You are on page 1of 20

50 MHz Long-Path Propagation

Jim Kennedy, KH6/K6MIO
Introduction 144 MHz, and has occurred more rarely
Operation over long distances on six at 222 and even 432 MHz.
metres can be very challenging. The fact The first recognition of this effect on
that F-layer ionospheric propagation is six metres appears to have occurred in
relatively rare provides some interesting late August 1947, near the peak of Solar
opportunities to observe some propa- Cycle 18*. On the 25 August, KH6/
gation modes in isolation that, while they W7ACS at Pearl Harbor worked VK5KL
may occur more often at lower frequen- in Darwin, to set a new six-metre DX
cies, are often masked by other propa- record. Two days later XE1KE in Mexico
gation modes occurring at the same City worked a stunned LU6DO in Argen-
time. tina [2]. At about the same time, sta-
Among the most interesting cases are tions in England and the Netherlands
Transequatorial Propagation (TEP) and worked stations in Southern Rhodesia
two TEP-related beasts: Transpolar Long- and South Africa [3].
Path Propagation (TPL) and Trans-equa-
torial Long-Path Propagation (TEL).
Under suitable conditions, these latter
modes can produce spectacular six-
metre openings spanning well over half-
way around the world, if one is lucky
enough to be in the right place at the
right time.
Transequatorial Propagation – A
TEP is a propagation mode that can
allow VHF stations located in the mag-
netic tropics on one side of the earth’s
geomagnetic equator to communicate Figure 1: Path footprints for the southern
more or less on a north-south line with Africa to Mediterranean TEP path, as seen
similarly placed stations on the other side by ZE2JV and 5B4WR in 1958 [3].
(Copyright ARRL, reprinted by permission.)
of the magnetic equator, over distances
of several thousand kilometres, gener- These patterns have been observed
ally in the afternoon or evening. The repeatedly since then, especially during
ionospheric skip points are located in solar maximum. While a total mystery
the F2 layer near the equator [1]. This at first, the basic mechanism began to
effect is very well documented at 50 and come into focus as the result of amateur

* This was the first solar maximum for the six-metre band. Prior to World War II, the band was five

20 Six News
and professional studies beginning with Thus, the wave appears to skip off the
the International Geophysical Year (1957- ionosphere and then come back to earth
58), and it continues to be the subject at some distant point.
of study today. Since the ionospheric electron den-
During that time, it has become ob- sity gradually increases with height (up
vious that stations with the good fortune to a point), the skip actually occurs as a
to be located within about ±40° of the more or less gradual bending or refrac-
magnetic equator can enjoy rather good tion of the wave back around toward the
propagation, often in the dead of night, ground, rather than a discontinuous re-
with their neighbours on the opposite flection as from a mirror. But for many
side of the equator. Notably, this effect purposes, this subtlety is not too impor-
occurs for a month or two in the spring tant. However, we will come back to
and fall around the equinoxes, centred the refraction concept a little later.
on March and October. Let’s consider the effect of the elec-
Moreover, the paths need not be tron density on a wave taken in isolation
strictly north-south. South America and from other effects, such as the earth’s
Hawaii work each other frequently, and magnetic field. To do this, we look at a
South American stations also work into radio signal being sent straight up. One
Southern Europe often. While these sta- can calculate the so-called critical fre-
tions are indeed on opposite sides of quency, fc, as the highest frequency that
the magnetic equator, there is a very sig- the ionosphere can reflect the signal
nificant east-west component to these straight back down again. This critical
paths, in addition to north-south. frequency† is given by:
fc = sqrt (Ne²/4π²ε0m)
Ions and angles
To best understand the causes of TEP or
(and TPL and TEL), one should note that fc = sqrt (N) * (9x10-6)
the Maximum Usable Frequency (MUF)
on a given path depends vitally on two N is the electron number density, e is
factors, the ion density (free electrons) the electron charge, ε0 is the permittiv-
in the reflecting layer, and the angle with ity of free space, and m is the mass of
which the radio wave encounters that the electron. Except for N, everything
layer. else has a known constant value.
Ionisation – When an upward-mov- The point is that the highest fre-
ing radio wave reaches, say, the F layer, quency that will skip vertically back
the electric field in the wave forces the down is the square root of the electron
electrons in that layer into a sympathetic density times a fixed number. So, for
oscillation at the same frequency as the example, in order to skip a signal at
passing wave. The oscillating electrons twice the current maximum frequency,
in the layer can reradiate the upcoming the number of electrons must be in-
wave downward, like a static reflector. creased by a factor of four.

Strictly speaking, the earth’s magnetic field leads to two critical frequencies, fo and fx, correspond-
ing to the ‘ordinary’ and ‘extraordinary’ wave propagation modes. These differences are not impor-
tant in the current discussion.

