Contacter l'auteur / Contact the author

Recherche dans ce site / Search in this site







HF Propagation tutorial

Van Allen Belts and the geomagnetic field.

by Bob Brown, NM7M, Ph.D. from U.C.Berkeley

Geomagnetic disturbances (VIII)

The end of the second volume of the book, "Geomagnetism" by Chapman and Bartels, has an interesting account dealing with the first days of magnetic observations in Sweden by Celsius and one of his graduate students. Knowing what we do now, I consider that as "Day One" of the Space Age. But I have to marvel that it took 75 years until Oersted came up with the idea of a current (like an ionospheric electrojet) giving rise to magnetic deflections (on the ground below an aurora) of a compass. Compare that time with the five years it took the French mathematicians to come to grips with the Biot-Savart Law for magnetic effects of currents. Interesting!

Finally, an excellent discussion of early auroral observations in Norway can be found in the last chapter of Brekke's book, "Physics of the Upper Polar Atmosphere" published by Wiley & Sons in 1997. Brekke, being a Norwegian, pays homage to the works and tradition of good auroral physics established by Stoermer. It's worth a bit of reading time, believe me.

At the end of the previous page, we made note that magnetic storms give rise to auroral disturbances, with optical emissions coming from above the 100 km layer, VHF reflections off the ionization in auroral displays, ionospheric absorption of signals going across an active auroral zone and strong magnetic disturbances observed on the ground from the current systems which develop along the ionized region. All that from an enhancement in the solar wind, perhaps coming at a greater speed, with a greater particle density or with the interplanetary magnetic field pointing south with respect to the earth's field.

Nowadays, we can read about all those changes on the Internet. But the most important one for magnetic storming has to do with the interplanetary field and its orientation. With the field pointing south, conditions when Bz is negative, the interplanetary field can merge with the terrestrial field (a non-classical concept) and field lines on the front of the magnetosphere then transferred to the tail region as the solar plasma sweeps by.

These ideas came forward in the '50s, thanks to the efforts of J. Dungey of the U.K. and others. As I said earlier, they go beyond the elementary considerations we get in classical courses on electromagnetic theory and are best left for the theorists to discuss. We only need to know what happens to the ionosphere and there, the news is BAD as the F-region loses ionization with the development of a magnetic storm.

But the E-region can gain ionization, with the penetration of auroral electrons. Those particles are from here inside the magnetosphere itself, not directly from the solar wind, and are accelerated locally, going from a fraction of an electron-Volt up to tens of kilovolts energy. And their flux can be quite large, resulting in electron densities of a million or more per cc from electron collisions with atmospheric constituents in the tens of kilometres above the 100 km level. The colors of the aurora are testimony to the collisions with the neutral constituents and the electron densities that result can give rise to signal absorption.

That last point may seem strange if you go back to the curves that were given page 2. There, the relative absorption efficiency per electron was dropping off quite rapidly above 100 km. But in the case of aurora, there are millions of electrons per cc up there and even if electron-neutral collisions are less frequent above 100 km, losses result just from the sheer amount of ionization that goes with an aurora.

But to give some numbers, auroral absorption of up to 5 dB or so is found in the riometer records of 30 MHz galactic radio noise coming in vertically. But that is just for one pass through the ionosphere. For amateur communications, say on 28 MHz, that should be doubled for a complete hop, increased even further by a factor of 3-4 for the oblique angle of the path and adjusted for the inverse-square frequency variation. At lower frequencies, that last adjustment shows even greater losses on those bands. So it should be no real surprise that auroral absorption represents an adverse factor for amateur communications.

This drawing illustrates an electromagnetic reconnexion (between the solar wind in yellow and the geomagnetic field in red) called a "crack" in the shield protecting earth. This is through such cracks, as large as a state and remaining open for hours, that the solar flux can penetrate into the ionosphere and create auroral events. Click here to run the animation (1.7 MB MPEG file) prepared by NASA/GSFC.

