LF Propagation and some effects of

Solar Disturbances

It is generally believed that propagation in the LF bands is stable and almost boring. This is only loosely the case. Historically the LF bands have been used were a wide service area is required which is relatively immune from interference and ionospheric effects such as fading. A lot of the research that has been carried out has had as its objective a determination of the severity of interference from signals beyond normal service range. In the years after WWII this was more to determine the reliable range of long distance navigation systems such as Decca Navigator, Loran-C and Omega. The criterial factor being the reliability with which the phase or timing of the signal could be determined. I have tried to approach my study from the standpoint of the ordinary radio amateur who is little concerned with service reliability. The Radio Amateur is willing to wait for and use whatever transient phenomena is available to achieve his ambition of long distance communication.

Data generally available

A study of many books will lead you to believe that a signal will increase in strength at night and is absorbed in daytime ....end of story. An HF forecaster has been quoted as saying that "there is nothing to predict about LF propagation". This has to be weighed against the particular interests of the professional workers. Most of them are involved in Navigation systems or Broadcasting. As a result their main interest in LF is a range out to about 1000kms, or at most 1500kms. Signals from beyond that distance are regarded as potential sources of interference, and as such the significance of these signals has a different value to ours, where we consider the distant signals as our goal. They are generally only really interested in calculating whether the distant signal poses an interference threat within their normal service area, and they wish to know how many nights a year this is likely to happen. As a result of the calculation they set the design megawattage of their transmitter to ensure the desired service area is attained. There is also the question of the level of interference they will provide to other stations sharing their frequency but outside their service area. The much quoted CCIR and ITU Recommendations are all aimed at providing the data to make reasonably meaningful calculations of these factors, and which can be accepted across the the world for frequency planning. The same situation occurs with those involved in Naviagation systems. These operate by enabling accurate phase (or timing) measurements to be made between pairs of transmitters. It has been found that the phase shifts caused by ionospheric reflection can destroy the accuracy outside the 1000kms range. Thus a lot of work was done in the 1950s and 1960s to determine what the maximum spacing between Nav. base stations could be, giving value for money without compromising their accuracy. Thus whilst the formulae in the learned papers are interesting they should only be regarded as a guide to what signal strengths may be achieved over the calculated path. They can be many tens of dB in error as far as determining whether a path is workable by amateur operators.

So what is the true situation as far as amateur operation is concerned?? The following plot show an accumulation of signal strength data for DCF39 ( 138.83kHz ) from about 3 weeks in Septemeber 2001. The plot is of average and peak signal strengths over that period. It was recorded by Brian CT1DRP, who is located near Porto in northern Portugal, and has a path of about 1950kms to the transmitter at Magdeburg.

There is 12dB enhancement between 0800z and 1100z, and there can be another 10dB enhancement at best. Night-time signals are 20dB better than day-time and can be enhanced by a further 10dB at best.

Now who says there is nothing interesting in LF propagation ??

Normal (quiet sun) Propagation

I need to define that I mean frequencies between 50kHz and 150kHz when I refer to LF signals. This part of the spectrum is agreed to behave differently to both the region below and the region immediately above this band.

Most radio amateurs will be familiar with the classification of the ionosphere into the D, E, and F layers, with the F layer at about 250kms altitude, being responsible for the DX signals and the D layer at 50-90kms for the daytime absorbing curtain below about 5MHz. The Ionosphere does not in practice exist as discrete 'layers' these are rather definitions of regions of altitude, and relate to levels between which the ionising processes change. When I started this study, I naively thought the E-layer was a daytime layer and that LF signals must see the F layer at night. In fact the E-layer is ionised day and night, and LF signals cannot penetrate it twice whilst retaining enough strength to be detected at extreme distance..

Most books about propagation will tell you that the sunlight ionises the D-layer during the day to make it absorbing so no signals will propagate. At night the ionisation decays and the ‘reflecting’ level occurs to about 100kms altitude (which is the bottom edge of the E-layer) to give ionospheric wave propagation at considerable strengths. It is very repeatable and very reliable. Much of the pre 1930s radio communication was based on this. Only frequencies below my designated lower frequency of 50kHz propagate by the often quoted "Waveguide Mode". Whilst there may be some similarities with the the 165-195kHz band allocated to our Antipodean friends, there can also be differences. In many articles on VLF/LF propagation, this band of 50 - 150kHz is often isolated for its own consideration.