Six News 21
But, what if the signal is sent at some
angle other than straight up?
Angle of attack – To answer this
question we must add the concept of
the angle of attack‡. This is the angle
that the direction of the moving wave
makes with respect to the ionospheric
layer. In the above example, the wave
strikes the ionosphere with an angle of
attack of 90° (i.e. going straight up verti-
cally into the horizontal layer). In the
Figure 2: M (cosec α) varies with angle of
more general case, the MUF is deter- attack. A smooth spherical ionosphere
mined by both the maximum electron gives angles between 10 and 20 degrees.
density that the wave encounters in the Tilted layers can produce lower angles,
ionosphere and the angle at which the and very exciting results.
wave hits the reflecting/refracting layer. ping stones off the surface of a lake. If
If a signal that is sent off very near you toss a rock into a lake so that it hits
the horizon (e.g. with a zero angle of ra- the water at a high angle of attack, it will
diation), due to the curvature of the iono- break the surface and sink. However, if
sphere around the earth, the signal will it hits the surface at a very shallow angle
normally hit the ionosphere at an angle of attack (grazing incidence), the rock
of attack between 10° to 20°, depend- will skip off the surface instead.
ing on the layer in question. The MUF So, in principle, the MUF approaches
in MHz (represented by fmax) can be cal- infinity as the angle of attack approaches
culated from: zero! However, such small attack an-
fmax = cosec (α) * fc gles are geometrically impossible to
achieve from a ground-based station ‘il-
or luminating’ a smooth, spherical iono-
fmax = M * sqrt (N) * (9x10-6) sphere. Under these circumstances, sim-
where α is the angle of attack. Note ple geometry would show that M H•3.4
the MUF depends on both the density at the F layer§. However, the operative
key words here are ‘smooth’ and ‘spheri-
of the electrons and the angle α.
cal’, there is a lot more to say about that.
The cosecant of α is called the ‘M
factor’. As the angle gets smaller, the How does TEP work?
cosecant gets larger. As a result, the It is not uncommon for north-south
smaller the angle of attack, the greater multi-hop F-layer openings on six me-
the MUF. tres to have no evidence of stations at
This is the radio equivalent of skip- the end of the first hop. This is often due

This is a borrowed aeronautical term. I prefer looking at the angle between the layer and the wave
direction, but physics texts normally use its cousin, the ‘angle of incidence’ – the angle between the
vertical to the layer and the wave direction (90° minus the angle of attack). The equations change a
little, but the answers are the same.
It is important to note that the angle of attack is also effected by the radiation angle of the antenna.
Hence, a low angle of radiation actually increases the MUF for a given system.

22 Six News
to an F-layer ionospheric bulge along bulge. An upcoming wave will hit the
the magnetic equator known as the equa- tilted near corner at a shallower angle of
torial anomaly. attack than it would have with the usual
Within ±20° of the earth’s magnetic spherical layer. This means that it will
equator there is a pronounced outward have a higher MUF for the same value of
bulge in the ionosphere. Though gen- electron density. The M factor is larger
erally regarded as an afternoon or early than the nominal 3.4 – perhaps by quite
evening phenomenon, it occurs at other a bit.
times as well. It is thought to be pro- The wave need not be bent all the
duced by a combination of a persistent way back toward the ground. If it is bent
thickening of the F layer near the equa- enough to cross the equator and hit the
tor, caused by the favourable angle of tilted layer on the far side, without com-
the Sun’s incoming radiation, and another ing back to earth, the second tilted sur-
effect called the afternoon fountain [4]. face may bring it down to earth again.
The afternoon fountain pumps elec- This skip from one corner of the bulge
trons upward from the E and lower F lay- to another is referred to as a ‘chordal
ers into the upper F2 region, significantly hop’. It will produce a much higher MUF
enhancing its electron density. This ef- than a traditional skip point [5].
fect appears to be the result of an inter- Since it really represents an F2-layer
action between the earth’s magnetic field ‘hop and a half’, the distance between
and west-to-east afternoon E-layer elec- endpoints can be a good deal greater
tric fields. than 5,000 km. It is also a low-loss path.
The equatorial bulge produces two Since the wave doesn’t come down at
regions, one north of the equator and the midpoint, it avoids two passes of D-
the other south, where the ionosphere is layer absorption that normal double hop
systematically tilted and the electron would have encountered.
density enhanced. This occurs at the In order for this form of propagation
points where the normally spherical iono- to function, both the north-side and
sphere is bent upwards to form the south-side tilted regions need to be ion-
ised enough to make the path work. If
either one is insufficient to skip, then the
whole path fails. Since there is little ioni-
sation margin at six metres, the best
chance for this to occur is when both
sides of the magnetic equator are equally
illuminated by solar radiation. This situ-
ation only occurs around the two equi-
noxes when the Sun is most nearly di-
rectly over the equator. As observed,
Figure 3: A diagram of a transequatorial the best months seem to be March and
chordal hop off the tilted north and south October.
skip points. These points each lie between
So, the afternoon fountain causes en-
about 0° and 20° north and south of the
earth’s magnetic equator and clearly cause hanced electron densities and tilted lay-
night time TEP in the tropics. ers to form within 20° of the magnetic

Six News 23
equator late in the day or early in the
evening. These conditions can persist
long into the night with some contacts
taking place long after local midnight
(common for South America to Hawaii).
They readily provide near grazing-inci-
dence chordal hops at six metres.
On any given day, the bulge may not
be exactly centreed on the magnetic
equator. Typically, the two corners will
be anywhere from 15° to 40° apart, and
each will be somewhere within 0° to 20°
from the equator on its respective side.
In order to access these skip points, the
Figure 4: In geomagnetic coordinates,
stations must be within one-half F hop
Hawaii is nearly due north of Australia and
of the nearest corner – this is the ‘TEP these TEP paths transverse north-south
zone’. paths often behave like conventional skip.
As noted, for operators who have the Signals are usually clean and stable. East-
good fortune to be in the TEP zone, the west signals from Japan and South
paths themselves do not have to be America can have heavy scatter modula-
tion, at times resembling aurora, and
strictly north-south. In the simplest case, display Doppler effects as well.
the two stations are on opposite sides
of the magnetic equator, although they out. From there, they may go either
can be at a considerable angle to the north or south, depending on which side
north-south line, as noted earlier. All that they find the ‘door’ out.
is necessary is that the two corners be A typical example of across-the-
at usable chordal skip points. equator TEP is the night time pipeline
In reality, many contacts made using that often exists between Hawaii and
TEP are a form of enhanced forward scat- Australia (for example, Hilo and
ter, or even side scatter. If the stations Townsville). Geomagnetically, this is
are substantially east or west of each nearly a north-south path. Usually sig-
other (in geomagnetic coordinates) their nals are pretty clean, and quite strong
signals will enter the region between the (50 watts and a long wire will do).
two chordal skip points at a consider- On the other hand, it is not that un-
able angle to a north-south line. When common for Hawaiians to hear Japan at
this happens, the signals may bounce the same time – and on the same beam
back and forth within the north and south headings as Australia – on what sounds
walls of the equatorial bulge, using the like backscatter. Japan and Hawaii are
bulge as a duct. on the same side of the equator and
East-west signals can be thought of mostly east-west of one another.
as zigzagging north and south in the Finally, propagation across the equa-
short term, but generally moving along tor, but largely along the equatorial
in an east-west direction under the bulge anomaly, can produce very strong sig-
until they find a weak point and break nals, such as the link between Hawaii

