UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Solar-terrestrial relations Hartz, Theodore Robert 1957

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1957_A1 H2 S6.pdf [ 7.39MB ]
JSON: 831-1.0085921.json
JSON-LD: 831-1.0085921-ld.json
RDF/XML (Pretty): 831-1.0085921-rdf.xml
RDF/JSON: 831-1.0085921-rdf.json
Turtle: 831-1.0085921-turtle.txt
N-Triples: 831-1.0085921-rdf-ntriples.txt
Original Record: 831-1.0085921-source.json
Full Text

Full Text

Faculty of Graduate Studies PROGRAMME OF THE F I N A L O R A L E X A M I N A T I O N ' FOR THE DEGREE OF D O C T O R O F P H I L O S O P H Y of T. R. H A R T Z B.A., University of Saskatchewan, 1945 B.Ed., University of Saskatchewan, 1946 M.A. , University of Saskatchewan, 1948 W E D N E S D A Y , SEPTEMBER 25, 1957, at 3:00 p.m. I N R O O M 300, PHYSICS B U I L D I N G Room 187A C O M M I T T E E I N C H A R G E D E A N G. M . S H R U M , Chairman A. M . C R O O K E R H . L. S T E I N R. E. BURGESS E. LEIMANIS K . C M A N N E. V . B O H N R. W. STEWART W . O P E C H O W S K I External Examiner: D. M . Hunten University of Saskatchewan SOLAR-TERRESTRIAL RELATIONS ABSTRACT The inter-relation of phenomena on the sun and in the earth's iono-sphere has been studied in order to establish a causal dependence which could result from travelling solar particles. In addition to examining a variety of data from a number of sources, an extensive R.F. noise record-ing program was undertaken to provide information on those regions of the solar and terrestrial atmospheres from which optical data were not available. The occasions on which there was an influx of such particles into the earth's ionosphere were established from magnetic data, from optical observations of the Polar Aurora, and from H.F. and V.H.F . radio observations close to the Auroral Zone. On the basis of these data the ionospheric disturbances for a 12-month period were related to prior distinctive solar events that were considered capable of producing the high energy particles. The selection was made from the optical, spectroscopic and radio data available. It was found that solar flares, eruptive prominences, and disappearing filaments were the most frequent sources of earth-reaching particles, and that large sunspots contributed only occasionally to ionospheric disturbances. Moreover, the radio noise data permitted a fairly reliable estimate of the particle veloci-ties in the solar corona, which could be used to predict the probability of subsequent terrestrial effects of the ejected particles. A correlation co-efficient of +0.65 was found between probabilities predicted from the solar information and observed ionospheric disturbances which occurred two or three days later. The observational evidence on the solar noise bursts and on ionospheric storms could only be explained on the basis of a considerable distribution of velocities for the particles ejected from the sun. It was shown that a Maxwellian distribution of corpuscular velocities was a probable one. Moreover, the observations indicated that particles are frequently emitted from the sun with a distribution of velocities, but that only in the case of very large phenomena are there sufficient particles with the high energies necessary to overcome the sun's gravitational field and reach the earth. PUBLICATIONS A S Q U A R E - W A V E M O D U L A T I O N M E T H O D FOR M I C R O W A V E SPECTRA T. R. Hartz and A. van der Ziel, Physical Review, 78, 473, 1950. A SOLAR NOISE O U T B U R S T O F J A N U A R Y 15, 1955 T. R. Hartz, Nature, 175, 908, 1955. T H U N D E R S T O R M SIGNALS A T V E R Y - H I G H F R E Q U E N C Y A N D U L T R A - H I G H F R E Q U E N C Y D. R. Hay and T. R. Hartz, Nature, 175, 949, 1955. RADIO STAR SCINTILLATIONS A N D T H E IONOSPHERE T. R. Hartz, Canadian Journal of Physics, 33, 476, 1955. V . H . F . A U R O R A L NOISE T. R. Hartz, G. C. Reid, and E. L. Vogan, Canadian Journal of Physics 34, 728, 1956. G R A D U A T E STUDIES FIELD O F S T U D Y : PHYSICS Electromagnetic Theory '. G. L. Pickard Electronics A. van der Ziel Special Relativity Theory W. Opechowski Wave Mechanics G. M . Volkoff Advanced Electronics A . van der Ziel Advanced Spectroscopy A. M . Crooker O T H E R STUDIES Theory and Applications of DifferentialjEquations W. H. Gage Topics in Applied Mathematics E. Leimanis Faculty of Graduate Studies .PROGRAMME OF THE F I N A L O R A L E X A M I N A T I O N FOR THE DEGREE OF D O C T O R O F P H I L O S O P H Y of T. R. H A R T Z B.A., University of Saskatchewan, 1945 B.Ed., University of Saskatchewan, 1946 M . A . , University of Saskatchewan, 1948 W E D N E S D A Y , SEPTEMBER 25, 1957, at 3:00 p.m. I N R O O M 300, PHYSICS B U I L D I N G Room 187A C O M M I T T E E I N C H A R G E D E A N G . M . S H R U M , Chairman A. M . C R O O K E R H . L. S T E I N R. E. BURGESS E. LEIMANIS K. C. M A N N E. V. B O H N R. W. STEWART W. O P E C H O W S K I External Examiner: D . M . Hunten University of Saskatchewan SOLAR-TERRESTRIAL RELATIONS ABSTRACT The inter-relation of phenomena on the sun and in the earth's iono-sphere has been studied in" order to establish a causal dependence which could result from travelling solar particles. In addition to examining a variety of data from a number of sources, an extensive R.F. noise record-ing program was undertaken to provide information on those regions of the solar and terrestrial atmospheres from which optical data were not available. The occasions on which there was an influx of such particles into the earth's ionosphere were established from magnetic data, from optical observations of the Polar Aurora, and from H.F. and V . H . F . radio observations close to the Auroral Zone. On the basis of these data the ionospheric disturbances for a 12-month period were related to prior distinctive solar events that were considered capable of producing the high energy particles. The selection was made from the optical, spectroscopic and radio data available. It was found that solar flares, eruptive prominences, and disappearing filaments were the most frequent sources of earth-reaching particles, and that large sunspots contributed only occasionally to ionospheric disturbances. Moreover, the radio noise data permitted a fairly reliable estimate of the particle veloci-ties in the solar corona, which could be used to predict the probability of subsequent terrestrial effects of the ejected particles. A correlation co-efficient of +0.65 was found between probabilities predicted from the solar information and observed ionospheric disturbances which occurred two or three days later. The observational evidence on the solar noise bursts and on ionospheric storms could only be explained on the basis of a considerable distribution of velocities for the particles ejected from the sun. It was shown that a Maxwellian distribution of corpuscular velocities was a probable one. Moreover, the observations indicated that particles are frequently emitted from the sun with a distribution of velocities, but that only in the case of very large phenomena are there sufficient particles with the high energies necessary to overcome the sun's gravitational field and reach the earth. PUBLICATIONS A S Q U A R E - W A V E M O D U L A T I O N M E T H O D FOR M I C R O W A V E SPECTRA T . R. Hartz and A . van der Ziel, Physical Review, 78, 473, 1950. A SOLAR NOISE O U T B U R S T O F J A N U A R Y 15, 1955 T . R. Hartz, Nature, 175, 908, 1955. T H U N D E R S T O R M SIGNALS AT V E R Y - H I G H F R E Q U E N C Y A N D U L T R A - H I G H F R E Q U E N C Y D. R. Hay and T . R. Hartz, Nature, 175, 949, 1955. RADIO STAR SCINTILLATIONS A N D T H E IONOSPHERE T . R. Hartz, Canadian journal of Physics, 33, 476, 1955. V . H . F . A U R O R A L NOISE T . R. Hartz, G . C. Reid, and E . L . Vogan, Canadian Journal of Physics 34, 728, 1956. G R A D U A T E S T U D I E S FIELD OF S T U D Y : PHYSICS Electromagnetic Theory : . .G. L. Pickard Electronics - A. van der Ziel Special Relativity Theory W. Opechowski Wave'Mechanics - : ......G. M . Volkoff Advanced Electronics A . van der Ziel Advanced Spectroscopy A . M . Crooker O T H E R STUDIES Theory and Applications of Differential Equations W . H . Gage Topics in Applied Mathematics E. Leimanis SOLAR-TERRESTRIAL RELATIONS by Theodore Robert Hartz B.A., University of Saskatchewan, 1945 B.Ed., University of Saskatchewan, 1946 M.A., University of Saskatchewan, 1948 A Thesis Submitted in Partial Fulfilment of the Requirements for the Degree of Doctor of Philosophy in the Department of Physics We accept this thesis as conforming to the required standard The University of British Columbia September, 1957 ABSTRACT The inter-relation of phenomena on the sun and in the earth's ionosphere has been studied in order to establish a causal dependence which could result from travelling solar particles. In addition to examining a variety of data from a number of sources, an extensive R.F. noise recording program was undertaken to provide information on those regions of the solar and terrestrial atmospheres from which optical data were not available. The occasions on which there was an influx of such particles into the earth's ionosphere were established from magnetic data, from optical observations of the Polar Aurora, and from H.F. and V.H.F. radio observations close to the Auroral Zone. On the basis of these data the ionospheric disturbances for a 12 month period were related to prior distinctive solar events that were con-sidered capable of producing the high energy particles. The selection was made from the optical, spectroscopic and radio data available. It was found that solar flares, eruptive prominences, and disappearing filaments were the most frequent sources of Earth-reaching particles, and that large sunspots contributed only occasionally to ionospheric disturbances. More-over, the radio noise data permitted a fairly reliable estimate of the particle velocities in the solar corona, which could be used to predict the probability of subsequent terrestrial effects of the ejected particles. A correlation coefficient of +O.65 was found between probabilities pre-dicted from the solar information and observed ionospheric disturbances which occurred two or three days later. The observational evidence on the solar noise bursts and on iono-spheric storms could only be explained on the basis of a considerable dis-tribution of velocities for the particles ejected from the sun. It was shown that a Maxwellian distribution of corpuscular velocities was a probable one. Moreover, the observations indicated that particles are frequently emitted from the sun with a distribution of velocities, but that only in the case of very large phenomena are there sufficient particles with the high energies necessary to overcome the sun's gravi-tational field and reach the earth. - i -In presenting t h i s thesis i n p a r t i a l fulfilment of the requirements f o r an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available for reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the Head of my Department or by his representative. It i s understood that copying or'publication of t h i s thesis f o r f i n a n c i a l gain s h a l l not be allowed without my written permission. Department The University of B r i t i s h Columbia, Vancouver 8, Canada. Date 2 7 , '9*7  FOREWORD AND ACKNOWLEDGEMENTS The research undertaken by the author has progressed in two stages. At the University of British Columbia considerable time was devoted to the building of sensitive electronic apparatus for the de-tection of weak R.F. signals. This work had a particular application to microwave spectroscopy. On joining the staff of the Radio Physics Laboratory of the Defence Research Telecommunications Establishment in Ottawa he continued the research, but devoted more emphasis to appli-cations outside the laboratory; and in particular to spectroscopic studies connected with radio astronomy. When i t was demonstrated that the well known spectral line of interplanetary hydrogen was the only one detectable with conventional equipment, he started a study of the earth's ionosphere by monitoring an extra-terrestrial noise source. Subsequently i t was found that much information on the solar and terrestrial atmo-spheres could be obtained through the use of techniques normally associ-ated with radio astronomy. The author wishes to acknowledge his indebtedness to those who contributed to the research. At the University of British Columbia he benefited from the guidance provided by Professor A. van der Ziel and from association with various students working in the research laboratories. At the Radio Physics Laboratory he was fortunate to have as associates such workers in the field of ionospheric and auroral research as Mr. J.C.W. Scott, Dr. J.H. Chapman, Dr. CO. Hines, and Dr. P.A. Forsyth who gave generously of their time and knowledge. In particular, the author would like to thank Dr. Forsyth for suggesting the problem and for his continued encouragement and advice. The helpful discussions on solar radio noise held with Mr. A.E. Covington of the National Research Council are also gratefully acknowledged. Mr. W.A. Sharf, Mr. W.E. Mather,rand various other members of the RPL technical staff assisted with the observations and undertook much of the routine maintenance of the recording equipment. The ambitious nature of the noise recording program would have been seriously curtailed without this assistance. In addition, Dr. G.C Reid scaled some of the scintillation recordings, and Miss M. Cain assisted with same of the data - i i -- i i i -reduction. T_e author acknowledges the unpublished solar data obtained from the following sources: the Fraunhofer Institute, Germany; the High Altitude Observatory, University of Colorado; the Central Radio Propa-gation Laboratory, U.S.A.; the U.S. Naval Observatory; the International Astronomical Union; the Royal Observatory of Belgium; the National Research Council; and the Dominion Observatory. Thanks are due to the Defence Research Board and to Mr. J.C.W. Scott, the Chief Superintendent of the Defence Research Tele-communications Establishment for permitting the use of the laboratory facilities in this connection. At the University of British Columbia the author was awarded two National Research Council studentships and a British Columbia Tele-phone Co. scholarship. TABLE OF CONTENTS Page ABSTRACT i FOREWORD AND ACKNOWLEDGEMENTS i i CHAPTER I - INTRODUCTION 1 CHAPTER II - RADIO TECBUIQPES AS RESEARCH TOOLS 5 2.1 The 50 mc Receiving Equipment 6 2.1.1 The Interferometer Type Receiver 1 2.1.2 Applications of the Equipment. . 13 2.1.3 Performance of the Receiving Apparatus . . . 16 2.2 The 500 mc Receiving Equipment 17 2.2.1 The Antenna 17 2.2.2 The 500 mc Receiver 20 2.3 Auxiliary Apparatus 23 CHAPTER III - THE DISTURBED IONOSPHERE .24 3.1 Methods of Studying the Ionosphere 25 3.2 Reception Quality on a Short Wave Communication Circuit 25 3.3 Radio Star Scintillations 39 3.3.1 Scintillation Dependence on Sidereal Time. . 42 3.3.2 Scintillation Dependence on Solar Time . . . 45 3.3.3 Correlations between Scintillations and Ionospheric Data hi 3.3.4 Scintillation Correlations with the Magnetic Field 49 3.3.5 Scintillations as a Measure of Iono-spheric Disturbance 49 3.4 The Planetary Indices of Magnetic Activity 50 3.5 The Aurora Borealis 52 3.6 Summary 53 CHAPTER IV - THE DISTURBED SUN 55 4.1 The Detection of Solar Particles 56 4.2 Noise Measurements at Times of Solar Flares . . . . 57 4.3 Noise Storms which Follow a Flare 61 4.4 Noise Storms and Sunspots 67 4.5 Solar Prominences and Filaments 7 1 4.6 M-Regions 72 4.7 Surmnary 72 -iv-- V -Page CHAPTER V - SOLAR PARTICLES 74 5.1 The Flare of February 8, 1957 76 5.2 Velocity Distributions 76 5.3 Particle Energies in the Corona 81 CHAPTER VI - THE IDENTIFICATION OF SOLAR PHENOMENA WITH IONOSPHERIC STORMS. . 86 6.1 Solar and Ionospheric Disturbances for 1956 . . . . 86 6.2 Flares as Sources of the Corpuscles 91 6.3 Surge Prominences and Disappearing Filaments. . . . 93 6.4 Sunspots 95 6.5 Other Solar Sources 97 CHAPTER VII - CONCLUDING REMARKS . . 99 7.1 Discussion. 99 7.2 Summary 103 BIBLIOGRAPHY 106 LIST OF ILLUSTRATIONS Page Figure 1. Block Diagram of 50 mc Interferometer Receiver 8 2. Reception Pattern of Two Spaced Aerials 8 3. Antenna Switch .10 4. Coaxial Switch Unit 10 5. Gnomonic Projection of Cassiopeia A Orbit showing Interferometer Pattern 11 6. Sample of the Scintillation Record 12 7. Azimuthal Equidistant Projection of Celestial Hemi-sphere with the Antenna Lobes from the Dipole Interferometer 14 8. Sample Recording of a Solar Noise Storm at 50 mc . . . . . 15 9. Block Diagram of 500 mc Receiver 18 10. The 500 mc Radio Telescope 19 11. Sample Record of 500 mc Solar Noise 22 12. Ionospheric Disturbance Data for January, 1956 . . . . . .27 13. Ionospheric Disturbance Data for February, 1956 28 14. Ionospheric Disturbance Data for March, 1956 .29 15. Ionospheric Disturbance Data for April, 1956 30 16. Ionospheric Disturbance Data for May, 1956 31 17. Ionospheric Disturbance Data for June, 1956. . . . . . . . 32 18. Ionospheric Disturbance Data for July, 1956 33 19. Ionospheric Disturbance Data for August, 1956. 31*-20. Ionospheric Disturbance Data for September, 1956 35 21. Ionospheric Disturbance Data for October, 1956 36 22. Ionospheric Disturbance Data for November, 1956 37 23. Ionospheric Disturbance Data for December, 1956 38 24. Samples of the Scintillation Amplitude Indices 40 25. Samples of the Scintillation Rate Indices 41 26. Scintillation Occurrence as a Function of Sidereal Time. . 43 27. Percentage Occurrence of Scintillations Vs l o g 1 0 s e c 9 • • ^ 28. Scintillation Occurrence as a Function of Solar Time . . .46 29. The Origin of R.F. Noise as a Function of Height in the Solar Atmosphere 58 30. Solar Noise Recordings for Feb. 8, I957 60 - v i -- v i i -Page Figure 31. Range of Solar Particles as a Function of their Initial Velocity 63 32. Time Required for Particles Starting at Rest to Reach the Sun as a Function of Starting Range 65 33. Time Required for Solar Particles to Reach the Earth as a Function of Initial Velocity 68 34. Magnetic Indices Following Two Importance 3 Flares . . . . 75 35. Assumed Maxwellian Velocity Distributions for the Solar Particles 78 36. The Distribution of Energy which the Solar Particles Introduce into the Ionosphere as a Function of Sun-Earth Travel Time 80 37. Distribution Curves for the Energy Introduced into a Region of the Corona at 2 x 10^ km Height by a Cloud of Solar Particles as a Function of the Travel Time of the Particles. 83 38. Distribution Curves for the Energy Introduced into Two Regions of the Corona by a Cloud of Solar Particles as a Function of the Travel Time of the Particles .84 39. Geomagnetic Indices Following a Number of Solar Flares . . 