Those remarks dealt with the electron density; one should also note the geometry and activity of the aurora. In regard to geometry, auroral activity at any given time is restricted to a narrow latitude range. (See Research Notes) But it can extend over a wide range of longitude and the type of activity varies from west to east. In evening hours, aurora tend to be quiet and not involve a lot of energetic particles (and ionization). Around midnight, the activity may increase dramatically, with displays flashing wildly overhead and in considerable motion. It is even possible to note from the distinct ray structures that the electron influx comes down the inclined magnetic field lines. Then in the morning hours, the aurora becomes more diffuse, shows some pulsating patches and more ionospheric absorption, slowly varying compared to that around midnight and much greater than before midnight.

HF signals that go across an auroral region will show effects characteristic of the activity - steady signals going across in local evening, considerable rapid absorption and flutter from the moving regions of ionization around local midnight and just strong absorption for local morning. Of course, all those ideas have to be tempered by the frequency involved, with devastating absorption on 160 meters and possible auroral reflections above the HF range.

The magnetic disturbances at high latitudes which accompany aurora give qualitative measures of the energy input to the magnetosphere from the impact of the solar wind. Nowadays, one can go to NOAA satellite data and obtain numerical values for the power input from observations of the influx of auroral electrons with energies up to about 25 keV. The numbers can be quite large, from 1 to 500 Gigawatts over one hemisphere. Such inputs can have profound influences, auroral heating and magnetic activity, but our concern is only with communications so we have to look at how frequently these events occur and if they can be anticipated.

Recent data published by NOAA gives a summary of magnetic storm activity over Solar Cycles 17-22 to suggest how the levels of magnetic activity might vary, year by year, in Cycle 23. Now when it comes to magnetic activity, indices are used to characterize what level of disturbance (from quiet conditions) is in effect, say in a 3-hour period or averaged over a day. In that regard, a number of magnetic observatories have been selected to provide data for use in making planetary averages. The actual data sets are normalized to common scales, 0 to 9 for the 3-hour Kp-index and 0 to 400 for the daily Ap-index.

One can obtain those data from the Internet and keep records to see if there is any recurrence tendencies. Indeed, there are and logging Ap indices is one way to anticipate possible disturbances that come from long-lived solar streams sweeping past the earth or stable active regions which are the source of increased levels of ionizing radiation.

Magnetic storminess is categorized in terms of Ap values and minor storms correspond to elevated levels of Ap while actual storms correspond to Ap greater than 40 and severe storms are when Ap is greater than 100. In that regard, the storm of May 3, 1998 had an Ap level of 112 while the greatest storm ever recorded was in September 1941 and had an Ap value of 312! Like the March '89 storm which put the Province of Quebec in the dark for a day, that one affected the power grid in the Northeast. Nowadays, the power industry is keenly aware of the magnetic storm problem and tries to anticipate problems by getting solar wind data from satellites, out there ahead of the earth and in the solar wind.

Anyway, both minor and major storms affect HF propagation for hours at a time or a day by their adverse effects on F-region ionization but severe storms reduce the bands to barren wastelands for days at a time. Propagation doesn't return until slow photo-ionization processes replace the F-region electrons.

Planetary K Indices

Geomagnetic Storm Level

K = 5

G1 Minor

K = 6

G2 Moderate

K = 7

G3 Strong

K = 8

G4 Severe

K = 9

G5 Extreme

Active:    K = 4
Unsettled: K = 3
Quiet:     K = 0, 1, 2

A = 100-400: Severe
A = 50-99 : Major
A = 30-49 : Minor
A = 16-29 : Active
A = 8-15 : Unsettled
A = 0-7 : Quiet

K-0 = A-0
K-1 = A-3
K-2 = A-7
K-3 = A-15
K-4 = A-27
K-5 = A-48
K-6 = A-80
K-7 = A-140
K-8 = A-240
K-9 = A-400

As we told at the end of the first page, the propagation aspects of magnetic activity are found on the SWPC website of the NOAA. On this site, scientists release daily reports and alerts related to the solar and geophysical activities as well as a 3-day Forecast. This product contains the Observed/Forecast 10.7 cm flux and K/Ap indices.

The effects of magnetic storming are the greatest, as you might suspect, at the higher latitudes and on the higher frequencies. For communications over any distance, differences in longitude mean that great-circle paths usually swing north and thus are at risk during magnetic activity. This is not too bad for short-path communications as the windows of opportunity can be rather wide. But that is not the case for long-path propagation; there, the path opens with the rise in F-region critical frequency with sunrise on the path and closes shortly thereafter as D-region absorption increases at lower altitudes. In short, if an opportunity is lost on a given day, one must wait for another day and try again. But having spent many happy hours in pursuit of long-path contacts, I can say it is worth it.