24 Six News
and South America. However, these sig- Sampling in Time
nals can be strongly distorted indicat- Before proceeding, some caveats are
ing significant scattering within the duct. in order. The data about to be presented
do not represent a controlled scientific
Cycle 23 Long-Path Observations experiment. For example, the stations
The peak years of Solar Cycle 23 of- at both ends of the various circuits were
fered a number of opportunities to ob- not all operational 24-hours a day, seven-
serve long-path propagation at 50 MHz. days a week. So some events, or the
The material that follows is the result of lack of events, were not observed. An-
the analysis of more than 314 contacts other factor is that the events all occurred
and their accompanying temporal, geo- during the peak years of the cycle, so
graphic, geomagnetic, ionospheric, and one cannot logically infer conditions
solar circumstances. Table I shows the during other phases of the solar cycle.
contributors to the data and the general
areas they worked into. Landmasses and Populations
The earth is mostly covered with wa-
ter. Given the long-path distances, and
Table I: Sources of Data Used in the
what appear to be constraints on the
Study, 1999-2002
propagation directions relative to a sta-
tion, there may not be any stations avail-
Reporting Working To: Contacts
able for some - otherwise technically
S79MXU South Pacific 1
possible - paths. Likewise, even on the
KH8/NØJK Africa, Middle East 2
world’s actual landmasses, there is not
8R1/W7XU Indonesia, Indian Ocean 3
a uniform distribution of amateur stations.
FY/W7XU Australia, Australian Maritimes 8
Both of these facts will affect the sam-
FO5RA Africa, India, Middle East 10
pling statistics; the conclusions must be
CEØY/W7XU India, Indonesia 11
viewed in light of these limitations.
9G/W7XU Japan, Philippines >30
KH6/K6MIO Europe, Africa, Middle East 249 Transpolar Long-Path Propagation
(TPL) – Really Stepping Out
These reported contacts were all ini- Late in the evening on 9th October
tially thought to be long-path contacts 1988, on the rapidly rising leading edge
by the reporting stations. However, each of Cycle 22, a six-metre station in Greece
report was carefully reviewed as to beam (SV1DH using the special six metre call,
headings and endpoint separations in an SZ2DH) worked a station in Japan
effort to discern whether they were, in (JG2BRI). What was especially amaz-
fact, long-path circuits. ing was that it was nearly midnight in
It became apparent that the contacts Greece and SV1DH was beaming south-
fell into one of three categories: long west, away from Japan, toward the south-
path via one of the Poles (TPL), long ern reaches of South America! The Japa-
path via the magnetic equator (TEL), and nese station was beaming southeast, at
short path via the magnetic equator the other side of the south end of South
(though over quite long distances). TPL America.
was by far the most common of the three The two stations completed a nearly
effects. 31,000 km long-path contact from north

Six News 25
of the magnetic equator southward en- coordinates, rather than geographic
croaching on the Antarctic near the coordinates. The map consists of three
South Pole, and then back north across ‘earths’. The central one represents the
the magnetic equator again and land- location of the reporting station, while
ing in Japan. The actual signal travelled the upper one represents the correspond-
about three-quarters of the way around ing station for contacts over the North
the world! [6, 7] Pole and the lower one represents the
corresponding stations for contacts over
Now Where Did That Come From?
the South Pole**.
Yes – where indeed? While perhaps
Notice that all of the stations, at both
not the first transpolar six-metre long-path
ends of each circuit, are in the TEP
contact, this example demonstrates the
profound propagation effects that can
occur. One plausible answer points back
to the power of grazing-incidence reflec-
Looking at 2000-2002
The peak years of Cycle 23 provided
a number of TPL openings to the de-
light of operators in southern Europe,
Africa, the Middle East, India, Hawaii,
equatorial South Pacific and elsewhere.
Perhaps the most widely known, if
only because of the number of contacts Figure 6: A diagram showing the seasonal
occurrence of both TEP and TPL as found
made, were the series of spring and fall
in the KH6/K6MIO log. Bars with square
openings in 2000, 2001, and 2002 be- ends indicate the range of dates over which
tween Hawaii and the Mediterranean and TEP was in evidence from time to time.
Southern Europe – over the South Pole. Bars with round ends indicate the range of
However, during this same period of dates during which TPL was sometimes
time, there were many other contacts
taking place in other parts of the world. Time of Year
These included contacts between Ghana Figure 6 shows a comparison of the
and Japan/Philippines (South Pole), seasonal occurrences of both TEP and
Easter Island and India/Indonesia (North TPL. There may well be other instances
Pole), Tahiti and the Seychelles/Ethiopia that occurred, but were not observed at
(North Pole). that station. The seasonal effects ap-
Figure 5 shows the relationship of the pear to be essentially the same for TEP
path endpoints for a sample of these and TPL, near the equinoxes, centred on
contacts. The map is in geomagnetic March and October.