92 40. Geomagnetic Indices Following a Number of Surge Prominences and Disappearing Filaments 96 CHAPTER I INTRODUCTION Almost any considerations on the inter-relation of solar and terrestrial phenomena seem to be subject to a high degree of speculation. To be sure, the theories on the subject are mainly based on observational evidence, but this has always been severely restricted by the experi-mental circumstances. In consequence, the investigator a l l too frequently is given to theorizing in the hope that logical deductions arising from the theory will be subject to observational confirmation. The result is that there are now extant a number of theories to explain the solar in-fluences on the terrestrial ionosphere, and no doubt the future will see many more. Probably the most obvious conclusion one can draw in this con-nection i s that the subject is a very complex one. However, auroras and magnetic storms have certain essential features in common. Moreover, both appear to be closely connected to the sunspot cycle of BOlar activity. It seems possible, then, to find an explanation for the gross features, even though the many fine details are not as readily tied in. One of the main features of magnetic disturbances and auroral displays i s that both are concentrated near the poles. This has led to a general belief that these phenomena are produced by charged particles whose motions are influenced by the earth's permanent magnetic field. The solar dependence can be explained i f the particles originate on the sun. An early laboratory experiment^ with charged particles impinging on a magnetized sphere confirmed that gaseous discharges of the same general character as the aurora could be produced in this manner. The mathematical analysis of the behaviour of a charged particle (2) in the earth's magnetic field was started by BirkelandN ' and then developed in great detail by Stdrmerv->/v . According to this theory the sun sends out a stream of charged particles, a l l having the same sign. If this stream is directed at the earth i t will be deflected by the mag-netic field and, under favorable conditions, the ions will precipitate near the polar regions, where they can produce aurora and magnetic storms by bombardment of the atmosphere. However, not a l l the particles are able to reach the earth; most of them are deflected away by the magnetic field. -2-Many of these latter particles spiral around the globe at a great distance and constitute a giant ring current which causes the world-wide magnetic storms and whose presence causes a shifting of the auroral belt toward lover latitudes. While the so-called Stdrmer theory has many attractive features, i t suffers from a fundamental difficulty. In order to produce the auroral and magnetic effects the ion density in the incoming stream must be high; but this is impossible since electrostatic forces would disperse such a dense stream long before i t reached the earth. Consequently, the theory has been neglected for many years and other, more acceptable theories have taken its place. Recently, however, there has been a revival of interest in the Stdrmer theory. This followed a suggestion by Bennett and Hulburt^ that a beam of ions of the same sign would not necessarily be dispersed by electrostatic repulsion. They were able to show that i f the particle density and velocity of travel were high enough, the particles would stay in a confined beam because of magnetic self-focusing forces. To overcome the difficulty inherent in the concept of a stream of particles of the same sign, Chapman and F e r r a r o ^ postulated a neutral stream: that is, a stream consisting of an equal number of positive and negative ions. If the size of the stream is large enough i t should act essentially as a conducting plane. They suggested that such a plane, approaching the earth, is capable of producing effects similar to the observed i n i t i a l phase of a magnetic storm. Moreover, the main phase of the magnetic storm could be explained on the basis of charge separation in such a stream due to the earth's magnetic field. The theory as i t was originally proposed was unable, however, to explain any of the details associated with auroras or magnetic storms. A later amplification by Martyn v , / showed that the charge separation and the resulting ring current could be used to give a very plausible explanation of almost a l l the features of the auroral and magnetic disturbance phenomena. Accord-ing to Martyn's theory the leakage of charges from the ring current into the polar regions of the earth is the responsible agency for the aurora. -3-Another theory based on the concept of a neutral solar corpus-cular stream is that due to A l f v e n ^ . While i t contains similar features to the Chapman-Ferraro theory for the i n i t i a l phase of the mag-netic storm, i t differs radically in the particulars. The important thing in AlfVen's picture is the polarization field set up in the in-coming stream. This field is neutralized through an electrical discharge which takes place along the earth's magnetic lines of force. On this hypothesis the aurora is due to an electric discharge and not to lumin-escence in the atmosphere produced by high speed ions as postulated in other theories. In opposition to the solar corpuscular theories outlined above (9) is the theory of Hulburtv . This suggests that the auroral phenomena and magnetic disturbances are due to charged particles of terrestrial origin which- are produced in the upper atmosphere by the sun's ultra-violet rays. Such corpuscles are then guided through the high ionosphere by the earth's magnetic field toward the polar regions where they descend to lower levels and cause the observed auroras and geomagnetic storms. The theories outlined above are by no means the only ones on the subject. From time to time others have been suggested but none has sur-vived as long or received as much consideration as these. However, even the principle theories have been subjected to severe criticism of various kinds and no one of them has yet gained universal acceptance. Many in-vestigations are now under way, and, as more evidence is gathered the situation will probably clarify i t s e l f . For instance, the observational data now leave l i t t l e room for doubt as to the solar origin of the corpuscles, thus refuting the ultra-violet light theory. It is not the author's intention to become involved in a dis-cussion of the theories of the aurora. Any theory, of course, depends on experimental data for its verification, and the present investigation has been confined largely to observational evidence. There is as yet much conflicting data on solar and ionospheric phenomena which must be resolved before an acceptable theory can be found. The author set out to examine the inter-relation of certain solar and terrestrial phenomena. He applied a new approach to an old problem; namely, that of using some of the modern techniques of radio astronomy to study the sun and the earth's ionosphere. The data that were obtained consisted of recordings of R.F. radiation from the sun on two frequencies, and recordings of the fluctuations in the R.F. radiation from a strong 'radio star'. The former provided a more complete picture of solar phenomena, while the latter furnished a measure of the fine scale ionospheric structure. These data were examined in conjunction with other data on the sun and on the ionosphere in an attempt to establish a causal link between the two. Much of the following is a consideration of cause and effect: the causes being eruptive phenomena on the sun, and the effects being ionospheric disturbances. An intervening mechanism of corpuscular trans-port is implied, but the details are neglected unless positive evidence is available. In addition the nature of the particles is neglected to a large extent. Meinel's observations^10^ suggest that the particles are protons, but other evidence implies that Ca ions are involved^ 1 1^. The author assumes only that ions of some sort travel from the sun to the earth, and that some, but not necessarily a l l , are protons. Chapter II is devoted to a description of the apparatus used and the nature of the recordings that were obtained. The next chapter is a consideration of the data which characterize the disturbed ionosphere, while Chapter IV deals with the observational data on solar disturbances. In Chapter V a short discussion on particle transport from the sun to the earth is presented, which serves as a background for the interpretation of much of the data. The solar data for a 12 month period is related to the corresponding ionospheric data in Chapter VI, and criteria are presented which permit the prediction of terrestrial effects following eruptive events on the sun. Finally, the last chapter is devoted to a brief dis-cussion of the results and a summary. CHAPTER II RADIO TECHNIQUES AS RESEARCH TOOLS When the radar operators on the c l i f f s of Dover in 19^2 found their receivers being jammed by interference from the sun^12^, a new era in solar research was born. In paying tribute to Hey and his associates who recognized the phenomenon for what i t was, one should not lose sight (It)(Ik) of the effort of such early pioneers as Ebert and Lodge* - > / v ' who postu-lated this sort of radiation and actually looked for i t as early as 1900. Their failure to detect i t was due, in part, to technical difficulties, and, in part, to their limited knowledge of the various parameters that affect the solar radiations. On the other hand, the success of Hey, and also that of Southworth and Reber only very shortly later^-5) (16)^ resulted from the remarkable advances of radio technology of the pre-ceding decade, which in turn permitted the correct assessment of the observed signal. In the years that have followed the war the study of solar noise radiation as well as other extra-terrestrial radiation has grown at a tremendous rate. Not only has a better picture of the sun and its radiation processes developed, but man's knowledge on other aspects of the galaxy is undergoing revision. The astronomer is now well aware that the significance of certain phenomena can be appreciated only i f they are studied simultaneously by radio and by optical instruments. The studies of solar radiation undertaken at DRTE by the author have been carefully correlated with other radio measurements and with data obtained by optical and spectroscopic means. It was thought that some of the conflicting ideas on the subject could probably be resolved i f a com-prehensive picture of various solar phenomena could be obtained, even i f only for a relatively short period of time. Such a comprehensive picture would result i f a great many data, obtained by a number of observers, were pooled and examined as a unit. A modest attempt in such a direction has been made in the present analysis, and, while i t is not nearly extensive enough, some worthwhile conclusions can be drawn. In particular, i t was found possible to specify some of the phenomena responsible for the emission of solar particles of matter. -5--6-The principle data used in this analysis are measurements of solar radio noise. The author was engaged in a program of noise record-ings at frequencies of 50 and 500 mc. However, these measurements are supplemented by observations from other parts of the world and at other frequencies in order to extend the period of the day during which results are obtainable from the sun. Various optical data are also examined in conjunction with the radio observations in an effort to arrive at some comprehensive conclusions. The situation is complicated in that the emission and propagation characteristics of light waves and radio waves differ greatly in the sun's atmosphere. This means that the optical obser-vations are usually made of chromospheric and photospheric phenomena, while the radio measurements are limited to the corona. In fact, at 50 mc, observations in the upper corona only were possible and no radiation was 5 obtained from regions below about 2 x lCr km from the photosphere. The radio noise measurements made in the present study pertain only to those phenomena which occur in the corona, and which could be con-sidered as indicative of the emission of particles from the sun. The exact nature of the mechanism by which a travelling cloud of particles initiates the radiation of R.F. energy in the solar atmosphere is of minor concern. It is sufficient to refer to the measurements of Wild, et a l ^ 1 ^ 1 ^ , which show that noise-producing "disturbances" travel outward from the sun with velocities compatible with the observed delays between a large solar flare and the accompanying terrestrial magnetic storm. In succeeding sections this matter shall be considered at greater length, and a number of probable sources of solar particles will be discussed. 2.1 The 50 mc Receiving Equipment In contrast with some of the elaborate "radio-telescopes" in use today to receive radiation from outer space, the apparatus employed in this investigation was quite simple. The antenna directivity and receiver sensitivity were so chosen that certain discrete extra-terrestrial sources were detected and any general background of cosmic noise was discriminated against, as were local interference and receiver noise. However, the dis-crete source could not be specified unambiguously in a l l cases: i t was possible to identify the respective sources only after a careful examin-ation of the recorded signal had been made. -7-Since this was the case, several receivers differing only very slightly from each other were used to observe several sources, and the correct identification was possible from the distinctiveness of the signal. In particular, i t was found possible to monitor the radiation from the 'radio star', Cassiopeia A, as well as that from the sun with similar (19 receivers. Several such receivers were built and operated by the author* a l l of which were based on the same principle. A brief description of one will be given here and modifications for the others will be mentioned. 2.1.1 The Interferometer Type Receiver The receiving equipment i s portrayed in idle block diagram of Figure 1. Some similarity to the original apparatus used by Ryle^ 2 0^ can be noted; however, the circuit details differ radically. The R.F. receiver consisted of two parts, a wide band (approx. 1.0 mc) amplifier, having a gain of 50 db, preceded by a narrow band (approx. 25O kc) pre-amplifier having a gain of approximately kO db. This particular combination was desirable because of the tuning problems en-countered in the presence of interfering signals. A wide band receiver was preferable since the noise power received is proportional to the band-width, but there was never any assurance that a receiver tuned to a pre-selected frequency would not receive radiation from distant transmitters on the same frequency. In practice, i t was necessary to search for a clear channel in the desired frequency range. This was done in the following manner: the main R.F. amplifier was tuned to a preselected frequency, then the pre-amplifier with its narrower response was tuned to the various frequencies within this band until no interfering signal was observed. With each tuning adjustment i t was necessary to permit the apparatus to record for a period of at least 2h hours. The R.F. amplifier was followed by an amplitude detector, and an audio frequency amplifier, tuned to the antenna switching frequency by means of twin T feedback networks. The bandwidth was chosen narrow enough to reject stray 60 cps voltages as well as excessive noise at unwanted frequencies which might have overloaded the following detector. The output of the audio amplifier was applied to the phase sensitive detector, and the detected voltage was integrated with a -8-A NT. SWITCH R.F. AMPL • AUDIO AND 1ST AMPL. DETECTOR PHASE SENS. DET RECORDING! METER INTEGRATOR AND D.C. AMPL. FIG. I - B L O C K D IAGRAM OF 5 0 M c . R E C E I V I N G S Y S T E M FIG. 2 - R E C E P T I O N P A T T E R N OF TWO S P A C E D A E R I A L S (a) C O N N E C T E D IN P H A S E A N D (b) C O N N E C T E D IN A N T I - P H A S E - 9 -conventional Miller circuit and recorded by means of an Esterline-Augus milliameter. A zero-centre meter was employed so as to permit deflections in either direction depending on the sign of the detected voltage. An R.C. phase shift oscillator was employed to drive the antenna switch and provide the reference voltage for the phase sensitive detector. An interference pattern was formed by the two antennas depicted in Fig. 1 when they were connected through electrically equal lengths of cable to the receiver. The switch, placed in one antenna lead, alternate-ly introduced a phase reversal in that lead, thus interchanging maxima and minima of the interference pattern. This is represented in Fig. 2 where the antenna response pattern i s sketched for each of the two switch positions. The two alternative paths for the energy received by the one antenna may be seen in the schematic drawing of the antenna switch shown in Fig. 3. The two paths by which radiation from antenna 1 reach the receiver differed in length by one half wavelength: the phase reversal was accomplished by directing the radiation alternately along one path and then along the other. The details of the coaxial switch unit are shown in Figure k. Standard coaxial connectors were used on the extremities, and appropriate lengths of RG 9/u coaxial cable were attached when the complete switch was assembled. The diodes employed in the unit were 1N91 germanium junction diodes, which, while conducting, had a very low impedance com-pared to the 50 ohm characteristic impedance of the R.F. transmission line. Two three-element yagi antennas were used for the interferometer system. These were separated by 200 feet in an east-west line and mounted 3/8 of a wavelength above the ground. The resulting antenna sensitivity-pattern was a series of lobes, as shown in Figure 5. This figure is a gnomonic projection of the celestial sphere in the direction of the pole, showing the orbi* of the radiating source with the antenna lobes super-imposed. (The lobes as shown are for one switch position; the maxima and irrfn-tmn interchange under the action of the antenna switch.) As the earth rotated, the source apparently moved through the successive lobes, and produced a recording similar to that shown in Fig. 6. In this figure the recorded intensity is represented by the deviation (irrespective of -10-ANTENA I. X •—' W—r-ANTENA 2 AL  COAXIAL LINES ARE R G 9/u CABLES F I G . 3 - A N T E N N A SWITCH 1000 MMF F I G . 4 - C O A X I A L SWITCH UNIT GNOMONIC PROJECTION OF CASS. A ORBIT SHOWING INTERFEROMETER PATTERN -13-direction) from the chart centre. As the sensitivity pattern of a three-element yagi antenna was s t i l l broad enough to include the discrete source, Cassiopeia A, at a l l times of the day, a continuous recording was obtained with the apparatus. 2.1.2 Applications of the Equipment The equipment described above was used to record the radiation from the radio star, Cassiopeia A, at a frequency of 53 mc. A second apparatus was used to record noise from the sun at a frequency of 50 mc. The second equipment differed only slightly from the f i r s t : the antennas used were simple dipoles instead of yagi antennas. The modification per-mitted the recording of radiation from a greater portion of the celestial sphere than was possible with the more directive antennas — the eastern and western horizon were the only parts not adequately covered with this system. To be sure, the radiation from intense radio stars such as Cassiopeia A and Cygnus A was also recorded with this apparatus, but the different phenomena could be readily distinguished on the charts, particu-larly i f the receiver gain and recording time constant were chosen appropriately. The nature of the recordings can be better understood by refer-ring to Figure 7. This shows an azimuths! equidistant projection of the celestial hemisphere at the latitude of Ottawa with the antenna pattern superimposed on half of the diagram. The orbit of Cassiopeia A is indi-cated, as i s that of the sun at several times of the year. The use of the overlay permits one to read equal time intervals from the figure. On this overlay mean solar times are given: in order to get actual solar positions, (21) a correction for the equation of time must be made* '. Since the noise from the normal, undisturbed sun was consider-ably weaker than that from Cassiopeia A, the records from this second receiver usually showed only the characteristic sinusoidal trace due to the radiation from the radio star. From time to time, increases in the solar radiation occurred and then the solar noise appeared on the record superimposed on the stellar noise. As the sun moved through the antenna lobes, the record traced out a sinusoidal pattern of somewhat different period from that due to Cassiopeia A. This may be seen in the sample record of a solar noise storm shown in Fig. 8. O9O0 Dec 21 Feb 2 3 -Oct 2 0 Mar 21 Sept 2 3 0 8 0 0 Hor izon S o l a r Orb i ts on dates shown Apr \7 Aug 28 JUNE 21 WE 9 T Zeni th Orbi t of ' C a s s i o p e i a Celest ia l Pole Hor i zon 7 0 0 0 5 0 0 0 4 0 0 Antenna Lobe Max imum 0 2 0 0 3 1 0 0 An tenna Lobe M in imum NO^TH F I G 7 - A Z I M U T H A L E Q U I D I S T A N T P R O J E C T I O N O F C E L E S T I A L H E M I S P H E R E W I T H T H E A N T E N N A L O B E S F R O M T H F D I P O L E I N T E R F E R O M E T E R -16-2.1.3 Performance of the Receiving Apparatus For an interferometer type phase switching receiver which de-tects the radiation from a discrete source, i t has been shown^2<^ that the signal-to-noise ratio at the receiver input is given by: rj ( = P Va3.agA1.A2 ( 1 ) ( a ^ + agAg)/2 + ^ ( O J F - L + OgFgJ/fe + N In this equation P is the power flux at the antennas due to the discrete source, Fj. and Fg are the effective values of the incident flux from the galactic background falling on the two antennas, Ai and Ag represent the effective areas of the two antennas for radiation from the direction of the discrete source, a^ and Og are the attenuation constants for the two antenna cables, X is the wavelength, and N is the noise power generated by the receiver. At a frequency close to 50 mc, and for the three-element yagi antennas employed, this ratio was approximately l / l 6 . How the maximum output signal-to-noise ratio, r^g, can be expressed in terms of D-i and the input and output bandwidths of the 1 (22) receiver, B^ and Bg respectively* ': rig = r j i B x ^ (2) For the Cassiopeia receiver, was approximately 25O kc and Bg approxi-mately l / 6 cps. Then rig _ * k.o x 10k That is, the amplitude of the meter deflection for the Cassiopeia A signal was k6 db in excess of the amplitude of the statistical fluctu-ations on the record. Consequently, the apparatus was able to measure the fluctu-ations, or scintillations of the radio star quite reliably without being limited by either cosmic or receiver noise. Likewise, the solar receiver was able to record small noise bursts reliably. In Fig. 6 sample records have been shown for the Cassiopeia receiver. Superimposed on the usual sinusoidal pattern are rapid -17-fluctuations, or scintillations produced by the ionosphere. That these are actually produced in the ionosphere and not on the source has been demonstrated by Smith, Little, and Lovell^ 2^ 2 1*') with spaced receiver measurements. Since the intensity of the source has been shown to be fairly constant, a l l the fluctuations observed by the author on the re-cordings were assumed to have been caused by ionospheric irregularities which diffracted or refracted the radiation passing through. It was found that the details of the recorded intensity were best portrayed i f a short integrating time was used: in the equipment under consideration two output channels, with time constants of 1.5 sec. and 6.0 sec, were employed. A sample record from the solar receiver is shown in Figure 8 for a period when a noise storm was in progress. A time constant of one minute was chosen for the solar receiver so as to smooth out most of the ionospheric fluctuations on the recording and show only the main features of the solar noise. In general, noise storms usually lasted from several hours to several days, but they were only one manifestation of a disturbed sun. At other times bursts of radio noise occurred which lasted for a few minutes to several tens of minutes. 