Turning to longer ranges in forecasts, the recent NOAA prediction for magnetic storminess during Cycle 23 is shown at right.

Given that forecast, we can look forward to major storm activity rising to about 2 per month by Year 6 (2002) in Cycle 23. That is not a good prospect but there are uncertainties in forecasts so one can hope for less and see what happens.

Note by LX4SKY. As expected the first months of the year 2002 were as disturbed as 2000 with a solar flux 5% higher (F10.6 of 220 SFU vs. 210 SFU in 2000) but decreasing rapidly with sunflares of X-class ejecting fast particles that produced indirectly some intense and highly colored aurora over Alaska, Canada and Finland.

The 10.7 cm solar flux is an indication of active regions on the solar disk and that is a quantity that warrants logging. Early in a cycle, new active regions begin to appear but later, some regions are quite stable, particularly around solar maximum, and knowing when the flux may peak again is quite helpful to DXers.

The origins of the magnetic activity differ throughout a solar cycle, however, with early part of the cycle giving more of the sporadic coronal mass ejections responsible for solar wind blasts hitting the magnetosphere. On the other hand, the latter part of a cycle is one characterized by fast streams from coronal holes sweeping past the earth. Those can be long-lasting so logging magnetic activity, with the A-index from Boulder for several solar rotations is a good idea, enabling one to avoid times of strong magnetic activity.

One aspect of strong magnetic activity is equatorward expansion of auroral displays, associated with the loss of magnetic field lines from the front of the magnetosphere to the magneto-tail. From the standpoint of propagation, that results in very low MUFs in the polar cap. But it is accompanied by an expansion of the polar cap that can bring on heavy, long-duration ionospheric absorption. That is the case with solar proton events, so-called polar cap absorption (PCA) events. Those events differ in striking ways with auroral absorption (AA) events but both can be present at the same time. Those events will be our next topic of discussion.

Research Notes

I have already given some words of praise for the book, "Physics of the Upper Polar Atmosphere", by A. Brekke. To that I would like to add that the front cover has an absolutely fantastic photo of an aurora taken from a satellite. There is a catch, however; the photo was made in Antarctica and the book must be turned upside down to get the aurora positioned over the polar cap. But like Confucius said, "A graphic is worth many kilobytes of text."

Geomagnetic storms and aurora

We are now into disturbances of propagation, those nasty things that can plague us, sometimes without our even knowing it. The last topic was magnetic storms and aurora. Those represent disturbances of the F- and E-regions, respectively.

The effects of magnetic storms can be world-wide in the sense that ionospheric electrons are removed from field lines, lowering the MUFs on paths across great distances. The part of the ionosphere which is disturbed the most is in the polar cap as that is the region whose field lines are most at risk. And recovery from magnetic storms is a slow process, requiring the electrons in the F-region be re-supplied by sunlight, a slow, tedious process which can take days after a severe storm.

At left, an aurora (curtain) pictured en 2002 by Dennis Anderson in Alaska with a 6x9 DSLR body equipped with a 50 mm f/2.8 lens. 35 sec exposure on Fuji Provia 400 F (the lake and the moon are "accidents" from a previous exposition). At centre, a marvellous aurora (mid curtain mid rays) also recorded in Alaska by David Fritts from PFRR. At right, an aurora (curtain) pictured by Jouni Jussila, from Oulu, Finland, on March 20, 2001 with a 24 mm f/1.4 lens on Fuji Provia 100F pushed at 200 ISO. 15 sec exposure. Remember that all these events are dynamic and evolve in time, changing usually of shape and brightness in a few tens of seconds only.

The effects of an aurora, by itself, are much more localized in the sense that the increased ionization is confined to the field lines that guided auroral electrons downward. Short of being in a full-blown magnetic storm, the effects tend to be brief, measured in minutes or hours, and when the aurora ends, it is a fairly rapid process. Essentially, the problem is to have the electrons in the ionization recombine with the positive ions which were generated by the influx of energetic auroral electrons.