We need to understand a little more about the ionisation processes to appreciate the changes that occur. These processes are very complex and I think a lot of professional Ionospheric researchers would admit that they still do not know precisely about all the processes involved. Most of what has been learnt has been as a result of satellite and rocket technology in the last 25 years. I will concentrate on the D and E-layers as those are the ones that will affect us most, and I will use a simplified model.

Ionisation is a term which describes the removal or addition of an electron to an otherwise neutral atom. This can be accomplished by the absorption of radiation if the energy is high enough or by bombardment by particles from space (Cosmic Rays). Once ionised, the lifetime of the ion depends mainly upon the gas pressure, since this determines the likelihood of a collision with another atom or ion, for that will either capture the electron, or neutralise the ion. At higher altitudes, where the pressure is correspondingly lower, ions and electrons have a longer ‘life’. At low altitudes (such as the D-layer) the pressure is such that when the main ionising source is removed (sunset) the ions rapidly recombine. It turns out that the radiation reaching the D-layer is mainly near-UV and it only has enough energy to ionise oxides of nitrogen, but not air molecules themselves. The higher energy radiation has been absorbed in ionising the F and E layers, where molecules of oxygen and nitrogen are ionised. There is also a substantial contribution to the ionisation of the E-layer from Cosmic rays which accounts for its all-day existence.

The band of ionisation we are interested in has different properties dependent upon the density of the ionisaton, which may be defined as the number of ‘free’ electrons per cubic metre. At low electron density the region is fairly transparent to radio waves but provides a little attenuation or ‘absorption’. At medium levels of electron density the region exhibits extremely high attenuation, and at very high levels of ionisation the region starts to ‘reflect’ the radio waves. I have puts quotes round the word reflect as the mechanism is not that of an optical mirror but is more a gentle bending of the direction of the waves known as refraction. The bending is a function of the changing refractive index of the region which is related to the electron density. For those of you who have come across the term refractive index in optics, this means that the refractive index of the ionised layer is less than unity and in fact is even negative.

What is the result of all this in a normal, disturbance-free day? I believe that as the daylight solar radiation penetrates to the D-Layer it starts to increase the electron density by photo-dissociation. The initial effect of this is an increase in absorption and this may account for the ‘morning dip’ ( 0630-0800z on the plot above) effect particularly on predominantly north-south paths. Eventually a point is reached at a level below the E-layer where the density becomes sufficient for 'reflection'. As the sun rises, the radiation from the sun enters at a higher angle, increasing the ionosation in this region. The result is that the ‘reflecting’ layer moves downwards, reducing the thickness of absorbing material the signal must traverse before it is turned back towards the ground. Thus the received signal strength increases. This is why there is a small amount of ‘skywave’ signal present in daytime on longer paths. This is D-layer propagation from an altitude of about 50kms !! As evening approaches and the angle of the sun’s radiation becomes more acute the level of ionisation decreases. The ‘reflecting’ region moves up in altitude, leaving an increasingly thick region of absorbing material below it. This leads to a dip in signal strength before the night-time conditions take over. In darkness the ionisation in the D-layer rapidly recombines, leaving a relatively low attenuation path below the ‘reflecting’ layer, which I believe is around the lower edge of the E-layer at about 95 to 100kms altitude.

Solar disturbances

X-ray Flares

These are emitted from areas of the sun’s surface that are usually referred to as ‘active’ regions and are sequentially numbered in the NOAA data. They correspond to ‘spots’ where, due to the violent movement of the surface, the magnetic field has become contorted and very ‘strained’. Occasionally the strain is suddenly relieved and there is a violent eruption of material and radiation. Although the event is referred to as an X-ray flare, a whole spectrum of electromagnetic radiation is produced. A real time plot is available on the NOAA SEC web-site http:/www.sec.noaa.gov/today.html covering either 3 hours or the last three days.