2.2 The 500 mc Receiving Equipment While an interferometer type receiver was suitable for solar noise measurements at 50 mc, the noise level at 500 mc was such that a different type of receiving equipment was desirable. At this higher fre-quency i t was more feasible to use a parabolic antenna on a steerable mount in preference to an interferometer formed by two fixed antennas. In consequence, the receiver was designed to measure total power in con-trast with the power comparison system used at 50 mc. The receiver design was based on an apparatus built by Machin, Ryle and Vonberg^22^. The complete equipment is depicted in the block diagram of Fig. 9 and a brief description of the component parts i s given in the following sections. 2.2.1 The Antenna A photograph of the antenna structure is shown in Fig. 10. The parabolic reflector was designed and built at RPL for this particular ANTENNA SWITCH ^^ANTENNA RECORDER NOISE DIODE i D.C. AMPLIFIER ROTATING C O N D . LIGHT CHOPPER -^f'^ MOTOR 30 cps SYNCHRONOUS AMPLIFIER H CD I PHASE SENSITIVE DETECTOR PREAMPLIFIER R. F. AND F AMPLIFIERS 30 cps AMPLIFIER F I G . 9 - B L O C K D IAGRAM OF 5 0 0 Mc. R E C E I V E R -19-FIG. 1 0 - T H E 5 0 0 Mc. RADIO T E L E S C O P E -20-purpose. It measured 20 feet in diameter and had an 11 foot focal length. The reflecting surface was composed of one-half inch hardware cloth, which was supported on a plywood framework. The antenna, feed at the focus consisted of a dipole backed by a reflecting disc. Measurements made on a similar antenna indicated a gain of approximately 27 db com-pared to an isotropic radiator, and a beam width of approximately 6° at the half power points. The parabolic dish was supported on an equatorial mount to facilitate the tracking of celestial bodies. This structure was designed at DRTE and built by the Dominion Bridge Company. It was so arranged that the antenna could be driven from east to west to maintain a fixed direction and thus follow the sun in its orbit. This was accomplished by a synchronous motor which ran on mean solar time. Minor corrections in the antenna setting were made every few days to compensate for errors introduced by the irregular motion of the sun and for changes in the sun's position with season. B G 8/u coaxial cable was employed to connect the antenna to the rest of the receiving equipment. A coaxial rotating joint permitted con-tinuous rotation in one direction and the parabola did not have to be reset on the sun at sunrise every day. 2.2.2 The 500 mc Receiver The signal from the antenna and the output of the noise diode were both applied to the antenna switch as indicated in Fig. 9* The switch was operated by a 30 cps synchronous motor in such a way that the two inputs were alternately connected to the receiver. The signal appear-ing at the switch output then consisted of noise modulated by a 30 cps amplitude component. This component was detected in the subsequent stages and provided a direct indication of the misalignment between the noise diode output and the signal from the antenna. In the succeeding stages this detected signal was used to adjust the level at which the noise diode operated to make i t equal to the noise power coming from the antenna. The antenna switch consisted of a rotating condenser which alternately produced open circuit and short circuit conditions on a set -21-of R.F. transmission lines as indicated in Fig. 9 . When the rotor plates meshed with the stator plates at A, an- R.F. short circuit to ground was provided at that point. This was reflected into an open circuit a quarter wavelength away, and permitted the signal from the antenna to reach the receiver. At the same time the stator plates at B presented an open cir-cuit to ground, which became a short circuit one quarter wavelength along the transmission line and prevented energy from the noise diode reaching the receiver. This situation was reversed on the next half cycle, when the signal from the noise diode and not that from the antenna was able to reach the receiver. For the main receiver an RCA Victor Co. C.T.R.-5OO receiver was employed with l i t t l e modification.. The R.F. and I.F. stages were left unchanged, but a pre-amplifier was added to improve the noise figure. Using a novel design the author was able to build a pre-amplifier having (25) a noise figure of about .5 ab at this frequencyx . In the 30 cps amplifier a twin T feedback circuit was used to achieve a narrow bandwidth at the switching frequency. This amplifier was followed by a phase sensitive detector, and a conventional Miller integration circuit. The detected voltage was applied to a D.C. amplifier which controlled the heater current in the TT1 noise diode. If any D.C. "error" voltage appeared here, i t was because the noise powers from the antenna and the TT1 were not equal, and the error voltage then produced a change in the heater current of the diode to make the two noise powers equal again. Consequently, within the limitations set by the integration time, which for this receiver was approximately 5 sec, the two noise powers were always maintained equal. The output power of a noise diode is proportional to the plate current of the tube. Hence i t was possible to measure the noise power coming in on the antenna by simply recording the D.C. current flowing through the noise diode. An Esterline-Angus recording meter was used for this purpose; a non-linear shunt having been applied to provide a scale which was approximately logarithmic. A sample recording made with this apparatus is shown in Fig. 11. This diagram shows the night-time noise level, the interference pattern at the horizon at sunset, and the noise 1200 1000 0800 0600 0400 0200 2400 TIME (GMT) FIG. I I - S A M P L E R E C O R D OF 5 0 0 Mc . S O L A R NOI§E -23-level from the quiet or undisturbed sun on two representative days. 2.3 Auxiliary Apparatus In addition to the noise receivers mentioned above, several other equipments were operated to provide extra information on the sun, or on the ionosphere. While some data were obtained from each one, not a l l of the ejqu|>pments were in continuous operation. Moreover, in most cases the desired data were readily forthcoming from some other agency and the apparatus operated at DREE was only used to provide an early indication. These auxiliary equipments included: (a) A recording magnetometer to provide a measure of the horizontal component of the earth's magnetic field at Ottawa^26). (b) A four inch telescope to provide data on sunspots. (c) A 27 mc interferometer receiver of the type described in the previous sections to record both scintillations of the Cassiopeian source and solar noise. (d) An optical recorder using a photo-multiplier tube directed at the northern sky to record the occurrence of aurora on clear nights. Besides this auxiliary apparatus which was operated specifi-cally for correlation purposes, results on ionospheric phenomena were available from other workers in DRTE. These proved invaluable for assessing the significance of many of the measurements made by the author with the above equipment. CHAPTER III THE DISTURBED IONOSPHERE The early workers in the field of radio had great difficulty in explaining the observed propagation over great distances around the curvature of the earth. This situation was eased by the proposals of Heaviside^ 2^ and Kennelly^ 2^ that an ionized reflecting layer existed at some distance from the earth's surface. That this "ionosphere" did in fact exist was soon proven experimental1y, and much of the early re-tea) search can be found in a text book on the subject by Mitra N , It has now been established that instead of being a single layer, the ionosphere actually consists of a number of distinct layers or ionized regions; the principle ones being the E layer at a height of about 100 km, the Fl layer at a height of approximately 200 km, and the F2 layer at a height ranging from 300 to 400 km. The E and PI layers are known to be daytime or sunlight layers, while the F2 layer exists at a l l times of the day and year. These ionized regions are formed by the sun's ultra-violet radiation which ionizes the molecules of air that exist at these particu-lar altitudes. Since the solar radiation is fairly well known for differ-ent regions of the earth at various times of the year, the normal conditions of the ionosphere can be fairly well predicted. There are, however, two other types of solar radiation which affect the ionosphere. By contrast, these may be considered as abnormal radiation since they cannot be predicted and since they produce abnormal ionization conditions in the ionosphere. One is ultra-violet radiation of very great intensity which penetrates to levels below the E layer and causes high ionization densities in the so-called D region. Such occur-rences usually last for several tens of minutes and seem to coincide closely with times of solar flares. The other form of radiation is corpuscular in nature and may produce extra-ordinary effects at a l l levels in the ionosphere. The present investigation is concerned with this latter form of ionospheric disturbance or abnormality. To a large extent the normal ionospheric conditions will be ignored and only those manifestations of solar corpuscular radiation will be considered. It is hoped that a study of such disturbances will throw some light on the relationship to the solar emission processes. -24--25-3.1 Methods of Studying the Ionosphere Traditionally information on the ionic densities in the iono-spheric layers has been gathered by radio sounding m e t h o d s I n this technique, pulses of radio frequency radiation at successively increasing frequency are directed vertically at the Ionosphere and the times for the respective return echoes are noted. The delay times for those frequencies that are reflected provide information on the ionization below the reflec-tion region. It is obvious that no data on the upper side of the various layers can be obtained with this technique. Moreover, when a disturbance occurs which produces greatly increased absorption in the D region, no reflected echoes from the higher regions are observed. Consequently, the sounding method, while quite satisfactory for investigating the normal ionosphere is not very useful in providing a complete picture of the abnormal ionosphere. The problem i s then: how does one specify the .conditions which exist at times of disturbances, and, in particular, during those distur-bances produced by corpuscular radiation from the sun? While a complete answer to this question is not possible at the present time, several manifestations of the abnormal ionosphere can be given which provide a partial answer. These also suffice to specify when an ionospheric dis-turbance or storm is in progress, and permit one a rough classification as to the intensity thereof. In the following sections some of these measures of ionospheric abnormality will be treated in more detail. 3.2 Reception Quality on a Short Wave Communication Circuit At a number of places on the earth's surface sounding trans-mitters are in operation sampling ionospheric echoes at a l l times of the day and night. The data from these stations, together with a measure of the solar ultra-violet radiation, are used for the prediction of the optimum point-to-point transmission facilities for high frequencies. For instance, RPL has such a prediction service, based on observations of the Canadian ionosphere, which is valid for communication circuits in Northern Canada^ Experience has shown that fairly accurate predictions are possible in this manner, and that the operators on transmission cir--26-cuits are able to maintain communications by compensating for the pre-dicted changes in ionospheric conditions. However, the solar corpuscular radiation is not predictable in the same manner, and when such radiation occurs i t is followed by disturbed ionospheric conditions which cannot be compensated for by the communication operators. At such time the transmission quality deterior-ates markedly. It follows, then, that the reception quality on a long range high frequency communication path can be used as an indication of ab-normal ionospheric conditions. The data given below were obtained from the Canadian Overseas Telecommunications Corporation for the trans-Atlantic circuit between Montreal and England. Hourly indices ranging from 0 to 5 had been assigned according as reception conditions were excellent, good, fair, poor, very poor, or "black-out". While, strictly speaking, this was not a quantitative measure since i t depended on such factors as the experience and s k i l l of the operator and the conditions of his receiver, i t could s t i l l serve as a reliable indication of abnormal conditions since the operator made a comparison to what would be normal reception for him.. The trans-Atlantic quality data for the year 1956 were smoothed and then plotted in the upper frame of Figures 12 to 23 inclusive. This smoothing process consisted of applying a running average to the data for each 2k hour period. In this way minor fluctuations were smoothed out, as were daytime effects of the sun's ultra-violet radiation, leaving only major variations which could be attributed to corpuscular radiation. Because the reflection process was involved, this particular measure provided disturbance data for only the lower regions of the iono-sphere. To be sure, similar data could have been obtained from vertical soundings but i t would have been in a less convenient form since the automatic recorder cannot make the comparison to normal ionosphere con-ditions as can the manual operator. Moreover, the trans-Atlantic path was preferred since i t skirted the auroral zone, which is the most likely region for the precipitation of incoming charged particles. . 3^  2^  TRANS-ATLANTIC QUALITY h4 SCINTILLATION RATE 5-4-3-2-I-0-h2 i ro i Kp INDICES MOST SOUTHERN LATITUDE \/FOR AURORA ,i I I I I «_ • • ' I I I I ! I 3 5 7 9 II ' . J i, i i, i i r i i i i i i i i i i i i i i 13 15 17 19 21 23 25 27 29 31 DAY OF MONTH FIG. 12-IONOSPHERIC DISTURBANCE DATA FOR JANUARY 1956 ho •50 -52 •54 3H TRANS-ATLANTIC QUALITY 6-5-4-3-2-H 0-A ^ i \ i \ / » J Kp INDICES A \ X / / v / / ASCINTILLATION RATE A MOST S O U T H E R N LATITUDE FOR A U R O R A h 4 k3 h2 T 1 '-J ',J 1 1 1—I I I I — l — 1 — I — I — l — 1 — i — r — j — i — | — | — | I 3 5 7 9 II 13 15 17 19 21 23 2 ^ 27 29 DAY OF MONTH FIG.I3H0N0SPHERIC DISTURBANCE DATA FOR FEBRUARY 1956 ho 50 52 h54 • 00 i 3-2-TRANS-ATLANTIC QUALITY SCINTILLATION RATE 6-5-4-3-MOST SOUTHERN LATITUDE o FOR AURORA o 1 i — i — r "i—i—i—r 7 9 i i — i — i — i i i i i I I I I i I I r i i i i M 13 15 17 19 21 23 25 27 29 31 DAY OF MONTH FIG.I4-I0N0SPHERIC DISTURBANCE DATA FOR MARCH 1956 h3 h2 i ro i ho •50 -52 =54 3H 2H 6-5-4 -3-2 H OH TRANS-ATLANTIC QUALITY SCINTILLATION RATE Kp INDICES o • • o o o MOST SOUTHERN LATITUDE FOR AURORA i i i r*~i i° i * i * i i i i — i — i — i — i — i — i — i — i — i — i — i — r 9 II 13 15 17 19 21 2 3 2 5 2 7 2 9 - 4 i i -O - 5 0 - 5 2 - 5 4 i I i i \ i i 1 3 5 7 DAY OF MONTH FIG.I5- I0N0SPHERIC DISTURBANCE DATA FOR APRIL 1956 3 J 21 6-5-4-3-2 -0-TRANS-ATLANTIC QUALITY SCINTILLATION RATE MOST SOUTHERN LATITUDE FOR AURORA V l J l J 1 ' I 1 I I I I I I I I I I I I I 1 | | | 1 3 5 7 9 II 13 15 17 19 21 23 25 27 29 31 DAY O F MONTH FIG.I6-I0N0SPHERIC DISTURBANCE DATA FOR MAY 1956 - 4 -3 -2 •0 - 5 0 - 5 2 - 5 4 5 -4 -3 -2 -oH TRANS-ATLANTIC QUALITY Kp INDICES SCINTILLATION RATE MOST SOUTHERN LATITUDE FOR AURORA o o • e © I 3 5 T~I i i I I i i i i — i i — i — i — i — i — i — i — i — i — i — i — i — r 7 9 II 13 15 17 19 21 23 25 27 29 DAY OF MONTH RG.I7-I0N0SPHERIC DISTURBANCE DATA FOR JUNE 1956 3H 2^  6 -5-4 -3 -2-I-0-TRANS-ATLANTIC QUALITY Kp INDICES V V ^ V A / V . / SCINTILLATION RATE MOST/ SOUTHERN LATITUDE FOR AURORA V • • • I.J I I I I I 1 3 5 7 9 II 13 15 17 19 21 23 25 27 29 31 DAY OF MONTH FIG.I8- I0N0SPHERIC DISTURBANCE DATA FOR JULY 1956 h4 h3 u> I ho -50 -52 •54 H 6-5-4-3-2-oH TRANS-ATLANTIC QUALITY /V SCINTILLATION RATE At V\ Kp INDICES MOST SOUTHERN LATITUDE FOR AURORA 9 9 I.1 ' I ' ' ' I I I 1 I I I I I I I I I I 1 I I I I I I I I I I I 3 5 7 9 II 1.3 15 17 19 21 23 25 27 29 31 DAY OF MONTH FIG.I9- I0N0SPHERIC DISTURBANCE DATA FOR AUGUST 1956 -4 -3 -2 I - 0 -50 -52 -54 6H 5 4 2 H 0 TRANS-ATLANTIC QUALITY SCINTILLATION RATE A Kp INDICES MOST SOUTHERISKLATITUDE FOR AURORA I i i i i i i i i i i i I I l l I I I I I I I I I I I I I I I I I 3 5 7 9 II 13 15 17 19 21 23 25 27 29 31 DAY OF MONTH FIG.20-IONOSPHERIC DISTURBANCE DATA FOR S E P T E M B E R 1956 h4 h3 h2 1 U) I ho •50 -52 -54 3 2-I-6-5-4-3-2-I-0 -TRANS-ATLANTIC QUALITY SCINTILLATION RATE Kp INDICES .MOST SOUTHERN LATITUDE • vFOR AURORA ' , • ' ' ' I I I I I I I 1 I I I I I I I I I 1 I I I • I I 1 | | | I I 3 5 7 . 9 1 1 13 15 17 19 21 23 25 27 29 31 DAY OF MONTH FIG. 21-IONOSPHERIC DISTURBANCE DATA FOR OCTOBER 1956 -4 -3 CO I hi ho •50 •52 •54 3-2-6-5-4-3-?-OH TRANS-ATLANTIC QUALITY V Kp INDICES © • O o • • I i » SCINTILLATION RATE MOST SOUTHERN LATITUDE FO /^AURORA I I I I I J I I l l I I I I I 1 I I I l I I 1 I I I I I I l I I 3 5 7 9 II 13 15 17 19 21 23 25 27 29 DAY OF MONTH FIG.22-IONOSPHERIC DISTURBANCE DATA FOR NOVEMBER 1956 -4 -3 -2 -I - 0 50 •52 •54 TRANS-ATLANTIC QUALITY 6-5-4-3-2-I-0-V ,4/1 AA / V SCINTILLATION RATE / \ V \ rs A i / * « V / A/ \ I \ Kp INDICES I 3 5 7 9 II 13 15 17 19 21 23 25 27 29 31 DAY OF MONTH FIG.23- I0N0SPHERIC DISTURBANCE DATA FOR DECEMBER 1956 -39-3.3 Radio Star Scintillations Some ten years ago during a survey of the distribution of galactic noise, Hey, Phillips and ParsonsXJ ' noted that the noise from the region of Cygnus showed irregular amplitude fluctuations with periods of seconds or minutes. They concluded that a high intensity variable source existed in the area, but subsequent measurements by other workers, while confirming the existence of the source, showed that the intensity variation was superimposed near the earth* * '. Many other discrete sources or radio stars have now been identified, most of which appear to twinkle or scintillate in a manner analogous to the twinkling of visible stars. Since the scintillations are caused by irregularities in the ionosphere, a study of their occurrence and nature was undertaken by the author to provide information on the small scale structure of the ionized regions. The scintillation recording program at Ottawa was divided into three distinct phases. The preliminary results indicated that the fluctu-ations differed from those reported by other workers: in particular, scintillations were observed at a l l hours of the day and not only at night as observed elsewhere. On the basis of this, recordings were made for an i n i t i a l two year period using a long time constant (one minute). This meant that only the large significant effects were observable, and the recordings were then analyzed simply on the basis of scintillation occurrence or non-occurrence. The data were correlated with other geo-physical phenomena in an effort to determine general trends upon which to base subsequent work. Some of the results of this investigation are summarized in the following sections. The second phase consisted of a series of recordings made using a time constant of six seconds. This meant that much more detail was available on the records, and the subsequent analysis was made by assign-ing two fluctuation indices to the record for each hour; an amplitude index and a rate index. The amplitude index is the ratio of the r.m.s. deviation of the amplitude to the mean amplitude, and the rate index is a measure of the number of fluctuations per unit of time. Samples of the respective indices are given in Figures 2k and 25. Some of the results from this investigation are summarized in the following sections. 