To listen: Auroral emissions

Audi CD to buy: Auroral Chorus I and II by Stephen P. McGreevy

But now we come to solar proton events. Those will affect the D-region and originate on the sun, with protons and other particles accelerated up to energies of millions, sometimes even billions, of electron-Volts (MeV or BeV). So solar proton energies, from acceleration on the sun, are high in contrast to those of auroral electrons which are accelerated locally, within the magnetosphere, up to tens of kiloelectron-Volts. The protons are accelerated in connection with some solar flares and then can leave the scene, passing through both the solar and the interplanetary field.

The interplanetary field generally points toward or away from the sun and the outward progress of protons depends on the degree to which they go along the field lines or perpendicular to them as they leave the sun. But the interplanetary field is not well-ordered like the geomagnetic field close to the earth so protons will diffuse through the region and their progress will depend on their momentum or the radius of curvature of their path. The more energetic protons will have radii of curvature which are large compared to the scale-size of field variations so those protons will follow more rectilinear paths. On the other hand, less energetic protons will have smaller radii of curvature in the field and their progress will be more like diffusion, scattered by the small-scale, organized portions of the interplanetary field.

All that is a way of saying that the high energy-protons will leave the region close to the sun faster and make their effects felt more promptly, albeit briefly. On the other hand, the low-energy protons will diffuse slowly through the field and their effects will be of longer duration. It should not be forgotten, however, that the duration of the acceleration process is of interest too. Generally, it is considered to be the same as the actual flare process but those can be brief, in minutes, or longer, measured in hours.

At left, prelude to an intense auroral activity, this radioheliograph reveals a strong eruption on the sun limb. It was recorded on April 4, 2000 at 1342 UT at Nancay Observatory. Emission came from the active region AR8948 and is directed straight to earth. At center the unbelievable impact of the solar flux on the magnetopause two days later by 16h UT. Click on the image to run the animation. A right what happended in the sky from April 7, 2000... The high atmosphere is burning under the intense radiation emitted by electrons discharging their energy. Documents Observatoire de Paris-Meudon, S.M.Petrinec/Pixie and Polar.

Another way of saying the same thing is if the flare region is off the to the east of the solar disk, solar protons heading toward the earth will have to stagger through the field lines which are more or less perpendicular to their paths. That is a slower process and protons can be held in the magnetic field region for times which are long compared to the acceleration process that started them. As an example, I had experience with one east limb event in August '79 where the solar protons finally reached the ionosphere 18 hours after the flare! Staggering, diffusion? Yep!

On the other hand, flare sites toward the west limb of the sun send protons out into the field which generally trails behind the rotating sun and we get "sprayed", as it were, by protons going along the field lines. That is called the "garden hose" effect. The Great Solar Flare Event of February 23, 1956 was a case in point, a west limb flare where the travel time was measured in minutes. Those were relativistic particles and had so much energy (over 10 BeV) that they penetrated to ground level, even at the magnetic equator! Been there, seen that!

But what are their effects? Given the remarks in the last paragraph, one can expect that the duration of the proton bombardment of the earth will depend on the location of the flare site. That is one propagation clue that NOAA provides with every announcement of a solar flare, the solar longitude involved. So that is one item of interest, east or west of central meridian.

Map of the D-Region Absorption Prediction. Compared to a situation where there is not the least attenuation (at right), it is rare that this map turns so red. It was calculated on September 11, 2017, some days after several sun flares of class X and M associated to several emissions of CME by the Sun, among which one was directed toward the Earth two days earlier. When high energy solar particles reached the geomagnetosphere, they affected a large part of the ionosphere. That day, V/UHF bands were closed at high latitudes with auroras in Alaska and up to mid latitudes (>55 with Kp=8). This time the effects of the CME created a radio blackout up to 15 MHz where the attenuation reached till ~35 dB; a S7 signal looked to a S1 signal ! You can get real-time update on SWPC website.

But as to the effects of the protons, those depend on their flux (or number per cm2/sec) and proton energy. The low-flux, low-energy solar proton events were only conjecture until the Space Age but are detected nowadays by satellites and one can see the data in the Tiger Plots on a NOAA website. But events with higher fluxes and greater energies can penetrate the Earth's field and get reach into the ionosphere, the atmosphere and, on rare occasions, they can reach ground level.

