radio circuits on great-circle maps based on different points such as Washing-ton and San Francisco in the light of our experiencewith communications from these points. As mentionedabove, the circuit path, Washington to Manila, or to

'm

i

rex

to 70 degrees. Intense magnetic disturbances howeverwill extend these zones out to geomagnetic latitudes 40to 50 degrees. When this happens radio circuits inter-cepted by this expanded zone of absorption probablywould experience unsatisfactory operation. Radio cir-cuits in the lower latitudes are not so greatly affectedduring magnetic storms although unusual conditions

l; i ,It . IV ,,IV Isasi 8 l r' 6i , . . ' ii ' X ' ; i;1. . dsii i " 4i, . V. k i 'A'lFNCWWS C/SE

Fig. 11-Ionospheric records, Kensington Experimental Station, during sporadic intense ionization of E region in late afternoon,showing development and subsequent disappearance of the effect.

the Far East, traverses a large portion of the auroralzone and is generally unreliable. Now let us examinethe circuit, San Francisco to Manila, or to the FarEast, as shown in Fig. 10. We see that the wave pathfalls well outside the normal auroral zone. The de-pendability of this circuit has been thoroughly estab-lished. A relative degree of freedom of interruptionfrom minor magnetic and ionospheric disturbances isassured by the distance between the absorbing zoneand the wave path.

For example, a magnetic disturbance which spreadsout the absorbing zone just sufficient to interfere witha Washington-to-London circuit probably would notaffect operation of a San Francisco-to-Manila circuit(if we had one) because of the greater separation of thelatter from the auroral zone. Undoubtedly the com-plete interpretation of circuit-disturbances is notnearly so simple as suggested above.We have seen that the zones of maximum auroral

and magnetic activity are roughly concentric with thegeomagnetic poles. Furthermore, these zones spreadout radially from the geomagnetic poles during mag-netic disturbances which seem to develop as a result ofbombardment of the polar regions by particles fromthe sun. During slight disturbances the expansion issmall, but during severe magnetic storms the zone mayextend 20 to 30 degrees beyond normal. The extent ofthe auroral zone is therefore measured by distancesfrom the geomagnetic poles. Another manner of in-dicating zones of equal distance from the geomagneticpoles is by a system of geomagnetic co-ordination oflatitude and longitude. The zones of maximum auroralactivity normally include the geomagnetic latitudes 65

of absorption and turbulence are experienced. In mostcases communication may be. maintained although themost satisfactory operating frequencies may be aboveor below those normally used depending upon the re-action in the ionosphere. Ionospheric recordings nearthe magnetic equator at Huancayo, Peru, show thatthe F2 region of the ionosphere frequently has a lowerconcentration of electrons during certain phases of amagnetic disturbance. This characteristic is also il-lustrated in Fig. 7. Occasions, however, of greater thannormal electron concentration have also been observed,especially during the preliminary phases of magneticdisturbances.

80

Fig. 12-Maximum usable frequencies over distance of 1000 milesduring sporadic E recorded in Fig. 11.

OTHER SOLAR EFFECTS ON IONOSPHERESolar and magnetic disturbances also seem to be re-

lated to the ionospheric effect known as sporadic

1943 155


E-region ionization. This is a condition which has pre-viously been mentioned as an indication of intenseionization. The stages of development are recorded inFig. 11. As normally observed, the sporadic E effect isrecorded at ionospheric heights slightly higher than the

the higher latitudes.'1 These characteristics of distribu-tion resemble auroral occurrences. It is interesting tospeculate that subsequent analyses may relate sporadicE as well as aurora more closely to solar corpuscularradiation.

T -I - , ---- -- r- , IA -- t . - , I--.o /5 /O ~~~ ~ ~~~~~20'25 30/.5_DECEMBER 25 JANUARY NUMBER OF DAYS AFTER IN/TiA DATE INDICATED Ar LEFrw /940_ /94/-

_-CJATNUAER 25

-/5

/5FEBRUAI?Y 2 [

/5 /w-At!CH 20

.I-APML 6S

MA 13

.0

-JUAE 9

16i_UGU5 7 2

|AUGUyST29ltos

*SEPRfEMfEP? 2,

IL1.sEL 2 =2~~~I/;NVEME /tfH la-

5 DECEMBER JS 1942

-0.'sLI5v]

Fig. 13-American character figure CA, Greenwich half days in 27-day sequences,December 29, 1940, to January 18, 1942.