3 4 5 FIG. 24-SAMPLES OF THE SCINTILLATION AMPLITUDE INDICES -42-The third phase consisted of a detailed examination of the individual fluctuations using a recording time constant of 1.5 seconds. Only very limited results have been obtained from this phase of the in-vestigation, and these were in general agreement with the results from the second phase. 3.3.I Scintillation Dependence; on Sidereal Time It was early recognized that many factors may contribute to the scintillations observed on the recordings. The most obvious, perhaps, is the thickness of ionosphere that the radiation must penetrate to reach the receiver. For a ray incident at the surface of the earth with a zenith angle 6, this path length is given by: L = [(R + h x ) 2 - R2 s i n 2 e] 1/ 2 - [(R + h 2 ) 2 - R2 s i n 2 e] 1/ 2 (3) where R is the radius of the earth and h^ and h£ are the heights of the top and bottom of the ionosphere above the surface of the earth. If, now, a dependence exists between L and the fluctuations, a graph of scintillation occurrence as a function of sidereal time should show a maximum effect when the source is at lower transit ; (maximum ionospheric thickness), and a minimum effect 12 hours later. That this was the case can be seen in Figure 26 where the percentage occurrence of scintillations for a two year period is plotted against sidereal time. Since, however, the variation of L with sidereal time could not be made to coincide with this curve for any reasonable values of h^ and h 2, i t was concluded that some other factors had yet to be considered. An interesting empirical relationship was derived for the data when replotted in Figure 27 in a slightly different form. . If 9 is the angle of incidence for a ray from the source on the "top" of the iono-sphere (taken as 400 km), the following equation was found to f i t the experimental points: % occurrence = 194 l o g ^ sec 9 + 10 (4) This may alternately be expressed as: P = 1.94 l o g 1 0 sec 9 + 0.1 (5) i IOH O 3 6 9 12 15 18 21 24 SIDEREAL TIME (HOURS FROM UPPER TRANSIT) FIG. 2 6 - S C I N T I L L A T I O N O C C U R R E N C E A S A FUNCTION OF S I D E R E A L TIME I •4=-i F I G . 2 7 - P E R C E N T A G E O C C U R R E N C E O F S C I N T I L L A T I O N S V s l o g | 0 s e c <j> -45-where P is the probability of observing scintillations from a source which subtends a zenith angle q> at the 400 km level of the ionosphere. It follows that: e 1.19Pi = g e c ^ ( 6 ) where Pj_ = P - 0.1 i s the probability in excess of 0.1 of observing scintillations. In Fig. 27 the line has been extrapolated to 100 per cent. This point on the graph corresponds to an angle of incidence at the 400 km level of 70°, which in the experimental set up was the largest angle possible i f the ray was to reach the ground. It follows, then, that a radio star on or near the observer's horizon should always be scintil-lating. A further point of interest may be seen in Figure 26. The curve is slightly asymetrical about the time of lower transit. This suggests that a dependence on the earth's magnetic field may exist since geographic Worth and magnetic North do not coincide. An examination of the more detailed recordings was made by Reid which showed that such a dependence actually exists 3.3.2 Scintillation Dependence on Solar Time The data on scintillation occurrence were summed for a two year period and plotted as a function of solar time in Figure 28. In doing this the sidereal component could be disregarded since over a yearly period its effect was averaged out, just as the solar effect was averaged out on the sidereal plot given in the previous section. (This was possible only because there was no noticeable seasonal dependence in these d a t a . ) i n Figure 28, the percentage occurrence of scintillations i s plotted against local standard time. It can be seen that there was a somewhat higher occurrence of scintillations at night than during the day, with a weak wa-iH rmim before midnight and a somewhat stronger one after mid-night. A similar result was obtained by Costain V J > / from measurements made at Saskatoon. Hewish^^, however, working in England, found high night-time values and very low daytime values. -k6-6CH 12 15 18 21 00 03 06 09 12 LOCAL STANDARD TIME FIG. 28-SCINTILLATION OCCURRENCE .AS A FUNCTION OF SOLAR TIME - 4 7 -The recordings obtained with the six second and 1.5 second time constant supported this observation. In no case was a clear preference for very high night-time fluctuation indices and very low daytime indices found. 3.3.3 Correlations between Scintillations and Ionospheric Data One of the most important questions about scintillations i s : at what height in the ionosphere do they originate? In order to answer this question, workers in the field have attempted to correlate their find-ings with data on the ionospheric layers obtained by vertical sounding methods. Since this latter technique requires that a radio wave be re-flected by the ionosphere, much longer wavelengths are employed, and consequently the normal ionospheric layers do not exhibit the same fine structure that is observed with 50 mc scintillation measurements. The abnormal ionosphere, on the other hand, does exhibit some fine structure even at the lower frequencies used by the sounding equipment, and such phenomena as sporadic E echoes and spread F echoes are well known. The reader is referred to an ionospheric observer's manual for examples of these phenomenav J 1'. The ionospheric records from the Ottawa station were examined by the author for distinctive characteristics for the two year period from September 1953 to September 1955. The occurrences of the following phenomena were noted and correlated to the occurrence of scintillations: (a) Sporadic E echoes. An E layer critical frequency in excess of 4 . 8 mc was taken as a positive indi-cation of a sporadic E echo; any lower frequency was taken as a negative indication. (b) High F2 critical frequencies, that i s , critical frequencies more than 15 per cent above the monthly median. (c) Low F2 criti c a l frequencies, which were those critical frequencies that were more than 10 per cent below the . monthly median. (d) Spread F echoes, which were denoted by a value in excess of 3 on the normal scale of 0 - 9. -48-(e) High v i r t u a l h e i g h t s o f the F2 l a y e r , which were taken here as b e i n g i n excess o f 10 per cent above the monthly median. A l o n g term comparison o f these phenomena w i t h s c i n t i l l a t i o n occur rence i n d i c a t e d t h a t the re was nega t i ve c o r r e l a t i o n w i t h spo rad i c E occur rence , bu t some measure o f p o s i t i v e c o r r e l a t i o n w i t h F2 l a y e r phenomena. A more d e t a i l e d examina t ion o f the F2 l a y e r r e l a t i o n s h i p was at tempted by l o o k i n g f o r c o r r e l a t i o n s on b o t h an hour-by-hour b a s i s and a day-by-day b a s i s . However, o n l y a v e r y s m a l l p o s i t i v e c o r r e l a t i o n (+0.19) was found between h i g h F2 v i r t u a l h e i g h t s and s c i n t i l l a t i o n occu r r ence . A l l the o the r c o r r e l a t i o n c o e f f i c i e n t s were so n e a r l y zero as t o be i n s i g n i f i c a n t . The c o n c l u s i o n drawn from t h i s was t h a t the method o f r e l a t i n g s c i n t i l l a t i o n occurrence t o i o n o s p h e r i c sounding da ta was not a r e l i a b l e means o f de t e rmin ing the h e i g h t o f the a c t i v e r e g i o n . The evidence seemed t o favour the F2 o r upper r e g i o n o f the ionosphere i n preference t o the E o r l ower r e g i o n as r e s p o n s i b l e f o r s c i n t i l l a t i o n p r o d u c t i o n , bu t i t was i n no way c o n c l u s i v e . Some r e s e r v a t i o n s shou ld be made i n t h i s c o n n e c t i o n . The i o n o -sphere da t a were ob ta ined f o r the ionosphere overhead a t Ottawa, w h i l e the s c i n t i l l a t i o n da t a p e r t a i n e d t o a r e g i o n c o n s i d e r a b l y t o the n o r t h . (At upper t r a n s i t the r a d i a t i o n from the source met the 400 km l e v e l about 90 km n o r t h o f Ottawa, w h i l e a t l ower t r a n s i t the d i s t a n c e was about 1150 km. The i o n o s p h e r i c p r o p e r t i e s may change a p p r e c i a b l y w i t h i n d i s -tances o f t h i s magnitude, p a r t i c u l a r l y near the a u r o r a l zone . ) The i o n o -sphere da ta were ob t a ined a l s o by a v e r y sho r t term sample, w h i l e the s c i n t i l l a t i o n da ta were averaged over an hour . The F2 l a y e r , whose a c t i v i t y seemed t o c o r r e l a t e be s t w i t h the s c i n t i l l a t i o n r e c o r d s , i s f r e q u e n t l y obscured by a lower l a y e r d u r i n g d i s t u r b a n c e s , and the da ta were thus m i s s i n g j u s t when they were l i k e l y t o be most u s e f u l . The s c i n t i l l a t i o n d a t a , moreover, r e l a t e t o s m a l l s c a l e phenomena which may be i n no way comparable t o the s c a l e observab le a t the l o n g e r wavelengths w i t h the i o n o s p h e r i c sounder. -49-3«3»^ Scintillation Correlations with the Magnetic Field A positive correlation was observed between scintillation occur-rence and high magnetic indices from the Agincourt observatory. While in i t i a l l y this was noted for the combined data covering an extended period*-^', subsequent examination showed that positive correlations were observed on a day-to-day basis. Magnetic K indices from the stations at Meanook, Alberta, which is near the edge of the auroral zone, and at Agin-court, which is approximately 25O miles southwest of Ottawa, were examined by the author in this connection. The planetary K indices, which are a measure of world-wide magnetic disturbances, were also investigated. Correlation coefficients ranging from +O.25 to 4O.3O were obtained in this manner^-3 , indicating that some relationship, though slight, did exist between the scintillation and the magnetic data. In order to highlight this relationship, the more detailed scintillation data obtained with recording time constants of six seconds and 1.5 seconds were correlated to the planetary K indices on a three hourly basis. The scintillation data were fi r s t smoothed to remove a dominant sidereal dependence. This was accomplished by applying a 24 hour running average to the data. The magnetic indices were similarly treated so as to make the data compatible. (Minor random fluctuations were thus smoothed out, as was any diurnal trend. However, a l l major significant features were retained in only slightly altered form by this process.) The correlation coefficient obtained in this manner for the results covering a 14 month period was +O.56. 3.3.5 Scintillations as a Measure of Ionospheric Disturbances The results outlined above have shown that scintillations are a definite manifestation of a disturbed ionosphere. The relationship with the occurrence of spread F echoes, when i t existed; the correlation with high F2 virtual heights; and the dependence on the planetary magnetic indices a l l support this argument. An examination of the scintillation indices at times when known ionospheric disturbances were in progress lent further support: both the amplitude and the rate of scintillations increased at such times. It is well known that during abnormal iono-spheric conditions a l l regions are affected in some way or other. In the -50-light of this, i t is l i t t l e wonder that the height of the scintillating region could not be established by comparison with ionospheric data! Since both the amplitude and the rate of the fluctuations in-creased under abnormal ionospheric conditions, either set of indices can be used as a disturbance indicator. Both exhibited the same dependence on the angle of elevation of the radiating source which has already been discussed above, implying that the thickness of the ionosphere was a con-tributing factor in each case. Consequently, before any other relevant factors could be sought in the scintillation data, this known diurnal effect had to be removed. This was accomplished by applying a 2k hour running average to the data as discussed in the preceding section. The smoothed data on scintillation rates for the respective months of the year 1956 have been plotted in Figures 12 to 23. These dia-grams also show other indicators of ionospheric disturbance, and a consider-able agreement between the respective curves can be seen. At the present time the reader's attention i s directed to the f i r s t two curves, which represent the reception quality on the trans-Atlantic short wave circuit and the scintillation rates. A detailed comparison between these data made on a three hourly basis for a period of Ik months yielded a correlation coefficient of +O.27. This value is fairly low, but no higher coefficient was expected. The short wave quality data was obtained for the lower regions of the ionosphere and is more properly a measure of absorption conditions in the D region at times when solar corpuscles are entering the atmosphere. The scintillation data, on the other hand, probably measured conditions at more than one level of the ionosphere, and perhaps only the higher regions, so a close relationship between the two phenomena need not necessarily exist. 3.4 The Planetary Indices of Magnetic Activity The magnetic variations as recorded at a point on the earth are partly "regular" and partly "irregular". The regular variations depend on the location of the station and on the time of day and season of the year. On the other hand, the irregular variations are supposed to indicate -51-the influence of solar corpuscular radiation on the earth. In order to provide a record of these variations, and hence of the solar particles, a three-hour range index, K, was introduced by Bartels, Heck and Johnson^ On such a scale an observatory assigns one of the integers 0 - 9 to each three-hour interval according to the range of the variations on their records. A scale, defining the K values, is adopted for each station in such a way that 9 corresponds to the greatest variation observed at that place. The K index is not completely free from a l l local features and can serve only as a regional index. There i s , however, a need for a measure of the disturbance for the earth as a whole. Such a measure is provided by the three-hour index, Kp, called the geomagnetic planetary index. It i s based on the average of the K indices for 11 regions which have been freed from local effects as far as possible by an appropriate standardization process. The values of Kp for 1956 were subjected to a 2k hour smoothing process to make them compatible with the other data, and then plotted in Figures 12 to 23. A comparison between the various curves in the diagrams shows that most of the large disturbance features are common to a l l the curves, numerical correlations between these data were made and the following results obtained: (a) the correlation coefficient between trans-Atlantic quality and Kp was +0.66; (b) the correlation coefficient between trans-Atlantic quality and scintillation rates was +0.27; (c) the correlation coefficient between scintillation rates and Kp was +O.56; (d) the correlation coefficient between trans-Atlantic quality plus acintillation rates and Kp was +O.74. In the light of these observations, the following explanation is put forth: the planetary magnetic indices are indicative of the total motion of charges in the ionosphere which arise because of the incoming solar particles. If abnormalities in the electron densities occur in the lower ionosphere, these affect the magnetic indices in addition to being -52-observable in the trans-Atlantic quality figures, and hence the positive correlation coefficient of -tO. 66. The motions of gpnail regions of ioni-zation at higher levels of the ionosphere, which probably produce the observed scintillation indices, also contribute to the currents which determine the Kp values, and hence the correlation coefficient of +O.56. (That higher regions are here involved i s inferred from the low correlation coefficient between scintillation data and the short wave quality figures.) However, when the trans-Atlantic quality data and the scintillation rate data are combined and then compared to the Kp indices, a higher (+O.74) correlation coefficient results, indicating that the planetary magnetic indices pertain to the motion of the total charge at a l l heights in the ionosphere. 3.5 The Aurora Borealis A zone of maximum auroral occurrence exists in each of the northern and southern hemispheres as a circular belt at approximately 23° from the geomagnetic pole^*^. A portion of the northern zone lies over the central regions of Canada, running in a great arc from Labrador in the (kl) east to Alaska in the north-westv '. An observer near this auroral zone 6ees the Aurora Borealis almost every clear night. Further south the frequency with which overhead aurora are observed decreases, until in Southern Canada i t is a fairly rare occurrence. When such an occasion does arise for an observer in southern latitudes, i t usually corresponds to a time of intensive bombardment of the ionosphere by solar particles. These incoming particles not only dis-rupt high frequency communications and produce magnetic storms, but also initiate brilliant auroral displays extending to great distances from the auroral zone. Consequently i t is possible to use the latitude at which aurora occurs as a measure of ionospheric perturbation by corpuscular radiation. Recordings of the positions of the aurora were made by a group of workers at the University of Saskatchewan. They employed an automatic photomultiplier system to scan a strip of sky from the northern to the southern horizon^ 2\ Knowing the height of the most intense part of the display, approximately 100 km, the latitude was easily specified. Their -53-recorder was ideally located for this sort of measurement since the 100 km level at practically any latitude from the auroral zone to beyond the U.S.A. border is observable from Saskatoon. Occasions did arise, however, when the aurora extended to more southerly latitudes than could be seen from Saskatoon. In such cases the record indicated only aurora at the latitude corresponding to their limit of observability. The Saskatoon data have been presented here as a further measure of ionospheric disturbance. In the lower frame of Figures 12 to 23 the most southerly extent of the aurora on any one night has been plotted in terms of geographic latitude. In the diagrams the closed circles represent a weak display, while the open circles represent a more intense display. Since an optical recording system was used, no data was available on those nights when cloud interfered with the measurements. It is readily seen from the diagrams that these data correlate well with other measures of ionospheric disturbance, and may be used to support other indicators that specify when particles are entering the ionosphere. 