Our interest, of course, is with ionospheric effects and being energetic charged particles, the protons will leave a wake of ionization as they plow through the atmosphere. The extent of the wake will depend on the relative numbers of protons in the various energy ranges - around 1 MeV, around 10 MeV, near 100 MeV and beyond. But generally, being both energetic and massive particles as compared to puny auroral electrons, protons penetrate deeper into the ionosphere (if they get that far through the geomagnetic field) and the heavy ionization near the end of their physical ranges can cause huge ionospheric absorption of signals because of the greater electron-neutral collision rate deep in the D-region.

For solar protons to get down to the ionosphere, they must first enter the geomagnetic field out at the magnetopause and then follow field lines, according on their momentum. The present view of these matters is in sharp contrast with the early days of ionospheric radio. Then, the dipole model of the earth's field was taken as the standard and all discussions about the effects of solar protons were based on work done by the Carl Stoermer, the Norwegian auroral physicist. So the idea was that protons were sorted out according to momentum (or energy) by the field and there was a sharp cut-off energy which varied with latitude.

Riometer recordings of a PCA event reaching 6 dB over Macquarie island (VK0, 54S) on April 16, 2002. Document IPS.

But with the IGY, things changed; the use of riometers, looking at ionospheric absorption due to the protons, showed that the cut-off idea was all wrong and the polar cap was wide open, full of low-energy protons, all the way down to the auroral zones where the cut-off energy was supposed to be 100 MeV. That was one of the first clues that the earth's field was not that of a dipole; then measurements made by satellite-borne magnetometers gave the final story, with the field configuration I've sketched earlier.

The coverage of the large polar cap area with solar protons is in sharp contrast with the narrow latitudinal coverage of the auroral zones by energetic electrons; beyond that, there is the difference in levels of absorption, tens of dB on 30 MHz for solar protons as compared to a few dB for the auroral electrons. So all in all, solar proton events that reach the ionosphere, so-called polar cap absorption (PCA) events, can be devastating when it comes to propagation across the high latitudes.

But there are few more aspects to PCAs to think about. For example, the access for solar protons to the polar cap is one thing but it has been found that solar protons can get into the magnetosphere via the magnetotail. And the access to the two polar caps is not always equal for solar protons, judging by satellite data. So there can be different ionospheric reports from the two polar caps, depending on sunlight on each and the access of the protons. All this makes propagation interesting and confusing!

When it comes to ham radio propagation, there is a propagation effect that can mask the access to the polar caps. Here, I refer to the fact that there is a reduction in ionospheric absorption in darkness, the number of dB in absorption going down by a factor the order of 5 or so. This is due to the fact that the electrons created by solar protons may attach themselves to oxygen molecules and form negative ions. Negative ions are so massive that they do not participate in the absorption process. So absorption in a darkened polar cap, at night or in winter, is less and might be interpreted as a low proton flux without satellite data to clarify the situation.

Riometer recording of auroral absorption events during substorms that occured on October 2, 1998. Documents DCS/IRP Group.

The electrons bound in negative ions are released when sunlight is restored to the D-region. That is the case for proton events but not for auroral electron events where the ionization is at much higher altitudes and electron detachment results from collisions with atomic oxygen, abundant above 100 km. So auroral absorption (AA) events do not show any day/night effect like PCA events.

To summarize now and put things in perspective: auroral absorption events are limited in time and space, found during magnetic disturbances, large or small. Polar cap absorption covers a wide range of latitudes, the whole polar cap, and can last for days at a time after some solar flares. And the ionospheric absorption is large, making PCAs a real threat to ham radio communications. And if the polar cap expands in size in the late phase of a magnetic storm, solar protons can then reach down to much lower latitudes and have even greater effects of our HF propagation.

The beauty of PCAs, if one would call it that, is that they are relatively infrequent. The real threat to ham radio communication is the effects of the solar wind, so I would say that magnetic storming is the thing to watch out for, by logging K-and A-indices to identify any possible repetitions and then by checking each day by whatever means are available. Magnetic storming is THE threat to our peace and quiet; what the sun provides in the way of higher critical frequencies by UV radiation can be taken away in a jiffy by a blast of the solar wind triggering a magnetic storm, minor or major.