** Due to the map projection, the actual great circle ground tracks are not accurately shown,
especially at the poles.
The TEP-zone boundaries are rough estimates. A few of the stations in the lower left of Figure 5 fall
around the northern boundary, but these are actually in the observed TEP zone, as can be seen from
Figure 1.

26 Six News
Time of Day and the Presence of evening when TEP is already in evidence
TEP over the usual paths (e.g. VK4), but the
A typical Hawaiian TPL opening oc- TEP is generally sporadic and not par-
curs near solar cycle maximum on an ticularly intense or widespread. Quite

Figure 5: A world map in magnetic coordinates. It shows 36 typical TPL contacts made in
2000-2002. The middle third shows the reporting stations; the upper third, the stations
worked over the North Pole; and the bottom third, the stations worked over the South Pole.
The curly lines in each third represent the approximate limits of the TEP zone. The map
projection distorts the actual signal path between the respective stations, especially near
the poles. But, it accurately shows the endpoint relationships.

Six News 27
strong backscatter coming from head- by clocks in the entire time zone.
ings of about 195° is very common, sug- On the night time side, they occur
gesting lots of ionisation and tilted lay- within an hour or two of local midnight.
ers. Television carriers on 48 MHz are Since the paths extend more than half-
heard at about the same headings, usu- way around the world, one would expect
ally with many offsets, often quite strong. that if it were midnight at one location it
Sometime within an hour or two of mid- would be near midday at the other. In
night, some very weak signals may show fact, on the daytime side, they occur
up on 50.110 MHz. After a few at- within an hour or two of 11:00 LST. Fig-
tempted calls, signals improve, callsigns ure 7 shows the distribution for the Ha-
are finally worked out, and the contacts waii-Mediterranean path.
begin into southern Europe. This pattern appears to be independ-
Usually, signals are very weak, and ent of the station’s location, and inde-
power is quite helpful. However, on a pendent of whether the path has gone
few occasions signals are enormous (e.g. over the North Pole or the South Pole.
50 watts to an indoor wire dipole in Por- However, the events seem to be an-
tugal working Hawaii), and even QRP is chored in the Night time TEP. It is im-
possible. portant to note that, for the Hawaiian
Examination of the Hawaii and other circuits, I have no reported cases of TPL
data show that there is a clear correla- when TEP was not actively present
tion between the occurrence of the con- around the time of the night time side
tacts and the station’s Local Solar Time opening.
(LST). This is particularly in evidence While it was not uncommon to hear
when one calculates the actual solar time a few Australian stations coming into
at the latitude and longitude of the sta- Hawaii before, after, and occasionally
tions involved, and not just the time held during the TPL, the Australians were not

Figure 7: This frequency of occurrence histogram estimates the likelihood of TPL as a
function of time of day. The average time on the night time side is 23:51 LST, and on the
daytime side it is 11:01 LST. Local solar time is that actual solar time at the station, not the
Time-Zone clock time.

28 Six News
hearing the Europeans at all, to their frus- In the Hawaiian openings, the night
tration. Interestingly, the same was true time openings were almost always close
for stations in New Zealand and the to the long-path great circle headings
South Pacific. to Europe. These headings are about
Except for normal TEP stations on the 15-20° west of south. Signals travelling
other side of the equator (who were not on these great circles will reach about
hearing the TPL), during the Hawaiian 70° S geographically before heading
openings, I did not observe a single in- north. This takes the path very close to
stance of signals from any intermediate the South Magnetic Pole (65°S, 139°E
point, despite the fact that the signals geographic, in 1988).
travelled the equivalent of many F-layer The observed Hawaiian morning
hops. openings had headings of about 160°,
The TPL did not appear to have come again quite close to the expected great
down to earth anywhere in between the circle route for Oman‡‡, about 20° east
two endpoints. of south. It was tempting to conclude
It would be proper to ask whether the that night time openings systematically
Hawaiian contacts were only made at went somewhat west of south, and day-
night. Well, not quite. All except three time openings went somewhat east of
contacts, were at night. During the day- south. Comparing these observations
light hours, there were two contacts with with the data from stations in other parts
Oman, and one with Italy, all made about of the world, the indications are that the
08:30 LST on various dates. These three characteristics seen in Hawaii were typi-
were all on east-of-south headings, rather cal.
than on the night time west-of-south All of the TPL contacts reported oc-
headings, but more on this in the next curred such that the daytime station was
section. always beaming somewhat east of the
On the night time side, the propaga- pole, and the night time station somewhat
tion always occurred in the latter part of west of the pole
the typical period for evening TEP, and Notwithstanding, the Hawaii log re-
some TEP was in evidence. veals that during a number of the night
time openings, strong 48 MHz TV carri-
Antenna Headings ers were heard from the 160° heading,
While the expression ‘over the poles’ but no stations appeared to be present
has been used frequently here, that is (there are very few in that part of the
really shorthand for ‘near the Poles’. The world, and it was well into the work day).
real poles of interest here are probably It is possible that this east versus west
the geomagnetic poles, rather than the phenomena was just a result of the land-
geographic ones. The geographic and mass and population effect.
magnetic poles are offset by about 11° Although the observed headings
of latitude. seem to vary little, nevertheless, a few
The Italian contact was certainly an anomaly. It was heavily modulated by multipath and very weak.
It occurred less than a minute before an Oman contact, which showed none of this. I am quite sure
that Italy in the morning was a case of sidescatter as described earlier for TEP. Except, scattering was
happening on the other side of the world!