3.6 Summary In the preceding sections, four different measures of the dis-turbed or abnormal ionosphere have been reviewed: the trans-Atlantic quality of a high frequency circuit, the rate of scintillation of a radio star, the geomagnetic activity, and the most southern extent of the Aurora Borealis. A graphical summary of the data for a 12 month period is presented in Figures 12 to 23. That these measures are not independent can be seen from the high correlation between the respective sets of data. Consequently, i t is believed that each one is a manifestation of corpus-cular radiation from the sun. In those cases when not a l l of the four indicators show the same degree of disturbance, variations in either the intensity or the velocity of the incoming stream of particles must be postulated which produce different effects at the various levels of the ionosphere. However, the large significant disturbances are portrayed reliably by any one of the four, and i t is only with the lesser disturbances that differences in the observed effects occur. A complete picture of a l l the disturbances cannot, -54-then, be obtained from just one measure, e.g. scintillations; inter-comparisons between the different measures are required. On the basis of the four sets of data, the different ionospheric storms for 1956 were graded in importance on a scale ranging from 1 to 6. On this scale, 1 was the smallest disturbance that could be clearly dis-tinguished on one set of data, while 6, the largest, showed up on a l l measures. This arbitrary classification was introduced to assist in the relation of ionospheric effects to the solar phenomena producing the particles. The l i s t of ionospheric storms is given in Chapter VI as part of Table I. CHAPTER IV THE DISTURBED SUN The sun is a mass of very hot gaseous material and i t s surface is continuously disturbed to some extent by eruptions and irregularities of various kinds. The author does not intend to discuss this random form of disturbance, which, by comparison, is of a mi nor nature. Rather, he will consider such major abnormalities as may produce particles having energies ,sufficient to overcome the solar gravitational forces and travel out to very great distances from the sun. In the previous section some means of determining the influx of such particles into the earth's atmo-sphere were presented; in the present section an attempt will be made to identify the particles in the space between the sun and the earth, and in particular, in the immediate vicinity of the sun following a visible dis-turbance on the photosphere. The particles which enter the earth's ionosphere have been linked to the sun by establishing similar cycles of recurrence for the respective phenomena. The astronomers have data on the sun's activity based on sun-spot counts for the past two centuries. From these an 11 year cycle of activity is clearly seen. The same 11 year cycles of magnetic and auroral activity are also observed* J / x '. Moreover, ionospheric disturbances tend to recur with a period of 27 days, which is just the period of revo-lution of the sun. However, attempts to- link individual ionospheric disturbances to respective sunspots have led to only moderate success. Within a period of from one to three days after the passage of a large sunspot group over the sun's central meridian, or the occurrence of a large flare, an ionospheric disturbance is sometimes, but not always, observed. At other times, disturbances are observed which cannot be thus related to sunspots, or to any other visible solar features for that matter. However, even these latter disturbances exhibit the 27 day re-currence cycle which links them to the sun. What solar features or abnormalities are then to be considered as probable sources of the emitted particles, and how can this emission process be observed on the sun? In the following sections such solar features as sunspots, flares, prominences, and M-regions will be considered, - 5 5 --56-and indications will be given of how the radio noise measurements made by the author relate to each of these phenomena. 4.1 The Detection of Solar Particles Measurements of the auroral spectrum have shown that high speed protons do enter the earth's atmosphere at times. Vegard^'^, Meinel^ 1^, G a r t l e i n ^ ^ and others have observed such incoming particles with speeds as high as 3000 km/sec, but have been unable to obtain evidence to link them directly to the sun. Moreover, their measurements show that the hydrogen spectrum lines are not a necessary condition, and many auroral displays do not exhibit these features. (47) Following a suggestion by Chapmanx ", several attempts have been made to detect matter in its flight from the sun. Spectroscopic studies in this respect have shown that Ca ions were found in the appropri-ate regions during a magnetic disturbance However, as only extremely weak spectral lines were observed the results were not conclusive. In a like manner observations made on the absorption of solar (49) (50) radio noise* and the orientation of comet's t a i l s x ^ ' supported the assumption that matter travels from the sun to the earth, but again the results were inconclusive. The problem of detecting particles in the immediate vicinity (51) of the sun is much simpler. Wild, et a l w , using a swept frequency receiver, have measured noise bursts which retain their character at successively decreasing frequencies with increasing time. Since the electron density decreases with increasing distance from the sun, and since the frequency of the radiation which can escape depends on the electron density, the study of noise bursts at a succession of frequencies provides a means of detecting sources which travel outward from the photo-sphere, and permits an estimate of their velocity. Wild's measurements did not show that a moving cloud of particles actually caused the radio noise originating at the respective levels, but this was inferred from the evidence. In order to determine the velocities of the travelling particles from such a set of measurements, the electron density and the magnetic -57-field in the various regions of the corona must be known. Values of the electron density have been deduced from optical and spectroscopic data by Jaeger and Westfold^ 2^-^, comparatively l i t t l e variation with time i s observed in these figures. On the other hand, the magnetic field is closely linked to sunspots and the active regions surrounding sunspot groups. In the field free case, the index of refraction i s given by • • ( - i f where e is the electronic charge, m the electron mass, K the electron density, and f the wave frequency. In the presence of a magnetic field the expression for the refractive index will be modified since both the magnitude and the direction of the field affect the propagation of radi-ation in an ionized medium. In order for radiation to reach the earth i t must originate no closer to the sun than a region where n i s zero in the absence of a mag-netic field. Furthermore, i f the generation process for a noise burst is a plasma oscillation as indicated by the measurements of Wild, et al., the radiation of frequency f originates only in the region where u = 0 for the frequencies f and f / 2 . These regions are indicated in Figure 29 for the various radio frequencies. Curve A i s for the field free case, while curve B gives an indication of the levels from which radiation could escape in the presence of a magnetic field of 36OO gauss. (This approximates the region above a very large sunspot.) Since i t is very difficult to specify the magnetic field at a l l times, one can deduce only approximate levels in the corona from which the radiation originates. Consequently, the velocities of the ejected particles cannot always be precisely specified. k.2 Noise Measurements at Times of Solar Flares The measurements of solar radio noise made by the author were at frequencies of 5OO mc and at 50 mc. Radiation at these frequencies which reached the earth originated in the solar corona, and consequently the occurrence of such radiation was taken as indicative of a travelling cloud of particles. The author felt particularly justified in this 10% I Co I 10 10^  10° 10' FREQUENCY (Mc/s) FIG. 29-THE ORIGIN OF R.F. NOISE AS A FUNCTION OF HEIGHT IN THE SOLAR ATMOSPHERE - 5 9 -conclusion in those cases when the noise followed a flare or some other unusual event on the sun's surface. For example, when a large flare occurred, an outburst of 5 0 0 mc radiation usually followed within a few minutes, which in turn was followed by a 5 0 mc burst of radiation within a short interval. The actual delay times depended on the velocity of the particles and on their direction of travel with respect to the sun-earth line. An example of such a sequence of events is shown in Fig. 3 0 for a flare which occurred on Feb. 8 , 1 9 5 7 - This flare, of importance 2 on a scale running from 1 - to 3 + , was observed to start at I55O U.T. and con-tinue until l 6 l 0 . The upper part of Fig. 3 0 shows the record from the 5 0 0 mc receiver for this period. It can be seen that a large outburst began at I55O and lasted for two minutes. The lower part of the diagram shows the record from the 5 0 mc receiver, also for the same period of time. A large outburst, similar to that at 5 0 0 mc, is seen to have started at 1 5 5 4 and lasted for approximately two minutes. This was later followed by a second burst at 1 6 0 9 lasting two minutes and then by a noise storm lasting for more than two hours. Since the generation process is a non-linear one, the second burst is undoubtedly due to a second harmonic from a s t i l l higher level as has already been established by Wild v^'. (The time separation is so short at 5 0 0 mc that the two bursts are not resolved.) The noise storm observed at 5 0 mc is absent from the 500 mc record and can probably be explained as matter falling back into the outer corona from the particle cloud. This latter feature will be discussed in a sub-sequent section. The eruptive process, however, was established from the time sequence of the start of the respective outbursts on the two frequencies. From Westfold's data as plotted in Fig. 2 9 the 5 0 0 mc radiation was known to have originated at about 1 0 km above the sun's surface, while the 5 0 mc radiation originated approximately 2 x 1 0 ^ km above the photo-sphere. Hence the particle cloud travelled almost 2 x 1 0 ^ km in four minutes with an average velocity near 7 5 O km/sec. This value is very approximate because the timing accuracy was not sufficiently great. A second estimate was obtained from the time difference between the two -60-- a -bursts on the 50 mc chart. Since the second burst was due to a harmonic which originated at a height corresponding to that of 25 mc radiation, i t was possible to compute a velocity for the particle cloud: this was 250 km/sec. Likewise, by considering the following noise storm to have been caused by matter falling out of the cloud back into the corona, i t was possible to deduce yet a third estimate of the i n i t i a l velocity of the particles. In this case velocities ranging from 100 to 450 km/sec were calculated. From velocities computed in these different manners for this and other similar cases, i t was apparent that a l l of the particles in an ejected cloud do not leave the sun with the same velocity, nor are a l l of them able to penetrate as far as the earth's orbit. Invariably, the greatest velocity was computed from the starting times of the out-bursts observed on the 500 and 50 mc receivers. The author found that observations of solar radio noise follow-ing a flare yield several different values for the velocity of the ejected particles. While the accuracy of the various measures was not too great, the discrepancies were too consistent to be explained in this manner. It was concluded, then, that a significant distribution of velocities existed in the cloud ofj.ejected particles, and that the timing of the commencement of the noise bursts on 500 and 50 mc yielded a measure of the most rapid particles. In such a fashion i t was possible to connect the pertinent solar phenomena and the ejected particle cloud to subsequent ionospheric disturbances, since only the most rapid particles could reach the earth. If the greatest velocity measured was sufficiently high, reliable fore-casts of ionospheric 'storms' were readily made. 4.3 Noise Storms which Follow a Flare Solar noise storms in general are associated with sunspots, and this feature will be discussed in more detail in a later section. There are, however, two types of 50 mc noise storms which follow the occurrence of a flare that cannot be related to sunspots. One type has a very abrupt commencement at approximately the time of the flare and may last for several hours or even days. The second type is illustrated in Figure 30: the noise storm starts some time after the short lived burst which accompanies the flare, and lasts for a number of hours or days. It is -62-suggested here that both of these types of storms are caused by matter falling out of the travelling cloud of particles and precipitating into the upper corona. This hypothesis i s supported by the observation that these types of noise storm are not too common at 500 mc; when they do occur they are fairly weak. It seems reasonable to suppose that not a l l of the particles in the ejected cloud have the same velocity, rather there i s probably a cer-tain velocity distribution about that deduced from the burst data. In consequence there i s a range of heights which these particles can attain under the gravitational influence of the sun; provided, of course, that the velocities are not great enough that the particles can escape entirely. Assuming radial motion, these distances are obtained from the relation: R' 2 / GMm . dR « l / 2 mV0 (8) R2 Ro where G is the gravitational constant, M is the mass of the sun, R is the distance of the particle from the centre of the sun, R Q i s the radius of the sun, R * is the maximum distance attained, m is the mass of the particle, and V 0 i s the i n i t i a l velocity with which the particle leaves the sun. This reduces to: R1 = 1 V (9) Ro 2GM By substituting appropriate values into this equation i t is possible to plot R' as a function of V Q for velocities less than the escape velocity. This graph is presented in Fig. 31. It may be seen from this graph that a particle requires an i n i t i a l velocity of 6I5 km/sec to Just reach the earth. If i t has any greater velocity than 620 km/sec i t will escape from the sun's gravita-tional field, whereas, i f i t has any less velocity than 615 km/sec i t will return to the sun before i t reaches the earth. (It is assumed, of course, that no forces other than gravitational forces are acting on the particle in a l l regions but the immediate vicinity of the photosphere.) If, now, -63-0 100 200 300 400 500 600 700 INITIAL VELOCITY (Km/sec) F I G . 3 1 - R A N G E OF S O L A R PARTICLES A S A FUNCTION OF THEIR INTIAL VELOCITY -6k-any reasonable velocity distribution exists in the ejected stream of particles a l l three situations may exist: some particles..may f a l l back to the sun before they get near the earth, some particles may enter the earth's atmosphere, and others may pass beyond the earth and escape entirely. In order to establish a time scale for those particles which return to the sun, equation ( 9 ) is re-written in the form: dR " I R' - R 0 and solvedfor t*, the time for a particle to rise from the sun's surface to a distance R* from the centre of the sun where its velocity.is zero. This yields: Note that t' is also the time required for the particle to f a l l from a height R' to the sun's surface, and as such has been evaluated and plotted in Figure 3 2 . Observations of the duration of a noise storm can be related to 2 t ' , the total time of flight, and may be used to provide an estimate of the distance that the slower particles travel from the sun. From this information an intelligent guess can be made as to whether or not other particles in the gaseous cloud which have only slightly greater velocities may be able to reach the earth. An example of such a situation is found in the large solar flare of January 1 6 , 1 9 5 5 . A large noise outburst was observed at 50 mc on that occasion, and on January 1 8 and 1 9 noise storms lasting most of the day-light hours were observed at this same frequency. The Mount Wilson records showed only a small sunspot group on the sun on January 1 8 , and no spot at a l l on the 1 9 t h , so these storms cannot be linked with large sunspots. On the assumption that these storms were caused by some of the slower particles from the ejected gaseous cloud as they re-entered the solar corona, i t is possible to deduce their velocity from the elapsed time since the burst on January 1 6 . It is seen in Figure 3 2 that this time delay signifies a range of approximately 1Q>' km, and from Fig. 3 1 one 1.0 10 10* I03 I04 TIME (Hrs) F I G . 3 2 - T I M E REQUIRED FOR PARTICLES STARTING AT REST TO R E A C H THE SUN A S A FUNCTION OF STARTING R A N G E -66-finds that this corresponds to an i n i t i a l velocity of approximately 600 km/sec. Prom the measurements made at the time of the flare of January 16, i t i s possible to derive an indication of the mean velocity of the ejected (5k) cloud. Covingtonmeasured the start of the outburst at 2800 mc as occurring at 2105 hours, while the author observed the 50 mc noise burst to start at 2110 hours. Using plausible values from Figure 29, this corresponds to a value of approximately 900 km/sec. The above example has shown how measurements of noise bursts at two frequencies permit one to calculate the velocities of the high speed solar particles. In addition, the duration of the solar noise storm that followed the flare permitted the calculation of the velocities of the slower particles in the stream. The values obtained were rather approxi-mate since the bursts were not timed precisely. However, that they were of the right order of?;magnitude can be shown by considering the subsequent geomagnetic storm. An indication of V 0 can be obtained from the delay between the flare and the commencement of disturbance conditions in the terrestrial ionosphere. The appropriate relation is derived by writing in place of equation (9): where V Q is the velocity at the sun's surface and V i s the velocity at a distance R from the centre of the sun. If this is written as dR dt 1/2 [ c * a - ^ > + ¥ ] ' ^ a solution can be found for t^ in terms of V0, where t^ is the time required for a particle, starting with velocity V Q at a distance R Q from the sun's centre, to reach the earth at a range R ^ . This i s : ^ " p ^3 l 0 S (VQ - A) (V x + A) ( U ) -67-where A = (V 0 2 - 2 « ) 1 / 2 a^d V l . ( A2 + <m )l/2 . By substituting the appropriate constants into this expression i t is possible to plot a graph of t-^  in terms of VQ. This is shown in Figure 33. The ionospheric and auroral effects which followed the flare of January 16 have already been reported by the author^5)^ These dis-turbances as well as a world-wide magnetic disturbance were observed to start approximately 2k hours after the flare. From Fig. 33 i t is seen that such a time delay fixes the starting velocity for these high speed particles at about 1700 km/sec. The ionospheric effects lasted for some two days, which again indicates an appreciable range of velocities for the incoming particles. The available evidence has been examined to try to determine the velocities with which particles of matter left the sun in the case of the importance 3 flare of January 16. While the accuracy with which such velocities can be determined is not as great as one might like, a l l the results indicate a velocity range of from 600 to 1700 km/sec. Moreover, the velocity of the slower particles as deduced from the long duration of the ionospheric storm was comparable to that deduced from the fall-back of particles into the sun. The solar noise recordings for a l l the flares listed in Table I were examined in a manner similar to the above example. Whenever possible the particle velocities were computed, and in a considerable number of cases a range of velocities was found. The computed values then served to indicate whether or not a significant number of the particles in the ejected stream would have sufficient energy to reach the earth and produce ionospheric disturbances. This information is given in Table I as the probability that a disturbance would be produced in each case. k.k Noise Storms and Sunspots Whether or not radio noise can originate at a point in the solar atmosphere depends on the electron density and the strength of the mag-netic field in that region. This is the case for synchrotron radiation I06-0 1.0 T 7o« 10 I02 I03 SUN-EARTH TRAVEL TIME (Hrs) FIG. 3 3 - T I M E REQUIRED FOR S O L A R PARTICLES TO R E A C H T H E E A R T H A S A FUNCTION OF INITIAL V E L O C I T Y -69-as discussed by Kruse, et a l ^ - ^ , and also for the plasma oscillations postulated by Wild, et a l * - ^ ' . In the previous section i t was assumed that noise outbursts were produced by the latter process, mainly because (66) Payne-Scott and L i t t l e v ' observed that the radiation was randomly polarized. On the other hand, they found that noise storms were frequent-ly circularly polarized, and hence are probably produced by a synchrotron process. Furthermore, the storms have considerably longer durations, and, unless they have very abrupt commencements, time sequences cannot be determined to designate the motion of the radiation source. Several types of noise storms which could be linked to solar flares have already been discussed. These storms presumably result from matter entering the solar corona from outside the sun. There i s a differ-ent class into which the majority of the 50 mc storms f a l l ; these are the (57) ones which can be linked to sunspots. Payne-Scott and L i t t l e w , / have shown that the source of the 97 mc storms observed by them lay in the corona directly over large sunspots. They suggest that the condition for the occurrence of a noise storm is the existence of a