So monitor/log the magnetic indices; they hold the key to success in high latitude DXing on the bands! But when the high latitudes are disrupted, try the other directions, say across the equator. That is pretty safe, the field lines there being shielded from the ravages of the solar wind. And there's a lot of rare DX there to make things interesting.

Polar Cap Absorption riometer profile recorded in Finland (69N) on April 21, 1998. The absorption increases when the D-region is exposed to solar UV radiation, and this day/night transition is a key feature for identifying PCA in riometer data. PCA events are normally very strong, often > 3 dB for sustained periods. Auroral absorptions are much less intenses. Documents DCS/IRP Group.

This is the end of the line and time to wrap up the discussion. It should be in two parts, the theoretical side which we compare with the experimental part. In regard to theory, the most general discussion would be one which uses ray-tracing with the best available model for the ionosphere and geomagnetic field. That is simple to say but as you know, words come easy. But let's look at how it's done and what it means to us. Then we can go to the experimental part.

Appleton's magneto-ionic theory

Now it may sound strange but the magneto-ionic theory that I mentioned earlier is all cast in terms of frequencies. Obviously, the operating frequency is of utmost importance. But then there are three other frequencies; how they compare with the operating frequency (QRG) determines features of propagation.

The first frequency is the plasma frequency; for a given position in the ionosphere, it is another way of specifying the electron density. Plasma frequencies in the lower ionosphere increase with height, up to the F-region peak, and decrease with latitude toward the poles. And, in a complicated way, they depend on the earth's magnetic field and sunlight. But for signals to be contained, not penetrating into the topside of the ionosphere, their effective vertical frequency (EVF) must be less than the plasma frequency at the peak of the F-region.

The second frequency is the collision frequency Fc between electrons and the neutral constituents which surround them. As you know, collision frequencies Fc determine ionospheric absorption and are greatest (<2 MHz region) in the lower ionosphere. The comparison of interest is the operating frequency QRG and Fc. If QRG >> Fc, then ionospheric absorption is not of great importance. And a good example of that would be up on the 10 meter band. But the plasma frequency is still of great importance as well as sunlight on a path.

Electron gyro-frequency between 630 to 1630 kHz. It is correlated with the lines of the geomagentic field and MUF and affects the top band propagation.

The third frequency is the electron gyro-frequency Fg, the number of times per second an electron goes around the local field lines. For the geomagnetic field, that ranges from 0.6 to about 1.6 MHz, in going from low latitudes to polar regions as displayed at left. And the comparison between QRG and Fg becomes very important down on the 160 meter band as 1.8 MHz is comparable to values of Fg along a path. The consequences of including the geomagnetic field in ionospheric theory are very important and should not be overlooked in thinking about propagation.

Before getting to them, we should recognize that geomagnetic effects have been neglected in almost all the discussion so far. True, it was pointed out that the earth's field serves to keep ionospheric electrons from running away, once released, but that was about it. So for most amateurs, theory is quite simple: some ionospheric absorption on the lower bands but otherwise, RF is linearly polarized, depending on the transmitting antenna.

But all that changed when Appleton embarked on formulating a more general theory which included the geomagnetic field. The results are not to difficult to obtain but hard to comprehend, given that the earlier theory is so deeply ingrained in our thinking. But let's take a look at a few of them and see how things go.

First, the strength and direction of the local magnetic field is important and propagation depends on the direction of wave travel relative to the magnetic field. That is a new idea to most hams but is the case as in the more general theory, RF waves are now elliptically polarized, depending on the direction of propagation. That may be hard to picture so think of a wave moving along with its E-field vector going around the direction of propagation but with varying amplitude as its tip traces out an ellipse.

Not only are waves elliptically polarized but there are two types, depending on the direction of rotation of the electric field - ordinary and extra-ordinary waves. The two waves propagate with different speeds and, oddly enough, are absorbed in the ionosphere (remember the collision frequency?) at different rates.