Six News 29
stations were worked whose great circle ing in the 193°-197° range. They also
bearings deviated from the above head- agree very well regarding the longitude
ings by as much as 30°. It is not clear range about the average. Both show a
whether, in the heat of rapid-fire DX ex- standard deviation of 10° and quite simi-
changes, the antenna heading could lar values for the minima and maxima.
have improved the signal if adjusted, or The centre of the ‘other-end’ footprint
if ‘steering’ of the signal was occurring would be about 194° of longitude away,
in the ionosphere, with modest great cir- with a core footprint width of ± 10° with
cle deviations. some endpoints as far as ± 30° from the
I suspect the answer is that a combi- centre.
nation of both was happening. Certainly,
when signals were ‘good enough’, the Table II: TPL Footprints Relative to
tendency was to work what was there. the Reporting Station, 2000-2002
On the other hand, there was at least
one unusually intense opening, where I North-Polar Paths South-Polar Paths
found the ‘best’ heading started at 200° Contacts = 14 Contacts = 234
but swung as far east as 170° and then Lat Lon Lat Lon
(deg) (deg) (deg) (deg)
back again. During this excursion there
Average Change 196 194 242 193
was no clear difference in the geographi-
Standard Deviation 6 11 4 10
cal location of the stations being heard,
Minimum 188 187 233 166
that is, it was not obvious that the foot-
Maximum 214 216 250 222
print was changing significantly at the
same time.
Generally, the paths involve offsets The latitude changes between the two
from north or south, but not more than tables are a different story. Here we see
about 20°. a difference of almost 50° between the
averages, although the standard devia-
The Path Footprint
tions are very similar. One notes that,
The characteristics of the path foot-
compared to the South Pole, there were
print are shown by the latitude and lon-
very few North-Polar contacts. Figure 5
gitude changes that occur between the
suggests a possible explanation for both
location of one station and the other.
the number of contacts and the latitude
Table II shows the average change in lati-
differences: the landmass/population
tude and longitude with respect to the
question. If the North-Polar paths do
reporting station, along with some meas-
have the same range of latitude changes
ures of variability. There are 14 North-
as the South-Polar paths, the footprints
Polar contacts, based on several report-
of the observed North-Polar paths would
ing stations and paths, on the right, and
fall mostly in the Indian Ocean, with few
234 contacts (all the Hawaii-Europe path)
stations to work, and badly skewed sta-
on the left, where at least the grid
squares of both stations, and hence, the
approximate endpoint latitudes and Footprint Changes in Time
longitudes could be known. There was some question as to
Both longitude tables are in good whether there was any systematic trend
agreement with the average change be- in the centre of the footprint as a func-

30 Six News
tion of time, either during a given open-
ing, or during the season. The results
were inconclusive. Although there are
some weak indications of trends in the
footprint centre during a single night’s
operation, there was no consistent pat-
tern from night to night, nor with the pro-
gression of the season.
Solar-Terrestrial Conditions
It is clear that solar and geomagnetic
activity have an important effect on the
occurrence of TPL. A careful examina-
tion of these conditions during all of the
reported TPL episodes confirmed the Figure 8a: South-polar TPL appears to
anecdotal information that a quiet require very quiet polar magnetic condi-
geomagnetic field, in the presence of tions and fairly quiet global conditions.
elevated solar activity, appear essential.
Comparisons were made to several
geomagnetic parameters and the con-
clusion was that TPL seems to be sensi-
tive to all of them, as shown by the Fig-
ure 8 analysis of the KH6/K6MIO data,
using the number of contacts in a given
opening as a measure of propagation
TPL appears to be most sensitive to
the high-latitude (polar) K index, where
Khi=0 was the dominant value for the
South-Polar path. The planetary K in-
dex, Kp, was also low, but the propaga-
tion was tolerant of somewhat higher Figure 8b: Although favouring solar flux
values (averaging about 1). around 195, TPL appears to be fairly
insensitive to the exact flux, occurring
94% of the Hawaii contacts were throughout the 150-250 range.
made while Khi=0. In every case, Khi and
Kp were 3 or less. Though not shown here, the same
The daily 10-cm flux ranged from 141 analysis of the South-Polar contacts from
to 229, with an average of 181. How- Ghana to Japan and the Philippines look
ever, the smoothed flux was confined to almost identical to the Hawaii data.
the rather narrower range of 168 to 196. As with the statistics for the path foot-
37% of the contacts occurred with daily prints, the 14 North-Polar contacts in the
fluxes between 141 and 188. As with data are anomalous. As with the South-
other F 2 propagation forms, elevated Polar paths, the Khi indices were about
smoothed values (168-196) seem more one point less than the Kp indices, how-
significant than the daily values. ever the median Khi was 2, with a number