Proton events

Protons are hydrogen nuclei, the major component of the sun. In a disturbance these can be ejected from the sun at enormous speeds, such that they reach earth within about 1 to 3 hours.

Coronal Mass Ejections

When the solar magnetic field gets ‘knotted’ up and suddenly attempts to straighten out, large quantities of corona are hurled up from the surface. Sometimes these follow the magnetic lines of force and form spectacular loops known as ‘prominences’ If the field lines do not return to the solar surface the coronal material is flung into space. It consists of a large cloud of very hot plasma (totally ionised gas, mainly hydrogen) and because it is a swirling mass of moving charges it carries with it a magnetic field.

Coronal Hole events

These are another source of plasma ‘globs’ that are flung off the surface by magnetic anomalies in the coronal region above the suns surface.

The Solar Wind

As the sun spins a small but steady stream of material in the form of ionised particles is thrown off and passes the earth on its way through the solar system. High wind velocities can lead to minor geomagnetic events, and the wind material often gets caught up and swept along with the much more rapidly moving coronal mass ejecta.

LF Radio effects

X-ray Flares

These have been reported by several people as producing an enhancement to signals at LF, whereas they often cause a blackout at MF and HF. The enhancement only occurs in the region of the ionosphere that is in daylight, and the effect is limited to the time of the event. The shape of the signal enhancement graph follows the X-ray flux graph very closely. The normal event will last from 10 minutes to, perhaps half an hour, tailing away slowly. These events cannot be accurately predicted, but they are most likely when active areas are near the centre of the solar disc. NOAA space weather centre will forecast the likelihood of a major flare. The level of enhancement in daytime makes the flares well worth searching for. It is possible to achieve an enhancement of up to 10dB above normal conditions for a period of about 15 to 20 minutes for a X-Class flare.

The intense burst of Ultra Violet radiation that accompanies the X-ray flare ionises much more of the D-layer. Possibly pushing the reflection level down as low as 50kms. There is now no absorbing region below the reflection height so the received signal is enhanced significantly. At low altitudes the lifetime of ions and electrons is small, so that as the intensity of the UV declines the level of ionisation declines and the reflection layer moves back up again and the signal must now pass through an absorbing layer again before it can reach the reflection level. The received signal strength drops slowly back to normal closely following the shape of the flare radiation flux. (see the NOAA GEOS satellite X-ray flux plots)

Plasma events (CME and Coronal Hole)

The glob of plasma carrying a strong magnetic field eventually collides with the Earth’s magnetosphere (provided that the glob was ejected in our direction...they are not always sent Earthwards ). The effect when this happens depends upon the orientation of the magnetic field carried by the plasma with respect to the Earths magnetic field. It is like the game children play with small permanent magnets. In one direction the glob merely bounces off the earths field. This probably gives rise to some minor geomagnetic changes as the earth's field is distorted by the collision. If the field is the the opposite direction, like magnets attracting, the lines of force of the plasma 'connect' with the Earth's lines of force and this enables streams of highly energetic particles from the plasma to be injected into the atmosphere at the two poles. These are the particles that cause aurora, and the the resultant shock to the earth field is referred to as a Magnetic Storm. A measure of the severity of the the magnetic disturbance is given in the A and K indexes. These are well described in some of the references, and in a tutorial available on the NOAA SEC web site. The magnetic storm has no immediate effect at LF though there may be some auroral effects at high magnetic latitudes. The real effect at LF occurs about 2 to 4 days after the collision with the plasma ‘glob’ when the injected particles have had time to diffuse down to the ionophere, in particular the D-layer, and spread out from the poles to lower latitudes. We do get a warning of plasma events as they are observed in the optical wavelengths, and the effect usually takes up to 72 hours to reach us. The effect is very complex but in general large CMEs, which produce the biggest geomagnetic disturbance, have the most profound effect on LF propagation conditions. The first noticeable effects are increased day-time signal levels from distant stations (at path-lengths greater than 1000kms). At the same time the night-time levels become depressed. The length of the period for which propagation conditions are affected is very much dependent on the severity of the event. The effects of minor storms ( Kp = 4 or 5 ) will dissipate in two or three days, whilst the after-effects of major geomagnetic storms ( Kp = 7 to 9 ) may persist for as much as 10 to 14 days.