sufficiently large sunspot, that i s , one having an area in excess of 400 millionths of the solar disc. Measurements made by the author at 50 mc bear this out, but suggest that the limiting area for an individual spot may be closer to (58) 300 millionths of the disc a r e a v ^ . The difference in frequency may explain this change in value. However, just because R.F. radiation is generated over a sun-spot does not necessarily mean that particles of matter are being ejected at that time, even though this may seem to be a reasonable assumption. The noise storm data only indicate that, in the presence of the magnetic field of the sunspot, the ionization in the corona (or that ejected from the spot into the corona) produces the R.F. noise. However, history has shown that the appearance of a large sunspot on the sun's central meridian is often followed by terrestrial magnetic and ionospheric disturbances. Consequently, the very large sunspots must be considered to produce particles which can escape from the sun. It is suggested here that the large noise storms which accompany the very large sunspots are indicators of this type of emission process. Probably particles are continually "boiling off 1 1 with velocities which in general are less than that required -70-to reach the earth, and only in the case of large sunspots are there sufficient particles with velocities great enough to produce terrestrial effects. On such a hypothesis, the particles would contribute to the noise storm on each of their passages through the corona. The 50 mc noise storms observed by the author have been examined in this light. It was found that more often than not those sunspot groups whose central meridian passage could be related to subsequent ionospheric disturbances were ones whose area exceeded 1500 millionths of the solar disp,which also had a single spot of area exceeding 690 millionths of the disc, and which were accompanied by an intense noise storm, in excess of 20 pabs , during most of the l i f e of the group. While this seemed to provide a working criterion for predicting ionospheric effects, there is no way at present fcd investigate whether the author's hypothesis on the generation process is valid. So far only noise storms at 50 mc have been considered. At 500 mc similar storms were observed except that the intensity increase over the quiet sun radiation was not nearly as great as at the lower fre-quency. The sunspot dependence was considerably more marked at 500 mc in that the total radiation depended strongly on the total area of sunspots on the disc and not simply on the largest spot as at 50 mc. These differences probably arise because the region of radiation generation is at a different distance from the spots in the two cases. Assuming a synchrotron generation process, and knowing that the magnetic field falls off at least as the cube of the distance from a sunspot, one would expect the contribution from the smaller spots to be negligible at distances of about 2 x 10^ km, and only the very large spots determine the 50 mc radiation source. On the other hand, at distances of about 10 km a l l the spots provide some contribution to the 500 mc source. Those 50 mc noise storms which follow a flare have been explained as due to the fall-back of matter from the particle cloud. However, the corresponding storm at 500 mc is not usually observed. At present the *This terminology has been introduced here for convenience. Bab is a concise way of writing "power per unit area per unit band width' and 1 pab is defined to be equal to 10~ 2 2 watts/m2/cps. -71-explanation for this is not known, but i t may come to light when more in-formation on the generation processes in the solar atmosphere is available. 4.5 Solar Prominences and Filaments Prominences are normally observed in hydrogen light on the limb of the sun as bright streamers of gaseous matter. When they occur on the disc, the light from the photosphere behind them is absorbed, and they therefore appear as dark lines. These are the so-called filaments. Despite the difference in name, there appears to be no difference in the phenomena. The prominence matter exhibits considerable motion, rising into the corona and then falling back to the sun. Optical measurements show that the velocities involved are usually no more than 200 km/sec. Reference to Fig. 31 shows that the height reached by such particles is not too great. Exceptional cases have, however, been observed when the velocities have exceeded 700 km/sec. These are the so-called surge prominences, or eruptive prominences. In such cases the gaseous material would, of course, reach heights exceeding the sun-earth distance. In fact, particles from such prominences may indeed reach the earth i f there is any lateral dispersion of the particle stream. On the limb a surge prominence is a very conspicuous phenomenon, but on the disc i t is observed as a sudden disappearance (disparition (eg) brusque) of a filament*-^'. If this occurs in the central portion of the disc, the probability of the high speed particles reaching the earth and producing ionospheric effects is quite good. In consequence, the author has focused extra attention on the disappearance of large filaments in an effort to establish them as sources of Earth-reaching particles. The situation is somewhat complicated in that not every disappearing filament signifies an eruptive prominence. Studies^) have shown that disappear-ances also happen in two other ways: prominences may either shrink without visible cause and disappear, or they may flow into the chromo-sphere over sunspots in well defined streamers. The radio noise recordings at 50 mc have been examined in an effort to clarify the situation. It has been found that noise storms do -72-frequently occur after disappearing filaments and surge prominences. These storms were usually rather spiky in nature. However, the 500 mc recordings were quite dissimilar and no estimate of the velocity of the expelled matter was possible from measurements at the two frequencies. On the other hand, an attempt to use the duration of the noise storm to determine the height which the particles attained and from this determine an i n i t i a l velocity led to inconclusive results because of the frequent presence of other solar phenomena which might have caused the noise storm. k.6 M-Regions As the result of a study of magnetic storms which recurred at intervals of 27 days, B a r t e l s * ^ postulated that invisible M-regions on the sun were the responsible agency. These were supposed to be some sort of restricted area which produced the geomagnetic disturbances, presumably by particle radiation. While at times he was able to associate M-regions with sunspots or some other solar phenomena, in general such identification was not possible. Whether or not M-regions do exist s t i l l seems to be an open question. Evidence persists that many ionospheric and magnetic distur-bances which cannot be associated with sunspots, flares, or prominences and filaments repeat at 27 day intervals. The author has noted, however, that in a number of such cases a 50 mc noise storm has preceded the dis-turbance by approximately three days. Such storms appeared to have a short l i f e , (of the order of one or two days at most) and to be un-connected with any other obvious solar phenomena. On the basis of such evidence i t seems possible that M-regions do exist which are invisible to the optical observer, but are detectable by means of a radio receiver. The latter condition must arise because of particles travelling through the corona on their journey to the earth. More evidence is required before the identification is firmly established. 4.7 Summary A number of solar phenomena have been considered as sources of particles which travel from the sun to the earth and initiate the iono-spheric disturbances. For many years there has been speculation about - 7 3 -Buch sources on the sun, but observations in the optical spectrum range were insufficient to resolve the issue. In the foregoing sections the author has shown that R.F. noise measurements can be made use of in many cases to specify the times when such particles are leaving the sun. Indi-cations have been given of the types of information available in the noise recordings during certain solar conditions, and in Chapter VI specific instances of particle emission as determined by radio and optical evidence will be presented. Just as in the case of the visual observations, the radio obser-vations of the sun contain much more data than can as yet be interpreted. This is to be expected since many of the solar parameters are not precise-ly known. In the present analysis the author has.ignored a number of dis-tinctive features in the visual data, and also many recognizable events in the noise recordings. In particular, such things as calcium plages, coronal streamers, small sunspots, and sub-flares have not been linked directly with the noise measurements, even though the possibility of some connection has not been ruled out. Attention has been focused on the large and unusual optical phenomena, so that a clear distinction could be made between such conditions and the more normal ones. A l l such instances occasioned careful examination of the noise recordings at or near the same time, and conclusions were drawn from the combined data whenever possible. CHAPTER V SOLAR PARTICLES Numerous accounts may be found in the literature of investi-gations into the velocities of travel of the solar particles. One such example is the work of Newton*^^^, who examined the magnetic indices following times of large solar flares. From data averaged for a great many flares, he was able to show that his results could be explained by particles entering the earth's atmosphere approximately two days after leaving the sun. However, the data on the smaller flares when averaged (64) showed no consistent results. Other evidence by Newton and Jackson* ' would suggest that a sudden commencement magnetic storm occurs about one day after a large flare. Such apparent contradictions seem to arise because significant differences in velocity exist between individual cases, and not a l l of the solar particles travel with the same or nearly the same velocity. Rather, the velocity is probably determined by the energy avail-able in the solar phenomena that produced the radiated particles. As a demonstration of such differences in velocity, the magnetic indices (Kp) following the importance 3 flares of Jan. 16, 1955, Feb. 23, 1956 have been plotted in Fig. 34. A 2k hour smoothing was applied to these data to remove minor random variations. It can be seen that the two curves are quite different in shape: the magnetic storm followed the flare much sooner in one case than in the other. Knowing the delay times, estimates of the velocities of the particles in the two cases can be obtained from Fig. 33. Those particles which contributed to the nifirinifl. in Fig. 34 l e f t the sun with velocities of 2000 km/sec and 900 km/sec respectively. Not only does Fig. 34 indicate that differences in particle velocity exist from one flare to the next, but the durations of the mag-netic storms indicate that an appreciable velocity spread exists for the particles of any one particular flare. This latter conclusion has already been discussed in a previous section. However, just what the particular distribution is in the case-of a certain solar event: lias not been speci-fied. Some of the available evidence will be examined in order to determine a general shape for such a distribution curve, and then a prob-able distribution win be considered to see i f i t leads to results similar to the actual observations. 7 H i i l 1 1 i 1 -I 0 I 2 3 4 5 TIME IN DAYS MEASURED FROM FLARE F I G . 3 4 - M A G N E T I C INDICES FOLLOWING TWO I M P O R T A N C E 3 F L A R E S -76-5 . 1 The Flare of February 8, 1 9 5 7 In Chapter IV an example of noise bursts on two frequencies which accompanied a solar flare was presented, and the method of com-puting particle velocities from the noise data was illustrated. In that case the findings were as follows: the time difference between the lead-ing edge of the bursts yielded a velocity of 7 5 0 km/sec for the most rapid particles. The maximum intensity of the ensuing noise storm on 50 mc occurred approximately 7 5 minutes after the flare, implying that the greatest number of returning particles were entering the corona at that time. Making use of Figures 3 1 and 3 2 i t was found that these particles had an i n i t i a l velocity of approximately 3 0 0 km/sec. The noise storm started shortly after the flare and gradually built up to a maximum and then died out gradually. The total duration was approximately two and one-half hours on the records. Those particles that contributed to the early part of the storm had velocities near 1 0 0 km/sec, and those contributing to the final phase of the storm had velocities computed at approximately 4 5 O km/sec. The above values are rather approximate and serve only to show what the general shape of the velocity distribution curve is like. Before a more precise determination could be made, one would require detailed information en the noise generation process at a l l heights in the corona. The curve shape deduced from these calculations is approximately that of a Maxwellian or a Rayleigh distribution. 5 . 2 Velocity Distribution The deductions in the previous section were quite approximate and served only to indicate the general form of the velocity distribution. Several natural distributions will now be examined and the form that the experimental observations should take in each case will be indicated. For the purpose of this discussion i t is assumed that gravitational forces and no others act on the particles after they have left the immediate vicinity of the photosphere. (Radiation pressure, for instance, was shown by Johnson to produce a negligible effect on protons, and mag-netic forces must be quite small since most noise bursts are known to be randomly p o l a r i z e d . ) -77-The assumption is made here that, following an eruptive process on the sun, the particles travel with an i n i t i a l velocity distribution given by: 2 2 KdVQ = KX V Q n e-V° dV0 (15) In this expression V Q is the velocity with which the particles leave the photosphere, and Kg are normalizing constants and n is a positive integer, n = 0 yields a Gaussian distribution, n = 1 yields a Rayleigh distribution, while n = 2 provides a Maxwellian distribution. Since the primary interest is in the high velocity portion of the curve where the shape is predominately determined by the exponential term, the value selected for n is of secondary importance only. In the following dis-cussion n has been chosen as equal to 2, but the results would be only slightly altered for n = 1 . In Fig. 3 5 a plot of 2 2 HdV0 = KX V 0 2 e" V o / K 2 dV0 ( 1 6 ) as a function of V 0 has been presented for three different values of Kg. Since only those particles with velocity greater than Vg = 6 1 5 km/sec can reach the earth, the portions of the curves to the lef t of this value of V D are of l i t t l e interest here. A l l those particles with V Q < Vg will f a l l back into the sun without entering the earth's atmo-sphere. It can be readily seen that the shape of those portions of the curves that l i e to the right of V Q = Vj; i s determined mainly by the exponential term in equation ( 1 6 ) , for the range of Kg's selected. For Kg much larger, this argument f a i l s . The time required for particles to travel from the sun to the earth has been expressed in equation (14) as a function of V0. This equation can be replaced by the approximate expression, *1 - R l (17) <V02 - Sgj ) 1 / 2 with negligible error i f values of V Q > 65O km/sec are considered. In this equation R;j_ is the sun-earth distance, R Q is the radius of the sun, -78-Km/sec. FIG. 3 5 - A S S U M E D M A X W E L L I A N V E L O C I T Y DISTRIBUTIONS FOR T H E S O L A R P A R T I C L E S -79-G is the gravitational constant and M is the mass of the sun. Now i t is possible to combine equations (16) and (17) and replace the velocity distribution of particles with a distribution of delay times with which the particles reach the earth. That i s : Ni dt x = Ki V 0 2 e' V o Z * 2 fjo dt-L (18) dt x P 2 /-p 2 a/2 1 / R L 4. 2GM^  dt x (19) Equation (19) is the distribution of particles arriving at the earth with delay times t^. However, any considerations of ionospheric disturbances should probably be concerned with the energy introduced into the ionosphere by the particles, rather than with numbers of particles entering the iono-sphere. Consequently, equation (19) should be re-written in the form of an energy distribution as a function of sun-earth delay time, t^. To a sufficient degree of accuracy for the present purpose, this becomes: K'*lk / R l 2 2GM\1/2 ' , Edti = ±- (-±- + 2 2 ) e K2 V t l * dti (20) t i 5 Uj 2 V In this expression E is the energy associated with those particles reach-ing the earth from the sun with travel times between tj_ and t^ + dtj_, and K' is a constant. This equation has been evaluated and plotted in Fig. 36 for three different values of Kg. The vertical scale in this graph is in arbitrary units (multiples of K'),. since the shape of the curves, rather than the absolute magnitude, is of interest here. As there were few particles having high velocities, the distri-bution curves in Fig. 36 show a small number at short delay times. There are many more particles with low velocities, but these are spread out over the increased delay interval, and the distribution curves again show a pm^ii number at long delay times. In between these two extremes a maximum exists whose position and relative magnitude depends on the value chosen -80-TIME (days) F I G . 3 6 - T H E DISTRIBUTION OF E N E R G Y WHICH T H E S O L A R P A R T I C L E S INTRODUCE INTO T H E IONOSPHERE AS A FUNCTION OF S U N - E A R T H T R A V E L T IME -81-for Kg. When taking observations of the ionosphere at the time of a dis-turbance, one is presumably tracing out such a distribution curve of particle energies at the various delay times. A further point of interest is the slope of the energy curve at short delay times. In consequence of the steep rise of the curve one would expect ionospheric effects to start abruptly. This is in agreement with observations of sudden commencement magnetic storms. On the basis of the preceding argument i t is obvious that the maximum of the energy-delay time distribution curve for corpuscles enter-ing the ionosphere does not correspond to the maximum of the velocity dis-tribution curve for the particles as they leave the sun. This is the case irrespective of the nature of the velocity distribution curve, within fairly broad limits. The question then arises: what part of the velocity distribution curve does one measure when timing the start of radio noise outbursts at two frequencies? Unfortunately this question cannot be completely answered until more infoimation is available on the generation process. However, i t seems probable that a consideration of the energies associated with the different particles will furnish the answer. 5.3 Particle Energies in the Corona The velocity distribution given in equation (16) for particles as they leave the sun can be re-written in the form of an energy distri-bution, since E dV0 = K3(NdVQ) V Q 2 (21) where E is the energy in the velocity range V 0 to V Q + dV0 and K3 is a constant depending on the mass of the particles. This energy distribu-tion becomes: E dVQ - % V 0 4 e"V° / K 2 dVo (22) Actually the time distribution of the energy at a level where the burst of noise originates is required. The time taken by a particle having a velocity V 0 to travel to a distance D from the sun's surface is given approximately by: t = D/VQ (23) -82-By combining equations (22) and (23) one obtains the time distribution of the energy, namely: Ejdt = *5 e-D 2/K2 2t 2 d t ( 2 J^ ) t In this expression Ej_ i s the energy arriving at the level of interest in the corona between the times t and t + dt. K^  i s a constant which will not be evaluated since the shape of the distribution curve only is of interest. Equation (24) has been plotted in Figure 37 for D = 2 x 10^ km and for three different values of Kg. On an energy basis only, and with-out considering the details of the generation process, i t is expected that a cloud of particles should generate a noise burst whose shape is that of the curves in Figure 37. To be sure, the finite l i f e of the original eruption will broaden the bursts somewhat from that shown since equation (24) was based on the assumption that a l l the particles le f t the sun at the same time. In any event, the very fast particles, that is , the most ener-getic ones, will s t i l l contribute to the leading edge of this burst, and i t is this leading edge which is used in the method of measuring velocities by timing the bursts. In connection, with the timing of noise bursts at two frequencies, a consideration of equation (24) for two different values of D should show the type of results expected. This was done by assuming that a cloud of particles originated on the photosphere in an instant of time, and that i t had the velocity distribution Ndv 0 - K l v D 2 e - < v ° A « » 2 av 0 (aj) As may be seen in Figure 35, the maximum of this distribution occurs at a velocity of 400 km/sec. The time distribution of the energy was com-4 5 / puted for the two cases D = 10 km and D = 2 x 10^ km (which correspond to radio frequencies of 500 mc and 50 mc respectively), and plotted in Figure 38. A comparison of these curves with the shape of actual noise bursts from the recordings showed much similarity. However, the 500 mc burst was usually not so much sharper than the 50 mc burst as is indi-cated in Figure 38. This difference would not be as noticeable i f - 8 3 -9-| 0 200 400 600 800 TIME (sec.) FIG. 3 7 - D I S T R I B U T I O N C U R V E S FOR T H E E N E R G Y INTRODUCED INTO A REGION OF T H E CORONA AT 2 x l 0 5 Km HEIGHT BY A CLOUD OF S O L A R P A R T I C L E S AS A FUNCTION OF THE T R A V E L TIME OF T H E P A R T I C L E S 8H 7H 6H UJ < o en >• < or 00 DC < T3 l±J 5H 4H -eh-D = I04 Km A 0 I I \ ^-D = 2xlO°Km I I I I \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ i I I 200 400 600 TIME (sec.) 800 1000 F I G . 