normal E region. This frequently blankets out or masksoff reflections from the higher ionospheric regions. Anal-yses of the occurrences of this effect from the limitedamount of world-wide data available have shown thatsporadic E is very seldom recorded near the geomag-netic equator, but that it occurs more frequently at

The sporadic E condition is very effective for com-munication purposes while it lasts because of the in-tense ionization and the low height. Occasions have

11 L. V. Berkner and H. W. Wells, "Abnormal ionization of theE region of the ionosphere," Terr. Mag., vol. 42, pp. 73-76; March,1937.

t56 A pril

Wells: Sciar Effect on Radio

frequently been recorded where radio communicationon wave frequencies up to 60 and 70 megacycles persecond could be supported for communication overdistances of 1000 miles or more. The graph of Fig. 12shows variation of maximum usable frequencies overa distance of 1000 miles during the sporadic E conditionrecorded in Fig. 11. In this case, a peak of 70 megacylesper second was reached at 17 hours 30 minutes. Thismaximum did not persist more than about 15 minutes,but wave frequencies exceeding 50 megacycles persecond were supported for more than an hour on thisoccasion. It should be added that communication overdistances of 1000 to 1500 miles represents a theoreticallimit for single-hop transmission from the E region.Communication over greater distances by such "freak"occurrences is much less probablesince reflection from sporadic E "'s 'effects at two widely separated pointsin the ionosphere would be involved.There is, of course, the possibility <which should not be entirely over-looked that the so-called sporadic E SUNSPionization is occasionally more gen- 42040Meral than at present believed. It is un-doubtedly this same effect whichmakes possible occasional long-dis-tance reception of frequency modu- slation and other stations on frequen- Xcies between 40 and 80 megacyclesper second.One of the factors which again re- - - r

lates the sun with the earth's mag-netism and the ionosphere is theoccasional tendency for certain types Fig. 14-Comparisof magnetic and ionospheric disturb- density of F2 regances to occur at intervals of 27 days.The sun completes one rotation in 27 days and it isreasonable to assume that active solar areas may becontinuously erupting streams of particles which sweepacross the earth once in each solar rotation. This 27-day recurrence tendency is illustrated by Fig. 13 whichplots magnetic activity in rows of 27 days each. Ina figure of this type, January 1, for example, is directlyover January 28, and the 27-day interval representsone complete rotation of the sun. Occasions have beennoted where moderate magnetic activity has recurredfor five or six consecutive solar rotations. However,studies of characteristics of the intense magneticstorms reveal that they seldom recur at all.12 Duringperiods of high sunspot activity the recurrence tend-encies are frequently difficult to determine since thevariations are masked by a large amount of randomactivity. However, during periods of low sunspotnumbers these recurrence tendencies are more readily

12 J. A. Fleming, Eleventh Arthur Lecture of Smithsonian Insti-tution: "The Sun and the Earth's Magnetic Field," presented Febru-ary 26, 1942. (In press.)

identified because of the absence of other activitywhich would mask the effect.The general relationship between sunspot numbers

and magnetic storms has already been discussed. Radioobservations of ionospheric characteristics for aboutone sunspot cycle of 11 years are now available. Thereis a very close correlation between sunspot numbersand average electron density in the ionospheric regionswhich is revealed by Fig. 14. The curve of electrondensity at the Huancayo Magnetic Observatory (Peru)for the F2 region measured at noon shows electrondensities in 1938 near sunspot maximum to be morethan double the densities recorded in 1933 near sun-spot minimum.12 This marked variation has an im-portant influence upon the selection and application

ion of annual average sunspot number with annual average electrongion measured at noon at Huancayo Magnetic Observatory, Peru.

of communication frequencies for established com-munication services. At the present time we are rapidlyapproaching another period of minimum solar activityand it is to be expected that communication circuitsfor the next few years will achieve generally satisfac-tory performance using the same ranges of frequencieswhich were found suitable about 1931. The time whenthe low of the sunspot cycle will be reached and theupward trend will start cannot be accurately deter-mined because of frequent irregularities in the lengthof the sunspot cycle which may range from 9 to 13 years.

In addition to the abnormal effects of solar disturb-ances on the ionosphere, there are other normal solareffects upon communications. The change from day tonight conditions is primarily a solar effect. Similarly thechange of ionospheric and communication characteris-tics with the seasons is primarly a solar effect. Otherobservations such as those made during solar eclipsesadd to our knowledge of solar and ionospheric relation-ships. Such matters, however, are beyond the scope ofthe present paper.

1943 157

Documents

Effects of Solar Activity on the Ionosphere and Radio Communications