According to the professionally published work, the delayed effect is due to the time taken for the charged particles to diffuse down through the upper atmosphere and spread outwards from the two poles.(More recent sudies I have found modify these ideas, suggesting the pool of energetic ions is held in a region called the equatorial ring current and is slowly dribbled into the ionosphere on the daylight side. ). I believe that that these ions add to those normally generated by sunlight, leading to a very much higher electron density lower in the D-region in daytime. This means that the 'reflection' level moves downwards, reducing the amount amount of absorption the signal experiences during its passage to the 'reflection' level. This shows as enhanced daytime levels for distant stations.

At night, the ionisation of the D-layer normally decays as the layer moves into shadow, this leaves a reflection height at 90 to 100kms altitude with little absorbing material for the signal to pass through. The extra injected charges have a longer life, possibly because they are very much ‘hotter’ that the normal photo-dissociated ions, they may affect the reflection height but they do leave a thick band of absorbing medium which persists all night, and that the signals have to pass through on their way to and from the reflecting level. Hence the lower levels of distant night-time signals, with two hop path signals suffering more attenuation than single hop paths..

There is a progression of radio effects as the extra iononisation slowly decays away (see comment above), which is seen most clearly on the Trans-Atlantic path. I divide the 'after-effects' into three distinctly recognisable phases, the first is severe absortion at night, greatly attenuating distant signals, starting 2-4 days after the peak of the storm.

The second phase starts to show as the high absorption levels die away, and distant signals show severe and very rapid fading. The fading period can be as little as five minutes from trough to trough, and the troughs can be as deep as 30dB below normal levels. This fading may be caused by reflections of different portions of the signal from regions at different levels in the layer. At critical electron density levels it may be possible for one part of a wavefront approaching steeply to pass through a reflecting altitude whilst a wave closer to grazing incidence will be reflected. Over relatively short paths like the 1950kms path from DCF39 to CT1DRP it is difficult to understand the mechanisms, but it is essentially multipath optical interference. There are reports and ionograms from sounding stations showing two distinct reflecting levels at LF.

It is noticeable during this phase of the event that from night to night the fading period starts to lengthen, until in the final phase the fading period is about 40minutes trough to trough, and the troughs are normally not as deep. During this final phase the constructive (optical) interference between signals reflecting from different heights can enhance a distant signal by up to 12 dB. This effect is very dependent on the geography of the path and whilst one station may see an big enhancement another (maybe as little as 30kms away) may have a disappointing signal level.

The above series of plots graph conditions triggered by Severe geomagnetic storms on the 3rd, 4th and 5th of October and is complicated by a minor storm occurring on the 14th before the conditions had fully recovered. The plot below for another sequence in December shows the height that levels can reach towards the end of these 'storm after-effects'

Proton events

As far as I am concerned the jury is still out on this one. I believe that I have seen good conditions on the evening of a major flare where there has been a Proton Event. It may just be that, at an active period of the sun, the best conditions are as long after the last geomagnetic storm as possible. Proton events are known to increase the absorption at Medium Frequencies in the BC band (see KN4LF site), which suggests that they could also provide the right input to enhance night-time signals at LF. The situation is so complicated near the solar maximum that we may have to wait for quieter conditions to really understand this effect.

Monitoring SXV and CFH

The equipment used to monitor SXV and CFH at my location is described in another short page (here). The equipment and procedures used by Brian CT1DRP to monitor a number of frequencies including DCF39, SXV and CFH, over a continuous period are described on his web site which is linked from my other page.

Results and Inferences

The plots of DCF39 from Brian CT1DRP show very clearly the enhancement in daytime due to X-ray flare activity. As this is an electro-magnetic radiation burst, it will only affect the area of the Earth which is in daylight. These events could provide enhanced paths for daytime DX into Europe. The daytime enhancement due to ions (electrons) injected into the D-layer during geomagnetic storms can provide a stable path enhanced by up to 10dB during the middle of the day. Several declared 'Good Weekends' during the autumn of 2001 have been correlated to this effect.