Six News 31
of contacts made at 4. The mean solar
flux value was 229. It must be borne in
mind that there is really very little data in
this case (14 contacts versus over 220
for the South-Polar case), which may
have skewed the statistics. In addition,
the indications are that the 14 contacts
correspond to the ‘edges’ of the nomi-
nal footprint, rather than the centre.
The TEP Connection
There is strong evidence to support
the proposition that there is an essential
connection between TEP and TPL:
1. TPL stations at both ends of the
circuits are in the TEP zone.
2. The TPL equinox seasonal effect
is the same as TEP, principally March and
3. TPL normally seems to occur only
when some evidence that TEP is present.
Figure 9: The footprints of the Hawaii TPL
An additional piece of evidence is to the Mediterranean, and the non-long-
seen in Figure 9. It shows that the TPL path South Africa contacts. Dots show the
Mediterranean footprint from Hawaii is Hawaii contacts. The ovals are the Figure 1
also an excellent match to the TEP Medi- TEP footprints for the south African-
terranean footprint from southern Africa. Mediterranean path. These strongly
suggest that both Hawaii propagation
Note that the great circle path coming
forms are related to some kind of TEP
from Hawaii goes north from South Af- insertion or launching effect.
rica and slightly west.
However, unlike TPL, the K indexes
The Hawaiian TPL signals came down
were systematically 4 or higher.
in just the right spot to suggest that the
This appears to be some form of
last hop was actually TEP!
propagation in between TEP and TPL.
Close, but no cigar It is much too far to be normal TEP. On
It is interesting to note that there were the other hand, although the path crosses
a few short openings between Hawaii and through the South Magnetic Polar re-
South Africa in 1999 and 2000. South gion, it is not TPL. It is not long path of
Africa is almost exactly halfway around any kind, because it is a few hundred
the world from Hawaii, and it is on the kilometres short of the halfway point
same great circle as the TPL path from around the world.
Hawaii to Mediterranean Europe. This all suggests that medium to high
Like TPL, they occurred during the values of K bring the polar signals back
solar-cycle-peak years, with comparable to earth in South Africa, while low values
solar flux levels, during the TEP season, of K may allow them to continue on to
when some TEP is present. the Mediterranean.

32 Six News
Figure 9 shows another curiosity,
during these short-path openings to Af-
rica, the Hawaiian signals land within the
starting footprint for southern Africa TEP
to the Mediterranean. That is to say,
when K is high, the Hawaiian signals end
at the point that the South African sig-
nals start from, on their way to Mediter-
ranean TEP. But, when K is low, these
signals seem to pass by South Africa and
land where the South African TEP lands, Figure 10: Refraction in the equatorial
in the Mediterranean basin. anomaly of rays transmitted at slightly
different angles can lead to some parts of
This latter point, together with the the signal escaping from the F2 layer, or
observation that TPL does not appear returning as TEP, or being injected at small
to come down to earth anywhere in be- angle of attack to produce further chordal
tween the two endpoints, suggests that hops.
the Hawaii-South Africa path and the Ha- In between these two extremes, some
waiian TPL signals have crossed the rays would be bent back below the F2
South Magnetic-Polar region at about layer but not enough to reach the earth.
the same point; but that in the first case, These rays would go forward until they
they were directed downward, and in the hit the underside of the F2 layer again,
second case, they were directed to (an- taking what is now a second, high-MUF
other) chordal hop. chordal hop.
How Might TPL Work? The significant point is that, the
Consider the following scenario. chordal hop from the equatorial anomaly
Suppose that it is during the spring or has injected a fraction of the signal en-
fall equinox period, and that TEP is ergy into the ionosphere at such a shal-
present. An upcoming wave from a trans- low angle that, even in the case of a
mitter within the TEP zone illuminates a smooth spherical ionosphere, the wave
range of the curved surface on the near- may now continue skipping around the
side TEP skip point. That wave is de- earth in a series of high-MUF chordal
flected, not at a single angle, but over a hops.
range of vertical angles as shown in Fig- In a smooth, spherical ionosphere,
ure 10 (based on the, technically more this signal would be trapped forever in a
correct, refraction model for skip). series of grazing incidence hops, and
A whole range of rays, at various an- never returning to earth. But, in reality,
gles, would proceed across the equator it will eventually encounter the equato-
and hit the curved surface at the far-side rial anomaly on the other side of the
skip point. Some rays would be bent earth. There, the ‘injection’ process
enough to go back down to earth as could be reversed and a fraction of the
normal TEP, accounting for the fact that arriving signal sent back down to the
some TEP is present around the TPL earth on the far side of the magnetic
openings. Some rays would be bent equator, as if it were TEP – but from very
very little and escape into space. far away.

Six News 33
This would mean that stations on the be shallow enough that, when bouncing
north side of the equator could commu- off the southern edge of the anomaly, it
nicate over the South-Polar region with never comes down to earth. Instead it
stations north of the equator on the other continues to bounce like a rock skip-
side of the earth. Stations south of the ping across a lake as the curving iono-
equator could communicate over the sphere keeps coming back to meet it
North-Polar region with other stations again.
south of the equator on the other side of If the same conditions seen south of
the earth. Hawaii also exist at the magnetic equa-
Figure 10 also explains why the sta- tor over northern Africa, the shallow skip-
tion on the right hears both TEP and TPL, ping wave will finally be bounced down
and the TEP station on the left does not out of the ionosphere by the northern
hear the TPL at all – it goes completely edge of the bulge, landing in Spain.
overhead. Since there is little D-layer absorption
Consider the practical example of a and the MUFs are very high due to the
path starting in Hawaii and ending in angle, the long path is actually possi-
Spain. The long-path link passes south- ble, while the short path, with its entirely
west from Hawaii, between Australia and traditional earth-sky-earth hops, is com-
New Zealand, the western edge of Ant- pletely out of the question.
arctica near the South Magnetic Pole, Other Possibilities - This picture is
Africa, and finally to Spain. The key fac- probably over simplified, and does not
tor here is that the first and last hops are contain all Nature’s of subtleties. There
off the equatorial bulge, as suggested may well be other kinky things thrown
by Figure 11. in.
If conditions are right at both ends For example, there are known F2-layer
(and in the middle) the chordal hop can tilts near the magnetic North and South
Poles in the vicinity of geographic lati-
tude 70°-80°, that may play a role in get-
ting the signal across the pole [8]. There
can be bumps that look a lot like the
TEP bulge, but on a smaller scale. These
could pass a chordal hop through the
polar region.
It is also true that near the equinoxes,
the signals cross the day/night termina-
tor (the grey line) near the South Pole at
nearly right angles, where there are also
tilts created by the day/night transition.
Any or all of the above effects may play
Figure 11: TEP can provide the launching a role.
points for shallow attack angle grazing
hops that cover long distances, with higher
Another related possibility is that the
than normal MUFs and low absorption. At trip over the Pole may be the result of
such times, long-path can be a superior ducting effects, such as a radio fre-
mode of propagation. quency ‘whispering gallery’ [9].