3 8 - D I S T R I B U T I O N C U R V E S FOR T H E E N E R G Y INTRODUCED INTO TWO REGIONS OF T H E CORONA BY A CLOUD O F S O L A R P A R T I C L E S A S A FUNCTION OF T H E T R A V E L T I M E OF T H E P A R T I C L E S -85-allowance had been made in the calculations for the finite duration of the solar eruption. If, now, in this example, one tries to take timing measurements from the leading edge of the burst, different results are forthcoming i f measurements are made at different heights on the recorded burst. If the 20 per cent points on the curves are chosen, a velocity of 1250 km/sec results, while i f the 50 per cent points are taken, the resulting velocity is 1050 km/sec. It should be noted, however, that the velocities thus computed are much higher than .400 km/sec, which was the maximum of the velocity distribution curve. It is these very fast particles which can penetrate to dis-tances of the earth's orbit and produce the disturbed ionospheric con-ditions. The timing of the noise bursts on successive frequencies pro-vides a fairly reliable means of specifying the velocities of such particles. CHAPTER VI THE IDENTIFICATION OF SOLAR PHENOMENA WITH IONOSPHERIC STORMS It was shown in Chapter III that the times at which solar particles enter the earth's atmosphere could be specified by noting ab-normal ionospheric conditions. The present chapter contains a l i s t of a l l such ionospheric disturbances, or storms, for the year 1956. These have been divided into six different classes on an arbitrary intensity basis, ranging from 1, for the smallest disturbance recognizable on the records, to 6, for the most intense storms. The availableooptical solar data were examined for a number of days preceding each of the above mentioned storms. Any outstanding event on the sun occurring from one to three days before the disturbance was selected as a probable source of the Earth-reaching corpuscles. The radio noise data, when available, was then employed to confirm whether or not particles had been ejected from the sun at that time, whether they had reached the high levels of the corona, and whether or not their velocity-was of the right order for a significant number to reach the earth. Such a method of relating solar and terrestrial events was, of course, not infallible, but the likelihood of error was quite low in the case of the very large disturbances. The solar phenomena responsible for the large storms were so outstanding as to make errors in assignment negligible. A number of ionospheric storms were observed when no outstand-ing solar disturbance was noted in the specified time. This is thought to result from insufficient visual observations since such occasions occurred more frequently in the early months of the year. More complete data were available for the later months. 6.1 Solar and Ionospheric Disturbances for 19!>6 A l l the significant solar events for the year were studied, and those which could be related to terrestrial effects are listed in Table I. The probability that an event would produce an ionospheric storm was estimated in each case, and these values (on a scale ranging from 0 to 9) are likewise presented in the table. In classifying the solar phenomena' on such a probability basis, consideration was given to the region of -86--87-occurrence on the sun and to a l l the available information from the solar noise recordings. For the detailed criteria the reader is referred to subsequent sections. A l l times given in the table are Universal Time. The timing of the ionospheric storms was rather approximate (decimal fractions of a day were used) since the starting times, except in the case of a sudden com-mencement storm, were fairly indefinite. The final column of the table is an indication of the time re-quired for the most rapid particles to travel from the sun to the earth. In each case the time is taken from the start of the solar event to the start of the geomagnetic storm. In several instances the association of solar and ionospheric events was somewhat arbitrary and possible errors may have occurred, particularly where several noteworthy solar phenomena happened in a short time interval. Whenever possible, however, the identification was based on the solar noise data, thus reducing the likelihood of ambiguity. The real problem in this regard was the lack of continuous noise data on the sun. The observations made by the author were for a fraction of the day only and observations for the rest of the day, made elsewhere, did not suffice for the present purpose. This difficulty could be overcome i f more observers in various countries around the globe were to make noise recordings at the longer wavelengths. -88-TABLE I Outstanding Solar Phenomena and Associated Ionospheric Disturbances for 1956 Date Time Solar Prob. Associ- VDate Delay (U.T.) Event of ated of (days) Storm Storm Storm 2 1.6- 2.2 - - - - 2 6.5- 6.9 -Jan 5 0940-6,0419 D. Filament 4 1 9.4- 9.8 -8 O3O8-O9OO D. Filament 5 4 10.3-11.9 2.2 - - - - 3 12.7-13.2 -_ _ 2 I4'l2-14.5 14 0830-15,1308 D. Filament 4 4 18.0-19.8 -19 O323-O458 D. Filament 3 3 21.8-22.5 2.7 20 0900 CM.P. Spot 5 4 23.7-25.3 3.3 - - - - 3 27.4-28.9 -Feb 7 0308-1040 D. Filament 5 1 11.0-11.3 3.9 9 0925-1105 S. Prominence 4 3 11.7-12.5 2.3 10 2110-2140 Flare 3 4 1 I3.2-I3.6 2.3 Ik 0539-0709 Flare 3 3 2 16.1-16.8 1.9 16 1805-2015 Flare 2 + 5 2 19.0-19.7 2.2 19 1430-1657 Flare 1 + 2 2 22.0-2217 2.4 23 0334-0414 Flare 3 6 5 25.1-26.1 1.9 - - - 2 27.8-28.3 -- - - - 3 28.8- 1.5 -29 2220-2309 Flare 3 6 5 3.0- 4.6 2.1 Mar 2 1220-1314 Flare 3 3 1 5.7- 6.4 3-2 _ - 3 10.8-11.5 -_ - - 2 14.2-14.7 -16 1730 Flare 1 + 1 1 19.9-20.2 3.2 19 0908-20,0943 D. Filament 4 4 21.0-23.4 1.7 _ 3 24.3-26.7 -_ - - - 3 28.7-30.0 -27 0900-1453 S. Prominence 4 2 30.8-31.3 3.* 30 0630-1350 D. Filament 3 2 2.4- 3.k 3.1 - - - 1 6.5- 7.4 -Apr 8 IO59-I3OO Flare 2 2 1 IO.3-IO.7 1.9 9 0940-1100 Flare 3 3 1 11.6-12.0 2.2 14 0847-1306 D. Filament 2 2 16.6-17.3 2.3 16 0736-0752 Flare 2 2 1 19.1-19.4 2.8 18 1313-1530 Flare 2 6 5 21.4-23.2 2.9 -89-Apr 24 O63O-0755 S . Prominence 7 6 26.2-29.6 1.9 27 2050-2150 Flare 1 3 2 30.0- 1.6 2.1 May 3 0915 Flare 2 1 2 5-2- 5.9 1.8 9 0052-0902 S . Prominence 4 4 11.9-13.6 2.9 11 I8IO-1852 Flare 2 5 6 15.O-17.7 3.2 17 2230-2404 Flare 3 5 4 20.3-21.3 2.4 20 0835-1535 D. Filament 4 2 22.3-22.9 2.0 21 1142-1217 Flare 2 4 6 23.4-26.1 1.9 26 0608 S. Prominence 1 1 27.9-28.3 1.7 27 0511 S. Prominence 3 3 29.6-30.7 2.4 30 0859-1200 D. Filament 4 2 1.0- 1.9 1.7 June k 0950-0954 S. Prominence 5 2 6.4- 6.7 2.0 5 0757-1054 S . Prominence 3 1 8.8- 9.4 3-5 8 0210-1207 D. Filament 4 2 10.6-11.6 2.5 12 0324 S . Prominence 4 2 14.8-16.0 2.7 _ 1 17.7-18.1 _ ' 20 234I-2I.O2O7 D. Filament 5 2 23.7-24.2 2.7 22 1525-1645 Flare 3 5 3 24.5-25.9 1.9 2k 0904-1131 S . Prominence 3 2 26.5-27.5 2.1 27 0003-0803 D. Filament 2 2 29.3-30.9 2.3 30 I2O4-I52O S. Prominence 2 2 3.0- 3.6 2.5 July 3 1418-1430 Flare 1 2 1 5.3- 5.6 1.7 5 0429-0518 Flare 2 2 2 8.0- 8.4 2.8 6 1155-1235 Flare 2 2 2 8.9- 9.4 2.4 8 0659-2336 S . Prominence 3 2 10-8-11.3 2.5 11 0818-1244 D. Filament 4 3 13.6-14.2 2.3 16 1526-163^ D. Filament 4 2 19.6-20.2 3.0 20 1050-1223 D. Filament 3 3 23.5-24.6 3.1 22 1700-2332 D. Filament 4 4 25.2-27.5 2.5 26 0333-1234 D. Filament 2 2 28.4-29.0 2.3 26 I6O8-I705 Flare 2 2 2 29.2-29.8 2.5 27 1258 S . Prominence 2 1 3O.O-30.4 2.5 29 O837-O843 Flare 2 2 1 31.5-31.6 2.2 Aug 5 1134-1141 Flare 2 4 2 8 .7- 8.9 3.2 7 1635-2313 D. Filament 3 3 9.4-10.2 1.7 8 1140-1322 Flare 2 + 2 3 11.0-11.9 2.5 9 1227-I355 D. Filament 5 2 12.2-13.O 2.7 15 0843-1125 D. Filament 2 2 I7.O-I7.6 1.7 19 O735-O8OO Flare 2 2 2 21.4-22.0 2.1 21 19^5-2200 Flare 2 5 5 23.4-26.4 1.6 26 1510-1520 S . Prominence 0 1 29.4-29.6 2.8 28 I520-I63O- Flare 2 - 3 2 3 l . k - 3 l . 9 2.8 31 1225-I63O Flare 3 6 4 2.0- 2.8 1.5 -90-Sept 1 1530-1710 S. Prominence 4 0805-1228 D. Filament 5 0236 S. Prominence 7 O332 S. Prominence 9 0712 S. Prominence 10 0813-1005 Flare 2 13 0210-1127 D. Filament 17 2013-2120 Flare 3 21 0804 S. Prominence 24 1348-1413 S. Prominence 26 0212 S. Prominence 29 1347-30,1330 S. Prominence Oct 1 2320-2750 D. Filament 3 0430 S. Prominence 5 0615-0645 Flare 2 7 1355-1^30 S. Prominence 9 0205 S. Prominence 17 1210-1308 D. Filament 20 0903-1325 D. Filament 23 0712-1455 D. Filament 25 0732-0948 D. Filament 27 0823-0830 Flare 2 31 1504 S. Prominence Nov k 0513 S. Prominence 7 1104-1354 Flare 3 11 0900 C.M.P. Spot 13 1430-1555 Flare 2 Ik 1045-1500 Flare 3 17 1427-18,0856 D. Filament 20 1002-1318 Flare 2+ 23 2100 C.M.P. Spot 25 1435-26,0737 D. Filament 30 1835-1915 Flare 2 Dec 6 I6OO-I635 Flare 2 10 0934-1053 Flare 2 22 0955-1058 Flare 3 25 2150-2215 Flare 2 27 1647-1805 Flare 2 4 3 3.0- 3-7 1.4 3 2 6.1- 6.7 1.8 3 3 8.2- 8.9 3.1 3 2 9.2- 9.9 2.1 2 1 11.1-11.5 1.8 2 2 12.8-13.4 2.5 3 1 16.5-17.0 3 A 8 4 20.1-22.9 2.3 2 2 23.O-23.5 1.7 3 2 25.9-26.5 1.4 1 1 28.1-28.3 2.0 2 3 1.9- 4.0 2.6 3 2 5.2- 5.5 3.1 1 1 6.3- 6.6 3.1 2 2 7.6- 7.9 2.3 3 2 8.8- 9.5 1.3 2 1 11.1-11.5 2.0 3 4 20.1-22.2 2.6 3 3 23.O-23.6 2.6 6 4 26.3-27.7 3.0 4 3 28.1-28.7 2.8 2 1 30.4-30.8 3.1 2 2 2.9- 3.6 2.3 1 1 6.1- 7.1 1.9 4 4 9.7-13.2 2.2 I } 5 14.0-16.5 3 2 17.5-18.4 3.1 2 2 20.5-22.2 2.9 1 3 22.5-23.7 2.1 4 2 25.5-26.I 1.6 2 2 27.8-28.4 2.2 2 1 2.0- 2.9 1.7 2 1 8.2- 8.6 1.5 - 2 10.0-11.0 -2 2 12.7-13.7 2,3 3 2 25.4-26.5 3.0 2 2 27.6-29.I 1.7 2 1 30.2-30.5 2.5 -91-6.2 Flares as Sources of the Corpuscles A glance at Table I shows that solar flares were responsible for about half of the ionospheric storms. 'Since the energy associated with this phenomenon is quite high, variously estimated at 10 2 ^ or 10 2^ ergs for a typical f l a r e ^ \ this conclusion is not surprising. Only a portion of this available energy contributes to the high speed particle cloud, the rest goes to generate other forms of radiation. These include the line radiations, particularly of hydrogen, the continuous emission in the visible and ultra-violet, extending into the x-ray region, cosmic rays, and the radio frequency radiation. It is quite probable that the division of energy between the various emissions varies from flare to flare, and that each flare must be treated as an individual case as far as terrestrial effects due to particle radiation are concerned. This individuality is exemplified in Figure 39 for several representative flares. In Fig. 39 values of Kp have been plotted for a five day period following each of a number of flares chosen from Table I. The effect of the solar corpuscles on the ionosphere can be seen in this diagram: in most cases the principle effects occurred from two to three days after the flare. This was the case for the majority of the flares in Table I, irrespective of magnitude. However, the smaller flares produced smaller magnetic disturbances than did the large flares, and very few flares of importance 1 are found in the table. Moreover, the time delay was not identical for each case, and examples can be seen of delays ranging from one day to four days, even though the majority were about two and one-half days. Not a l l the flares, nor even a l l the importance 3 flares produced recognizable terrestrial effects. A careful examination indicated that several other factors must be considered in this connection. The region of origin on the sun and whether or not radio noise emissions from the upper corona occurred in conjunction with the flare determined which flares were apt to produce Earth-reaching particles. The detailed criteria are given below. These were deduced empirically on a portion of the data, and then used to predict disturbances for the rest of the year, as shown in column, k of Table I. I 2 5 4 5 0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 DAYS DAYS DAYS DAYS FIG. 39—GEOMAGNETIC INDICES FOLLOWING A N U M B E R OF S O L A R F L A R E S - 9 3 -The probability that a given flare will produce terrestrial dis-turbances through particle bombardment is given by the importance of the flare, i ts position on the sun, and by the associated radio noise in the following manner: Flare importance 1 - to 2-Flare importance 2 and 2+ Flare importance 3 -Flare importance 3 Flare importance 3 + If the flare occurs in the sun's central region (a circle, centred on visible disc, of one-half radius of disc) If in outer region of the sun If a large burst of 5 0 0 mc noise occurs followed by a large 50 mc noise burst, and i f the de-duced velocity is appreciably above 600 km/sec, multiply above total by If the deduced velocity is less than 600 km/sec, multiply by If a large burst of 50 mc noise, or a large noise storm occurs in conjunction with the flare, If a 50 mc noise storm follows the flare and per-sists for longer than k hours If a large cosmic ray increase is associated with the flare probability is 0 probability is 1 probability is 2 probability is 3 probability is k add 1 add 0 add 1 add 1 add 1 The total is the probability (on a scale 0 - 9 ) that an iono-spheric disturbance will follow the flare with a delay of the order of two or three days. (Note that values greater than 9 are taken as 9 on this scale.) 6 . 3 Surge Prominences and Disappearing Filaments These two solar phenomena were responsible for the majority of the terrestrial effects listed in Table I that were not associated with solar flares. Both have been taken to be identical processes despite the difference in terminology. Optical evidence has shown that the emitted matter from such phenomena is not necessarily ejected in a radial direction, and particles may reach the earth even though the eruption occurs on or near the sun's limb. The author has found that both have -9k-been the cause of ionospheric disturbances as indicated in Table I. In the case of the surge prominences, the criteria for esti-mating the probability of a subsequent disturbance is based on the size of the prominence, its location on the sun, and the associated radio noise in the following mariner: If the surge prominence is small If the surge prominence is medium If the surge prominence is large If i t occurs between 50 °N and 50 °S on either limb, If outside these limits If a large noise storm at 50 mc is associated with the event, and continues for at least k hours and i f there is also increased noise at 500 mc, If only a storm or bursts at 50 mc, If the surge prominence is observed to extend to great heights, If the observed motion i s very rapid, probability is 1 probability is 2 probability is 3 add 1 add 0 add 2 add 1 add 2 add 1 The total is the probability (on a scale of 0 to 9) that an ionospheric disturbance will follow the eruptive prominence with a delay of two or three days. In the case of disappearing filaments, the criteria for pre-diction is much the same. If the filament that disappears is small, If the filament that disappears is medium, If the filament that disappears is large, If i t occurs in the sun's central region (a circle centred on visible disc of radius one-half that of disc) If in outer region, but between 50°N and 50°S If in polar region, If a large noise storm on 50 mc is associated with disappearance and lasts at least k hours, and i f a noise storm occurs on 500 mc, If only a noise storm, or bursts occur on 50 mc probability is 1 probability is 2 probability is 3 add 3 add 1 add 0 add 2 add 1 -95-If the disappearance is very rapid add 1 If the filament was over a centre of activity add 1 The total i s the probability (on a scale 0 to 9) that an ionospheric dis-turbance will follow the disappearing filament with a delay of approxi-mately two or three days. As in the case of flares, each surge prominence and disappearing filament has its own individuality, and the character of the subsequent terrestrial disturbance differs in each case. This is seen in Fig. kO where values of Kp have been plotted for several disturbances which followed such solar events. It may also be seen from Fig. kO that the delay times differed somewhat from one event to the next. In general the delays appeared to be much the same as in the case of flares, ranging from two to three days, with several lasting somewhat longer. This would imply that the particle velocities were of the same order of magnitude as those associated with flares. Moreover, there was.lno reason to believe that the velocity dis-tribution was markedly different in character from the Maxwellian curve considered for the case of flares. To be sure, the author was unable to observe single, clean-cut bursts of noise which could be used for velocity determinations, but subsequent noise storms were observed from which velocities were deduced in the manner described. Such measurements were in keeping with his hypothesis of a distribution of velocities for the ejected particles. 6,k Sunspots Only in a few cases was the author able to relate the central meridian crossing of a sunspot to a subsequent ionospheric disturbance, and in those cases the sunspots were large and noisy. It is suggested that matter is being shot out from sunspots with fairly low velocity most of the time. As the area of the spot increases, more energy is available and greater velocities result for the particles, until, for the very large spots, a significant number of particles have sufficient energy to reach the earth. This hypothesis was supported by radio noise measurements at the long wavelengths. The author found that i f the size of the spot was great enough (greater than 1500 millionths of the solar disc area) and i f 4 Kp F I G . 4 0 - G E O M A G N E T I C INDICES FOLLOWING A N U M B E R OF S U R G E P R O M I N E N C E S AND DISAPPEARING F I L A M E N T S -97-the accompanying noise storm was intense enough (greater than 20 pabs at 50 mc), then the probability of a subsequent ionospheric disturbance was quite good. The evidence indicated that ionospheric disturbances followed some two or three days after the central meridian passage of such an active region on the sun, implying that a significant number of solar particles were ejected with radial velocities in excess of 620 km/sec. This result i s in agreement with the suggested Maxwellian distribution. 