Rather than leaving things as they stand at this point, it should be noted that the wave polarizations go over to simpler cases when propagation is along or perpendicular to the field direction. To use modern advertising parlance, there are also cases in the "not exactly" category, quasi-longitudinal and quasi-transverse propagation where the waves are close to, but "not exactly", the strict limits mentioned above. That makes magneto-ionic theory less stern and forbidding as the elliptically polarized waves are close to circular or linear in those cases.

That is a brief summary of what happens to RF when the QRG is comparable to the electron gyro-frequency, say around 1.8 MHz. Added to that is the idea of limiting polarizations where RF enters or leaves the lower ionosphere. So there could be a mis-match between wave polarization at launch and the limiting polarization at the bottom of the D-region. In that case, the mis-match between the two polarizations means the coupling of RF into the ionosphere is less than 100 %. That is part of the "bad news" at the low end of the amateur spectrum. Of course, there is also the question of the how the polarization of the emerging wave matches that of the receiving antenna. And the other "bad news" is one mode, the extra-ordinary polarization, is heavily absorbed over distance, meaning that more power could be lost from that effect.

All this emerged when Appleton worked through the more general theory of how ionospheric electrons respond to RF in the presence of the geomagnetic field. Once that is done, the next step is to incorporate the results into the "equations of motion" for waves and do ray-tracing with the best field model available. The consequences are interesting, as you can imagine, with the important result that ducting is possible just with the typical electron density gradients present in the ionosphere.

All this is probably more than you wanted to read about but you should know that the simple ideas that are abroad are not the final story. But one idea from magneto-ionic theory that applies at frequencies way beyond the electron gyro-frequency is the rotation of the plane of wave polarization.

The Faraday rotation can be measured in placing a linear polarizer on each end of a solenoid containing a transparent material and cross them at 90. This rotation angle can be measured as a function of magnetic field, length of sample and wavelength of light.

Ordinarily, changes in HF polarization are attributed to ionospheric tilts, not an effect from the magnetic field. But it is real, seen with satellites on VHF.

The idea comes from sending linearly-polarized signals along the field direction. If you think about it, a linearly-polarized wave is the same as the sum of two circularly polarized-waves of equal amplitude but rotating in opposite directions. The rest is straight-forward as the two circular polarized waves travel with different speeds, meaning that one gets ahead of the other, and the polarization of the resultant linearly-polarized wave is rotated as it travels along. That is Faraday Rotation and is an important part of work on VHF where two circular polarizations can be present with essentially equal amplitudes.

But a problem with Faraday Rotation comes up on the lower bands as the extra-ordinary wave is heavily absorbed and over any great distance, the ordinary wave is the only one that survives. So it is not so much a question of Faraday Rotation on 1.8 MHz but one of the remaining ordinary polarization and how it compares with the limiting polarizations at the bottom of the ionosphere and antenna polarizations.

As for the experimental side, that really deals with what we know about our surroundings. Starting from the ground and going up - the geomagnetic field, the neutral atmosphere, how solar radiation affects the atmosphere and creates the ionosphere, the solar wind and its effects on (or in) the earth's field, the solar magnetic field and solar activity. There's a lot to know and more to the point, it's important to appreciate that we're dealing with a coupled system. So any effect that is dealt with in isolation may not be well understood.

The present situation as far as propagation is concerned depends on the use of computers and that brings up the question about the programs that are available. For the geomagnetic field, there is the International Geomagnetic Reference Field (IGRF) while the models of the ionosphere are found in the Internation Reference Ionosphere (IRI-2001 at U.Leicester and at NSSDC). Those two serve as research sources but also find their way into software such as PropLab Pro or DXAtlas.

Then there are also the various propagation programs that are available at present. Viewed by themselves, they are efforts done in isolation with quiet-day representations of the ionosphere. So additional consideration must be given to the details of the critical frequencies all along a path and also the geomagnetic circumstances and any unusual ionization, say from solar protons. That's where mapping programs and the SWPC websites on the Internet prove their value. Without using that information, it is hardly possible to make a realistic prediction of anything.

As an example, the week of Nov. 8-14 was characterized as one of considerable magnetic activity and solar activity. Thus, the following A-indices were reported from the Boulder magnetometer: Sun: 68, Mon: 78, Tues: 6, Wed: 4, Thurs: 4, Fri: 60, Sat: 38

Without that knowledge, the results for propagation conditions from a computer program, using only input with regard to sunspot counts, would make you think you live on a different planet as they would have little bearing on actual conditions.