I have often referred to a 'morning dip' in the signals, this is most noticeable on paths with a predominantly north-south orientation. It can clearly be seen in the plot of the average level of DCF39 from CT1DRP above. I believe that as the sun rises it first grazes the earth and reaches the D-layer from underneath. This is a little surprising but the geometry is correct and the effect was described by Bob Brown in a paper in QEX on the effects of ozone. This means that the initially weak rays start to ionise the D-layer from below what is at that time the reflection level (probably 100km altitude, at the lower edge of the E-layer). The rays this early are weak and some of their strength is absorbed by two passages through the ozone layer, but they are strong enough to start producing absorbing levels of ionisation. Then as the sun rises and starts to illuminate the D-layer from above the enhanced ionisation drives the reflection level downwards, reducing the amount of absorbing medium that the radio wave must pass through. So the signal level rises and the daytime skywave strength will peak at (solar) mid-day at the centre of the path. After this the strength of the solar radiation decreases and the 'reflection' level rises, leaving an thickening layer of absorbing medium below it. So the signal level slowly declines up to dusk, when the absorbing medium decays and the refection level moves up to around the lower edge of the E-layer at 100kms altitude and the signal levels rise to their best levels at distances over about 700 kms. Thus just after dawn is not the best time for DX on LF, particularly on N-S paths, even though it may sound quiet. That is because all the skywave noise and interference is being absorbed as well.

Propagation will nearly always be better at night, particularly at extreme range, and when the absorption is low. The best conditions have seemed to be around about 14 days after a big event when the effects in the ionosphere are dying away but there is still some multiple path propagation to elevate signal levels by constructive (optical) interference. On the best nights trans-Atlantic communication is possible with QRSS (3 second dot ) morse for stations running in excess of about 250mW ERP, and I feel it is only a matter of time before conventional qsos are made, maybe on the tail of a QRSS or JASON contact.

Some ideas on Transatlantic Propagation Modes

The path of a radio signal is often likened to a tennis ball bouncing between the ionosphere and the ground. I suspect it is not quite like that for signals at extreme range. If you consider a signal leaving the transmitter site at 0 degrees (i.e. horizontal) it will eventually reach the ionosphere and be turned back towards the ground. At extreme range it will approach the ground at grazing incidence, almost horizontal. Thus it does not need to 'reflect' from the ground. That part of the wavefront not contacting the ground will continue on, beyond the 'point of contact' upwards to another meeting with the ionosphere. A very simple geometric calculation on this suggests that for a 'reflecting layer at 100kms the first 'hop' is about 2400kms. This model of a 'mirror' reflection is not accurate and the wavefront may stay within the ionosphere for anything upwards of 1000kms being slowly refracted groundwards. Thus there could be some signal on the transatlantic path which arrives (in the UK anyway) by a one hop path. This may sound a bit like heresy but there is some evidence to support it. On many occasions the signal from CFH rises to a peak and then drops sharply again before rising to the full strength for that night. The position of this first peak corresponds well with the position of the solar shadow, at 100kms altitude, reaching the mid-path point (i.e. opening a one hop path) Note that at this time the position of the first 'reflection' point nearer to Canada is still in full daylight. The path reaches full strength as the shadow reaches a point about 1200kms off the Nova Scotia coast so that the eastward three quarters of the path is now in darkness. Thus allowing the two hop mode to propagate. This is much stronger than the one hop mode due to scattering and diffraction effects. The dip is due to destructive interference as the two hop signal reaches the same strength as the single hop signal, before increasing to almost swamp the single hop.. It is coincidental that at this time the evening shadow is just reaching the Nova Scotia coast at ground level. The simple model is complicated by the ground following trait of radio waves. There is good published evidence that the effective takeoff angle can be as low as -9 degrees, due its following the earth's curvature as it leaves the transmit site.