6.5 Other Solar Sources A number of magnetic and ionospheric disturbances were listed in Table I which could not be directly linked to sunspots, flares, or prominences and filaments. On these occasions, the solar noise data were insufficient to permit identification with any solar phenomenon, and the author was forced to assume that M-regions were responsible. In several cases, particularly early in the year, the optical data were somewhat scanty for identification purposes. On the whole, a comparison of the predicted probability of a disturbance, made according to the above criteria, with the observed dis-turbance classification showed considerable agreement. This correlation is shown in Table II. In the table, a zero under observed ionospheric disturbances signifies that no disturbance was observed, while a zero under predicted probabilities means that no disturbance at a l l was pre-dicted. -98-TABLE II Correlation between Predicted and Observed Ionospheric Disturbances Predicted Probability of Disturbance O I 2 3 4 5 6 7 8 9 Observed Ionospheric Disturbance ON ui -p" OJ ro H o 3 2 0 2 7 2 0 5 1 5 1 0 4 2 1 1 1 8 1 3 8 4 1 2 6 4 1 2 4 3 2 1 2 3 1 1 1 A correlation coefficient of +O.65 has been computed for the data in this table. Consequently, the estimated probabilities seemed to be a fairly reliable means of forecasting the occurrence of ionospheric dis-turbances. In particular, a probability of 4 or more almost invariably signified an ionospheric disturbance of some kind, and probabilities in excess of 5 were associated with intense disturbances. CHAPTER VII CONCLUDING REMARKS 7.1 Discussion The investigations described in the foregoing sections have large-ly been restricted to what actually happened, and the author has tried to avoid too much speculation on the process by which i t happened. However, in such a comprehensive subject as solar-terrestrial relations, there is s t i l l ample opportunity for conflicting viewpoints on what did happen in particular instances. Occasions have arisen when different observations of magnetic, auroral, and ionospheric phenomena were at variance. It is thought that this state of affairs resulted principally because of geo-graphic differences in the observing stations, and could have been resolved by a world-wide measure of the respective phenomena. Such a step is an impractical one; the best that can be done is to have an average value computed from a number of stations around the globe which are measuring the same effect. The planetary magnetic indices are obtained in this manner. However, the readings made at any one station are a better indi-cation of physical behaviour in that particular region than the planetary values. It is quite possible that certain localized effects may occur in one region before they happen in another region, and that the world-wide effects may occur at an entirely different time. The data gathered during the International Geophysical Year by the many participating countries should resolve much of the seemingly conflicting evidence of this sort. Among other things one may be able to find the time of the f i r s t indi-cation of an ionospheric disturbance at any station on the globe in re-lation to the time of the solar event, and in this way determine the velocity of the most rapid solar particles. The author has used the Kp data to indicate precipitation of particles on the earth. In addition he has used three other indicators to support the Kp measurements, namely the amount of absorption on a H.F. communication circuit, rate of scintillation of a radio star, and the latitude at which aurora occurred. The justification for using these data in this way lay in the good correlation that was found between them and the Kp values. This, together with the fact that they were obtained close to the Northern Auroral Zone, implied that each could be taken as - 9 9 -- 1 0 0 -an indication of incoming particles. Direct evidence of solar corpuscles entering the ionosphere is not readily available, and such indirect evidence as indicated above must be relied upon. The only direct measure seems to be the observation of hydrogen lines in the auroral spectrum. Meinel^^ showed that such spectral lines, which occurred in certain auroras, were due to high speed, protons. Moreover, he showed that the direction of travel and the velocities at the 1 0 0 km level of the ionosphere could be deduced from the doppler shift of the H^  line. Such velocities, however, are not those which the protons have at the top of the ionosphere since they suffer considerable deceleration in penetrating the atmosphere. It has been shown that the incoming protons must have velocities of at least 2 0 0 0 km/sec in order to penetrate to the level where most of the auroral light originates, and any with velocities less than this would not pene-trate to the E region. In consequence, the presence of the Hg line in the spectrum of the low level aurora implies that protons of solar origin have velocities in excess of 2 0 0 0 km/sec as they approach the earth. Spectroscopic data obtained in the auroral zone by Montalbetti* ' during a four month period in 1 9 5 6 were examined by the author for B Q occurrence in the wake of solar eruptions. It seemed likely that such positive evidence should provide confirmation for the velocities deduced from Table I for particular events on the sun. In a number of cases a strong hydrogen line appeared in the spectrum at times when a sudden com-mencement magnetic storm was in progress. However, such occasions followed at least two days after the eruptive solar event which released the particles. One is led to conclude that either the H Q light of the aurora occurs very high in the ionosphere or else that an acceleration mechanism, such as the ring current postulated by Stflrmer and Martyn, exists close to the earth. The suggestion has been made by Dagg^^ that scintillations result from ionospheric irregularities produced by solar ultra-violet radiation. The correlation found by the author between scintillation rates and geomagnetic indices would rule out this possibility. There now seems no doubt but that the radio star fluctuations observed near x -101-the auroral zone at least are due to incoming solar particles. Hewish*-3 ' has placed the diffracting region responsible for the fluctuations at a height of 400 Ism. However, the evidence in support of this conclusion is far from conclusive. The author has shown that high scintillation rates are characteristic of a disturbed ionosphere, and i t is well known that F2 region conditions are greatly affected at such times. This does not necessarily mean that the scintillations occur in the F2 region, since the D and £ regions also exhibit unusual activity at times of ionospheric storms. It seems likely that a consideration of the energies of the particles entering the ionosphere and producing the irregularities re-sponsible for the scintillations would provide a solution to the difficulty. On such an argument, those particles with velocities in excess of 2000 km/sec at the top of the ionosphere would be able to penetrate to the E region before losing a l l their energy through collisions. On the other hand, particles with lower velocities would be stopped at higher levels and contribute to irregularities at those heights. The height of the scintillating region would not be fixed, but could vary from one time to the next. To be sure, the vertical transport of ionization might affect this argument, but l i t t l e evidence exists to support such movements. It is possible that scintillations may be due to two separate agencies. While the high scintillation rates and amplitudes may be attri-buted to high speed solar particles, the low indices for the more normal ionosphere might arise from much slower particles impinging on the iono-sphere. These latter particles may be galactic hydrogen atoms with velocities of about 33 Km/sec as suggested by Harrower^0^ or only the very slow solar particles. Again, more observational data is required to clear up this issue. The evidence presented above has led to the conclusion that, in general, only the large eruptive events on the sun release sufficient energy in the form of high speed particles to affect the earth's iono-sphere appreciably. Such eruptive events are associated with a centre of activity on the sun. The concept of a centre of activity was intro-duced^ 1^ to represent the totality of a l l visual phenomena accompanying -102-the formation of a sunspot group. These phenomena include sunspots, faculae, chromospheric flares, structure of the chromosphere, filaments or prominences, and variations in the surrounding corona. A centre of activity results when a magnetic disturbance reaches the solar surface, and these magnetic forces provide the high velocities of the emitted particles. The details of the evolution cycle differ from case to case, but the appearance of such a centre of activity on the sun inevitably implies the release of energy in the form of high speed particles. Measurements of the velocities of the solar particles, such as (72) those made by Ellison* 1 have yielded values of the order of several hundred km per sec. Occasionally values in excess of the escape velocity (620 km/sec) have been recorded. The author has shown, however, that not a l l the particles travel with the same velocity, and that particles of matter leave the centre of activity with an appreciable range of velocities. Furthermore, he has shown that i f this velocity distribution is taken to be a Maxwellian distribution the observed noise bursts from the sun and observations of ionospheric disturbances can be readily explained. Statistical studies of geomagnetic activity and such solar features as relative sunspot numbers, areas of prominences, areas of filaments, dark and bright hydrogen flocculi, and calcium flocculi have been made by Kiepenheuer*-7^, Roberts and T r o t t e r F e l i c i n ^ 7 ^ ) , and others. While various degrees of correlation have been reported, the results lack significance since no consideration was given to the velocity with which the particles were emitted from such solar phenomena. No doubt there is corpuscular radiation of some degree at such times, but i t is the author's contention that, in general, the velocities are much below the escape velocity. The measurements made in the present study have indicated that large numbers of high speed particles are able to reach the earth only from a very large sunspot group, a large surge prominence (or disappearing filament), or from a large solar flare. Pre-sumably at other times the maximum of the velocity distribution curve for the ejected particles lies at a much lower value, as evidenced by visual and spectroscopic measurements of prominences. -103-The solar noise recordings at 50 mc have contributed much to the understanding of particle emission from a centre of activity. Most of the excitation in the corona seems to occur over sunspots whose area exceeds 300 millionths of the disc area, indicating that matter is being ejected into the corona. In addition, flares and subf lares in a centre of activity contribute to noise bursts, and frequently spiky noise is observed at times of large filaments and prominences. The larger flares and sunspots, the disappearing filaments and surge prominences also produce noise of the same general character, but of somewhat greater magnitude. The energy of the ejected particles is only great enough for them to reach the earth in the case of these latter phenomena. Consequently, by taking suitable observations of the sun i t is possible to predict ionospheric disturbances with considerable reliability. There are s t i l l many important problems to be solved before the complete story on solar-terrestrial relations is known. The data dis-cussed in this thesis are only a very small sample of the sort of obser-vations required in this connection. At the present time an intensive period of observations on our planet, known as the International Geo-physical Year, has just commenced. Numerous scientists in many countries are participating in investigations aimed at resolving this and other geo-physical problems. Their findings may provide the final conclusion on the influence of solar particles on ionospheric disturbances. 7.2 Summary The author has confined his considerations primarily to those solar-terrestrial relations which could be explained on the basis of particles of matter being ejected from the sun and impinging on the earth. Most of the details of what happened to the particles in transit were glossed over and the emphasis was placed on measurable effects of the particles in the sun's atmosphere and in the earth's atmosphere. In this connection a variety of data covering a 12 month period when solar activity was quite high have been critically examined. The author sought to establish just when particles were entering the earth's atmosphere, and just when the particles had lef t the sun. •104-The following measures were used as indications of incoming corpuscles: the degree of ionospheric absorption observed on a long-range high frequency communication circuit, the rate of scintillation of a 'radio star', the planetary indices of magnetic activity, and the most southerly extent of the Aurora Borealis as observed in Central Canada. Correlations between these different measures were sufficiently high as to leave no doubt that they are causally related. A l l the ionospheric disturbances were then compared to the solar data for the same period to establish their respective origins. The solar data in question consisted of both optical and radio measurements. The former indicated when eruptive phenomena were in pro-gress on the photosphere, while the latter provided information on the ejected matter travelling through the sun's corona. It was found that solar flares, eruptive prominences and disappearing filaments were the most frequent sources of Earth-reaching particles. Only occasionally could large sunspot groups on the sun's central meridian be associated with subsequent ionospheric storms. From radio noise measurements at 500 mc and 50 mc i t was poss-ible to estimate velocities of travel for the emitted particles. Such information, together with a knowledge of the region of origin on the sun permitted the author to deduce reliable criteria for predicting subsequent ionospheric disturbances. While these criteria were empirically deter-mined in retrospect for a portion of the year's data, they were later applied to the remainder of the data with considerable success. A corre-lation coefficient of 40.65 was found between the probability of a sub-sequent ionospheric storm, predicted on the basis of solar data only, and the actual observed storm intensity. Probably the most important feature of the foregoing work has been an indication of the wealth of information obtainable from the noise recordings. The scintillation work permitted an examination of the structure of the ionosphere, particularly at times of abnormal conditions, in regions from which l i t t l e information was previously obtainable. This tool has not been thoroughly exploited yet; here the results have been used only to specify the precipitation of particles into the upper atmosphere. -105-On the other hand, the solar noise data have provided consider-able information on the processes in the corona that are associated with flares and surge prominences. The present work has shown that the ejected particles from such an eruptive phenomenon do not a l l have the same velocity. Rather, there is a distribution of velocities which is similar in character to a Maxwellian distribution. By assuming a Maxwellian distribution of velocities for the particles, i t was shown that the energy distribution in time available at a certain height in the corona was consistent with observations of the shape of the noise bursts from that level. Moreover, the energy distribution in time for particles which are able to reach the earth is consistent with observations on the intensities of magnetic and ionospheric storms. A study of sunspots and the associated 50 mc noise storms also pointed to a velocity distribution for the solar particles which are con-tinually being ejected from the spot, rise to great heights in the corona, and then f a l l back to the sun's surface. As the spot size increased, the mnv-jmnm of the distribution curve shifts to higher velocities, until, eventually, an appreciable portion of the particles are able to reach the earth. This is consistent with observations that only sunspots above a certain size produce terrestrial effects. -106-BIBLIOGRAPHY 1. K. Birkeland, Arch. Scl. Phys., 1, 497 (1896). 2. K. Birkeland, Norwegian Aurora Polaris Expedition, 1902-03, Christiania, 1 (1908). 3. C. Stormer, Arch. Sci. Phys. Geneve, 24, 317 (1907)• 4. C. St&rmer, Phys. Rev., 45, 835 (1934). 5. W.H. Bennett and E.O. Hulburt, Jour. Atm. Terr. Phys., 5, 211 (1954). 6. S. Chapman and V.C.A. Ferraro, Terr. Mag. Atmos. Elect., 37, 147 (1932). 7. D.F. Martyn, Nature, 167, 92 (1951). 8. H. Alfven, Kungl. Sve. Vetenskaps Akad. Handl. Stockholm, III, 18, No. 9 (1940). 9. E.O. Hulburt, Phys. Rev., 31, IO38 (1928). 10. A.B. Meinel, Astrophys. J., 113, 50 (1951). 11. R.S. Richardson, Trans. Amer. Geophys. Union, p.558 (1944). 12. J.S. Hey, Nature, I57, 47 (1946). 13. H. Ebert, Astronomy and Astrophysics 12, 804 (I893). 14. Sir Oliver Lodge, "Signalling across Space without Wires", (3rd Ed.), Electrician Printing and Publishing Co., London, p.33 (1900). 15. G.C. Southworth, J. Franklin Inst. 239, 285 (1945). 16. G. Reber, Astrophys. J., 100, 279 (1944). 17. J.P. Wild, J.D. Murray, and W.C. Rowe, Nature, 172, 533 (1953)-18. J.P. Wild, Aust. Jour. Sci. Res. A., 3> 399 (1950). 19. T.R. Hartz and W.A. Sharf, nA Radio Star Scintillation Recorder", RPL Project Report 23-1-2, (May, 1956). 20. M. Ryle, Proc. Roy. Soc. A, 211, 35I (1952). 21. "The American Ephemeris and Nautical Almanac, 1957". United States Government Printing Office, Washington. 22. K.E. Machin, M. Ryle, and D.D. Vonberg, Proc. Inst. Elect. Engrs. I l l , 99, 127 (1952). 23. F.G. Smith, Nature, I65, 422 (1950). 24. C.G. Little and A.C.B. Lovell, Nature, 165, 422 (1950). 25. T.R. Hartz and W.S. Cook, "A U.H.F. Coaxial Line Amplifier", RPL Project Report 23-0-2, (September, 1955). 26. J.H. Meek and F.S. Hector, Can. J. Phys. 33, 364 (1955). -107-27. 0, Heaviside, Etxcy. Brit. (Ninth Ed.) 33, 215 (1902). 28. A.E. Kennelly, Elect. World and Eng., 39, 473 (1902). 29. S.K. MLtra, "The Upper Atmosphere", 2nd Ed. (1952). The Asiatic Society, Calcutta. 30. G. Breit and M. Tuve, Phys. Rev. 28, 554 (1926). 31. "Prediction of Optimum Traffic Frequencies for Northern Latitudes", RPL Report 1-1-3 (November, 1954). 32. J.S. Hey, J.W. Phillips, and S.J. Parsons, Proc. Roy. Soc. 192, 425 (1948). 33. G.C. Reid, Can. Jour. Phys. in press. 34. T.R. Hartz, Can. Jour. Phys. 33, 476 (1955). 35. C.H. Costain, "Radio Star Scintillations close to the Auroral Zone", M.A. Thesis, Univ. of Sask. (1955). 36. A. Hewish, Proc. Roy. Soc. A., 214, 494 (1952). 37. J.H. Meek and C.A. McKerrow, "Ionospheric Observers' Instruction Manual", DRTE /RPL Ottawa (1951). 38. T.R. Hartz and G.C. Reid, "Investigation of the Ionosphere by Means of Radio Star Scintillations", RPL Project Report 23-1-1 (Feb., 1956). 39. J. Barbels, N.H. Heck, and H.F. Johnston, Terr. Mag. Atmos. Elect. 44, 441 (1939). 40. C. Stflrmer, "The Polar Aurora". The Clarendon Press, Oxford (1955), P.14. 41. E.H. Vestine, Terr. Mag. Atmos. Elect. 49, 77 (1944). 42. D.M. Hunten, Can. Jour. Phys. 31, 681 (1953). 43. S.K. Mitra, op. cit., p.398. 44. W. Brunner, Astronomische Mitteilungen, pp.131-136 (1939). 45. L. Vegard, Geofys. Publ. 18, No. 5 (12 March, 1952). 46. G.W. Gartlein, Phys. Rev., 61, 463 (1951). 47. S. Chapman, Mon. Not. Roy. Astr. Soc, 89, 456 (1929). 48. H.A. Brtfck and F. Ruttland, Mon. Not. Roy. Astr. Soc, 106, 130 (1946). 49. J.S. Hey, S.J. Parsons, and J.W. Phillips, Mon. Not. Roy. Astr. Soc, 108, 354 (1948). 50. L. Biermann, Zs. f. Ap. 29, 274 (1951). 51. J.P. Wild, J.D. Murray, and W.C. Rowe, Aust. J. Phys. 7, 439 (1954). 52. J.C. Jaeger and K.C. Westfold, Aust. J. Sci. Res. A, 3, 376 (1950). 53. K.C. Westfold, J. Atm. Terr. Phys., 1, 152 (1951). -108-54. A . E . Cov ing ton , p r i v a t e communicat ion. 55. T . R . H a r t z , Na ture , 175, 908 (1955). 56. V . E . Kruse , L . M a r s h a l l , and J . R . P i a t t , A s t r o p h y s . J . , 124, 601 (1956). 57. Ruby Payne-Scot t and A . G . L i t t l e , A u s t . J . S c i . Res . A , 4, 508 (1951). 58. T . R . H a r t z , "50 mc/s S o l a r Noise f o r the P e r i o d J u l y 1 t o December 31, 1955", REL P r o j e c t Repor t 23-2-2 ( A p r i l , 1956). 59. Mar i e - Joseph M a r t r e s , G . O l i v i e r i , and J . R a y r o l e , Camptes Rendus, 244, 1724 (1957). 60. K . O . Kiepenheuer , "The Sun" e d . by G . P . K i u p e r , U n i v e r s i t y o f Chicago P res s (1953) P.408. 61. J . B a r t e l s , T e r r . Mag. Atm. E l e c t . , 37, 48 (1932). 62. H.W. Newton, Mon. N o t . Roy. A s t . S o c , IO3, 244 (1943). 63. H.W. Newton, Mon. N o t . Roy. A s t . S o c , 104, 4 (1944). 64. H.W. Newton and W. Jackson , "Seventh Repor t o f the Commission f o r the Study o f S o l a r - t e r r e s t r i a l R e l a t i o n s " , Hemerle, P e t i t and C o . , P a r i s (1951) p.107. 65. M . C . Johnson, Mon. N o t . Roy, A s t . S o c , 86, 300 (1926). 66. Ruby Payne-Sco t t and A . G . L i t t l e , A u s t . J . S c i . Res . 5, 32 (1952). 67. K . O . Kiepenheuer , op . c i t . p.393. 68. R . M o n t a l b e t t i , p r i v a t e communicat ion. 69. M . Dagg, J o d r e l l Bank A n n a l s , 1, 99 (Dec. 1956). 70. G . A . narrower , Na ture , 179, 608 (1957). 71. K . O . Kiepenheuer , op . c i t . p.432. 72. M . A . E l l i s o n , Mon. No t . Roy . A s t . S o c . , 116, 624 (1956). 73. K . O . Kiepenheuer , A s t r o p h y s . J . , 105, 408 (1947). 74. W .0. Rober ts and Dorothy E . T r o t t e r , J . Atmos. T e r r . P h y s . , 6, 282 (1955). 75. F . N . F e l i c i n , As t ronomichesk i Z h u r n a l , 17, 21 (1940). 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items