34 Six News
Transequatorial Long Path – another Europe and the Mediterranean. The
way to go far beam headings were 045° toward south-
In addition to TPL, there is at least ern Europe.
one other kind of long path. This phe- Samoa, Jordan, French Guyana, and
nomenon occurs more or less along the Queensland/Northern Territories are all in
equator – Transequatorial Long Path the TEP zone, but the great circle routes
(TEL). were not transpolar at all. In both cases,
In April 2000 a station in American the mostly eastbound signals from the
Samoa (KH8/NØJK), beaming east, con- west stayed in the TEP zone, crossed
tacted a station in Jordan at 0700 LST. the magnetic equator once into dark-
A few months later in October 2000, be- ness, landing at the other end of the cir-
tween 0600-0800 LST, a station in French cuit between about 2000-2100 LST. All
Guyana (W7XU), also beaming east, the contacts were definitely long path,
worked a number of Australian stations traversing longitudinal distances of 190-
in Queensland and the Northern Territo- 206°.
ries. The later contacts happened dur- Time of Day
ing a time that there was a daytime F2 Since the path was between a day-
opening between French Guyana and time station on the west working east to

Figure 12: This map shows three examples of TEL (VK8 and VK4 overlap). The west-most
station in each case is in daytime and the east-most station is in night time. Note that only
one crossing of the magnetic equator has happened in each case. The map projection
distorts the actual great circle routes and headings.

Six News 35
a night time station, where the night time How does TEL work?
path was apparently provided by nor- There is evidence that there may ac-
mal TEP, one would expect to see a nor- tually be two different mechanisms at work
mal TEP time-of-day pattern for the night here.
time station. Figure 13 shows that this In the first case, as in the discussion
is very much the case for the seven iden- of Figure 4, fairly long east-west TEP is
tified TEL contacts. quite common. The paths from South
For the daytime side, the requirement America to Europe and Hawaii to South
is really that the station be far enough America are typical examples. These
west that the contact actually stretch particular circuits are more or less along
more than halfway around the world (or great circles. They cross the magnetic
else it would not be classified as long equator at very shallow angles. The path
path). Thus, one would need at least a stays within the north-south boundaries
12-hour time change. of the equatorial bulge for a significant
If the night time pattern is early fraction of the trip. The end-to-end path
evening, then the daytime pattern must length is typically 10,000-12,000 km.
be early in the daylight hours. Figure This is about 60% longer than traditional
13 shows that this is indeed the case, north-south TEP, although it is far short
with most of the contact times cluster- of long path (the half-way distance is
ing about 0700 LST. about 20,000 km).

Figure 13: The frequency of occurrence of TEL as a function of Local Solar Time shows
that the night time (east-end) station has the typical early-evening TEP pattern.
The daytime (west-end) station must be up early in the morning in order to be far enough
from the night time station to produce long-path.

Solar-terrestrial conditions It typically occurs in the mid-to-late
There were only seven instances of evening hours on the east end of the cir-
true TEL in the data. Generally, the K cuit and in the afternoon on the daytime,
indices were much higher than for TPL. west end of the circuit. In this regard it
Six of these contacts occurred when the resembles TEL, except that, being a
mid-latitude index Kmid = 3 and Kp=4. shorter path; the daytime end is fewer
The remaining contact occurred when the time zones away.
values were 1 and 2 respectively. Thus it is afternoon, rather than first