If the two hop signal swamps the single hop signal why do we get deep fading occasionally on the transtlantic path? I believe I can explain this in terms of the extra persistent ionisation of the D-layer after a geomagnetic event. The geomagnetic storms inject hot ions into the ionosphere which do not decay as quickly as the UV photo-dissociated ions. These hot ions increase the absorption in the D-layer at night. This acts to favour the single hop signal, which only passes through the absorbing material once, whilst the two hop signal must get two doses of absorbtion. The result is the two path are much nearer to similar strengths at the receiving site, and thus the fades are deep and narrow. The longest steady conditions are when the sun is quiet and the two hop path prevails, but when the one hop path is reasonably strong constructive interference between the two paths can give a 6dB improvement over the normal levels. This occurs on the tail of a geomagnetic-storm stimulated disturbance, just before things return to 'normal'.

This idea is one for 'Peer Review'.... it is set up to be shot at!! BUT please make sure you have good data, because I have quite a lot that needs to be explained some way.

Acknowledgements

This study would not have taken place but for the enthusiasm and support of a number of LF amateurs. In particular Vaino Lehtoranta OH2LX who 'goaded' me into setting up the computer driven ADC data logger, and endured many long e-mails as I struggled to understand the physics. Laurie Mayhead G3AQC who exchanged many plots from his chart recording system with me in the early days and who provided the first review paper whose references enabled me to chase through a lot of the major professional published work on LF Propagation. Out of all those papers the work of Jack Belrose VE2CV shines out as solidly practical, and a great example to try to follow. Rik Strobbe ON7YD for hosting the WEB posting of the early daily CFH and SXV plots. Markus Vester DF6NM for doing some interesting analysis on the CFH data. Whilst the early work was done on CFH from my home near Ipswich (JO02PB), it would all have stalled in mid 2001 when SXV stopped it regular transmissions but for the sterling work of Wolf DL6YHF and Brian CT1DRP. Wolf liaised with Brian to include textural data-logging in SpectrumLab, and Brian has forwarded me the 24 hour data-log of DCF39 every morning for nearly a year now. These 24 hour plots of DCF39 have enabled be to correlate the X-ray effects and geomagnetic effects that give daytime enhancement, as well as continuing the study of the night-time paths. Finally to Mike G3XDV and Dave G3YXM ( and many others) for their encouragement, and interest. It was my contention from the early days of my involvement with 136kHz that Trans-Atlantic communication should be possible. It has been a great pleasure to see the successes of the past 18 months and to feel that I have had some small part in it all.

References

General solar weather
Burch J,, " The Fury of Space Storms ", Scientific American, Apr. 2001, pp 72-80

Luhmann J. " Space weather: physics and forecasts" , Physics World, July 2000, pp 31-36

"A Primer on the Space Environment", N.O.A.A. S.E.C available on http://www.sec.noaa.gov/primer/primer.html

Williams. G., "Propagation at the Solar Maximum" , (RSGB) RadCom, Jan. 2000, pp 27-29

Newton C., "Solar Data Explained" (RSGB) Radio Communication, Nov. 1997, 38 -39

LF Propagation
Burgess b. and Jones T.B.,"The propagation of l.f. and v.l.f. radio waves with reference to some systems applications", The Radio and Electronic Engineer, Vol.45, No.1/2, Jan/Feb 1975

Belrose J.S. and Thomas L., "Ionisation changes in the middle lattitude D-reagion associated with geomagnetic storms", J.of Atmos. and Terr. Physics, 1968, Vol. 30, pp 1397-1413.

R.R.Brown, NM7M, " More on Atmospheric Ozone and Low Frequency propagation", QEX (ARRL), Jan/Feb 2001, pp28-36

Lauter E.A. "A Survey A 3 - Absorption Measurements at Low and Medium Frequencies", Annal. Geophys., Vol. 22, 1966, pp 289-299

Belrose J.S. " Low and very low radio wave propagation" , AGARD Lecture series XXIX 'Radio Wave propagation' pp IV-I to IV-115 (Technical Editing and Reproduction, London, 1968)

Reports and Recommendations of the CCIR, 1990 Vol. VI, " Propagation in Ionized Media "

and many others that can be referenced from the above papers.

return to main page