Top 50 solar flares, Spaceweather

At left, a picture recorded by SDO in EUV at 131 and 171 of a chromospheric eruption of class X9.3 occuring on the Sun on September 6, 2017 between 11:53 and 12:23 UT in the active region AR 2673. This was the strongest since September 7, 2005 (class X18). It was associated with the emission of several CME and several other sun flares in the next days. At right, a coronal mass ejection (CME) disturbs solar wind currents and creates magnetic disturbances that hit sometimes the earth in a catastrophic way. The Wide  Field and Spectrometric coronograph LASCO onboard the SOHO satellite has observed many CMEs. The spectacular event of September 12, 2000 displays above created a halo event giving the feeling that the whole sun was surrounded by the CME. Such halos are generated by sunflares (eruptions) directed toward Earth. Click on the image to run the animation (.GIF of 595 KB).

But that is not the whole story as the CME that was responsible for the magnetic activity also produced a solar proton event on November 14. Then, 10 MeV protons, which are capable of reaching the ionosphere, appeared at satellite altitudes around 0600 GMT. The proton flux peaked at 300 p.f.u. (proton flux units or protons/sq-cm/sec/ster) around 1245 GMT and continued coming out of the interplanetary field for more than a day. Also, there was a weak flux (6 p.f.u.) of 100 MeV protons, capable of reaching balloon altitudes (about 30 km), was present. In addition, there was a strong increase in 1-8 A X-ray background on the 13th.

As I said, these are coupled systems and we have to look at more than one limited aspect if propagation is really our interest. Of course, as we go toward solar maximum, this will be the case more and more often. But on the cheery side, the week of Nov. 8-14 has to be an exception. For example, in the year that I spent in my long-path study around the maximum in Cycle 22 , something like 80 % of the days were free of any significant disturbance and even with minor or major disturbances on the rest of the days, I was able to make a long-path contact on over 90 % of the days.

That suggests a cautious but optimistic approach is called for, watching all the disturbance indicators on a regular basis, "going for it" when propagation looks good and even "looking around" when conditions may not be the most promising. I like to say "DXing is an intellectual pursuit" so it's worth a bit of study; that makes the rewards all the more enjoyable.


I think I've said all I wanted to so let me close with words of a great man that I'm sure you'll recognize: "That's all folks!"


Bob Brown, NM7M (sk)

For more information

On this site

PDF version of this document (for printing purpose, without images)

Real-time status of solar, geomagnetic and auroral activities

What can we expect from a HF propagation model ?

Review of HF propagation prediction programs

On the web


An Introduction to HF propagation and the Ionosphere, ZL1BPU

Radio wave propagation (chapter 2), TPUB

HF Radio Propagation Primer, by AE4RV (Flash presentation)

ON5AU's Propagation pages, Marcel De Canck

Introduction to HF Radio propagation, PDF file from IPS, released in HTML format at G3YRC Radio Club or N1QS

Propagation Studies, RSGB

ARRL' Shop

RSGB' Shop

The DX Magazine


Propagation and Radio Science, Eric Nichols (KL7AJ), ARRL, 2015

The High-Latitude Ionosphere and its Effects on Radio Propagation, R.D.Hunsucker/J.K.Hargreaves, Cambridge University Press, 2007

Physics of the Upper Polar Atmosphere, by A. Brekke, John Wiley & Sons Inc, 1997

The Little Pistol's Guide to HF Propagation, by Robert R. Brown (NM7M), Worldradio Books, 1996

The New Shortwave Propagation Handbook, by Jacobs, Cohen and Rose, CQ Communications, Inc., 1995

Radio Amateurs Guide to the Ionosphere, by Leo F. McNamara, Krieger Publ.Corp.,1994

Ionospheric Radio (IEE Electromagnetic Waves Series, Vol. 31) by K.Davies, Inspec/Iee, 1990

Radio Wave Propagation (HF Bands): Radio Amateur's Guide, by F.Judd, Butterworth-Heinemann; 1987.

Back to Menu

Page 1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 -

Back to:


Copyright & FAQ