36 Six News
thing in the morning, on the west end. It TEL, but not long enough. The contact
is entirely possible that some TEL is an times on the night side ran from 2000 to
extreme case of this same east-west TEP 0000 LST. On the day side, they ran from
effect. 0900 to 1400 LST. These generally oc-
On the other hand, some of the con- curred with Kmid between 0 and 3, Kp be-
tacts are consistent with a daytime multi- tween 1 and 3, and the solar flux between
hop F2 link that connects to a final, classi- 157 and 217.
cal night time TEP hop.
The multiple F2 hops are probably en- Discussion and Summary
hanced by the normally more strongly- At the outset of this study, the plan
ionised north and south equatorial- was to try to characterise the transpolar
anomaly ridges, as the signal moves form of TEP-enabled long path (TPL)
eastward. However, it may well leave based on the observations and perhaps
the TEP zone altogether for a short while to offer some plausible explanations for
before the great circle brings it back the effect. However, it became evident
south again. there was a second form (or forms) of
The characteristics are that the day- TEP-enabled long path, which I lumped
time station will be looking east into an broadly into the transequatorial classifi-
active F2 opening slightly to the north, if cation (TEL).
north of the equator, and slightly south First, some caveats
if the daytime station is south of the While there were hundreds of contacts
equator. To get the longest throw, early in the TPL database, there were just
morning is best (0600-0900 LST). seven in the TEL category, along with
If conditions are right and the F2 hops only 18 contacts in the apparently re-
go far enough, a link to TEP from the lated long-reach short-path mode. As a
daytime station’s side of the equator to result of the small number of samples in
TEP across to the other side of the equa- the TEL data, one should be cautious
tor can occur. Unlike TPL, normally about drawing very strong conclusions
there is only one magnetic equator cross- about its characteristics.
ing (nearer the eastward, night time sta- There were only 14 contacts for the
tion). North-Polar TPL case. Consequently,
Almost, but not quite there may be too few to describe the
Like the first form of TEL, not all nF2- statistics of the effect. There are also
TEP links are long path. It just depends significant landmass effects. North-Po-
on how far the multi-hop F2 can go be- lar TPL connects stations in the magnetic
fore finding the night time TEP zone. Southern Hemisphere with stations on the
Examples of long-throw short paths are other side of the earth but also in the
Hawaii to St. Helena Island and Tahiti to magnetic Southern Hemisphere. The
the Arabian Peninsula. Not long path, problem is that, to a much higher de-
but probably the same propagation gree than the Northern Hemisphere, the
mode. Southern Hemisphere is mostly water!
Altogether there were 18 cases of very So, the spatial distributions of the ex-
long short-path contacts linking to TEP pected footprints are not well sampled
one way or the other – very much like either.

Six News 37
TPL Where can you get TPL? – This is a
The TPL contacts noted in 2000 - matter of looking at where paths start
2002 occurred between stations that and where they end and then asking
were both in the TEP zone, essentially whether there will be anyone there at the
on the same side of the magnetic equa- other end of the path.
tor. It appeared as if they were working The experience is that both stations
TEP across the equatorial anomaly – must be in the TEP zone. Generally, the
except that they were linked to each paths involve offsets from true north or
other on the far side of the equator by a true south, but not more than about 20°.
very long-range propagation mechanism The centre of the ‘other-end’ footprint
that involved a more or less great-circle would be about 194° of longitude away
path near one of the poles. (east or west), with a total width of ±30°.
The observational evidence is gener- The latitude change can range from
ally consistent with a model calling for about 190° to 240° on centres with a
the TEP mechanism to inject a portion range of ±15° or so.
of the upcoming signals into shallow- If one takes a look at the Figure 5
angle of attack chordal hops on the far world map and picks a proposed start-
side of the equator. However, this is re- ing location in the TEP zone in the cen-
ally not certain. tral ‘earth’, then the possible TPL path
For example, the day-night termina- options would be those areas in a wedge
tor for these paths was being crossed at about ±30° from the same magnetic lon-
nearly right angles in the vicinity of the gitude in the upper or lower earth, be-
respective Pole. It is possible that a tween the TEP-zone lines. One should
Grey-line hop may have played a role be looking at paths that cross the mag-
there. Such a picture calls for enough netic equator near the starting station and
ionisation at the day side of the anomaly again near the ending station.
to allow some form of morning TEP. North-Polar TPL Route – Recalling
Another issue is with conditions on that this is a Southern Hemisphere to
the daytime side of the circuit. Morning Southern Hemisphere path, a look at
TEP is not very common, but one should Figure 5 will show that the only signifi-
consider Figure 10, with signals travel- cant landmasses in the southern TEP
ling in the opposite direction. Here, the zone are Australia and the southern
incoming day-side wave is arriving from halves of South America and Africa.
a chordal hop, not the earth’s surface. Australia and Africa mostly match up with
Thus, it needs a much shallower refrac- Atlantic and Pacific Oceans, respectively.
tion angle to get back to earth, than a Nevertheless, North-Polar TPL ought
normal TEP signal. Thus, the day-side to be possible between the vicinity of
electron density would not have to be the northern strip of Australia and south
at the higher night-side levels in order to eastern South America (southern Brazil,
redirect the signal back to earth. Paraguay, Uruguay, and north eastern Ar-
In any case, this requires the coinci- gentina). If this were to occur, one would
dence of favourable conditions over a expect the South American end would
very long path, and it is no surprise that be in local early morning and the Aus-
it is relatively rare. tralian end near local midnight.

38 Six News
Referring back to Figure 5 will also Nevertheless, it can provide very long
reveal a number of South Pacific islands paths in a generally east-west direction.
that have plausible paths to southern A consequence of the day-to-night
parts of Africa and South America. connection is that the daytime station is
South-Polar TPL Route – This ap- aiming generally east and the night time
proach connects the northern TEP zone station is aiming generally west. In all
to its northern counterpart on the other probability, the F2 hops are aided by the
side of the earth. Figure 5 suggests that enhanced ionisation along the equato-
circuits from Southeast Asia to Central rial-anomaly ridges. If good F2 condi-
America and northern South America tions exist beyond the TEP zone, the
should work, along with the well-known path may leave the zone and re-enter it
Hawaii to Mediterranean path. again as it approaches the other station.
TEL Long-throw transequatorial short
Whether extreme east-west TEP or path
multi-hop F2 to TEP links, TEL seems to The east-west TEP and nF 2 -TEP
be rarer, since it appears to depend on modes are fairly common within the TEP
the long reach of F 2 mechanisms for zone. They appear to result from the
much of the propagation distance. Con- same phenomena that cause TEL, ex-
sequently, it requires not only a geometri- cept that the links are shorter. Like TEL,
cally usable connection to TEP on the the daytime station is aiming generally
eastward end of the circuit, but also east and the night time station is beam-
good daytime F 2 conditions as well. ing generally west.

Six News 39