This is the third in a series of articles exploring the possible causes of two somewhat different, although very long short-path propagation phenomena that appear to involve Es somewhere along the path (see Kennedy CQ VHF Fall 2011 and Kennedy and Zimmerman CQ VHF Winter 2012).
EWEE: An East-West Effect
The previous article in this series, Range 50-MHz Es: East-West (EWEE),”printed in the Winter 2012 issue of CQ VHF magazine, discussed the first of these two effects. It is an east-west propagation that occurs in the local summers in the Northern and Southern Hemispheres. Unlike the common variety of Es, it produces paths of 8,500-11,000 km. This is equivalent to five or six Es hops (see Table 1). Hints of this sort of thing have been reported at times since the late 1970s. However, the number of these occurrences seems to have increased significantly since about the year 2000. Han Higasa, JEIBMJ, argued that five or six hops was beyond the workable range of Es, and labeled the effect Short-path Summer Solstice Propagation, or SSSP (Higasa 2006 Japanese, 2008 English).
Nevertheless, the propagation seems to occur exclusively during the hemispherelocal summer, strongly suggesting that some form of Es mechanism(s) is at play. Kennedy (2010 and 2011) subsequently noted that, like ”Es, these paths are accompanied by a very reliable time-of-day pattern. The west-end station is within its Local Solar Time (LST) morning Es peak (roughly 0630-1230), and the east-end station is in its LST afternoon-evening Es peak (roughly 1600-2200). Evidence was presented that this was truly an Es effect, and that, at different times, it might be either entirely normal multihop Es (nEs), or a combination of nEs with some variation of chordal-hop Es (Kennedy and Zimmerman 2012). They called it East-West Extreme Es (EWEE).
This Study: North-South Across-The-Equator Effect
Another form of propagation which showed up in 2009, 2010, and 2011 is even more surprising, although anecdotal reports have been around for a long time. This is a very weak signal mode that always crosses the equator and has path ranges from 10,500 km to beyond 15,000 km. One end of the path is in the Southern Hemisphere during its local summer, and the other end lands in the Northern Hemisphere during its local winter. Unlike EWEE, the paths cover very long north-south distances, in addition to large east-west distances, between the two end-point stations. If this is the result of ”nEs multihop, then Table 1 shows that this would require an unprecedented six or seven hops.
Curiously, the only documented paths are between ZL/VK (and environs) and NA. It is tempting to speculate that this second phenomenon might somehow be related to EWEE. For example, they do involve a station on one end during its summer major Es season, and a second station during its winter minor Es season. However, it will be seen that there may also be something else involved. What follows is a discussion based on recent observations and some comparisons with a data-driven ionospheric model. However, first here is a quick look at the general characteristics of both the E and F ionospheric layers.
The E layer lies roughly between 90 and 130 km above the Earthsurface, although its boundaries are rather fuzzy and its effects are occasionally observed as high as 150 km. There are two main sources of E-layer ionization, the Sunsoft X-rays and ultraviolet radiation during the daytime, and vaporizing meteors at all times. Since the Sun is a large contributor, there is a big puddle of E-layer ionization that follows the Sun around the Earth, with the ionization maximum being very near the moment-to-moment subsolar point on the Earth. The mid-latitude summertime peak in Es appears to be related to an observed large seasonal peak in the influx of random meteors (Haldoupis, et al. 2007).
Even so, the normal background amount of E-layer ionization is much lower than that required to skip 50-MHz signals. Even in the middle of a good day, the Maximum Usable Frequency (MUF) is rarely above 22 MHz. Six-meter propagation requires the assistance of additional processes to produce ”E ionization that reaches much higher levels. That is, Es results from a combination of events, all of which must be in place at the same time.
There are wind patterns in the E region, and like most other winds, they generally flow more or less parallel to the Earthsurface. Some of these winds flow along roughly east-west or west-east lines, and others along north-south or south-north lines. When the wind speed changes with height, blowing faster at one level than another, or in opposite directions, there occurs a wind shear at the boundary between these two wind streams.
In the presence of the Earthmagnetic field, the wind shear can cause the free electrons above and below the center of the shear to be pushed together near the shear. This has the effect of sweeping lots of electrons out of the large volumes of low-electron-density space above and below the shear, and compressing them into a small, very thin layer of high-electron-density space near the shear level.
This process can result in the formation of thin Es clouds with electron densities ten to a hundred times more than that of the same space before the wind-shear compression. As a result, these clouds end up with a much higher MUF than the ”E-layer ionization.
The well-known diurnal pattern of Es (figure 1 ) seems to be the result of a family of atmospheric tides that are also the result of the Sun. In this case, the Sun shining on the Earth below heats up the lower atmosphere and causes the entire atmosphere at that point to expand, well up into the E and F layers. The principal modes of these tides have periods of 24, 12, and 8 hours, respectively (Haldoupis, et al. 2004). These overlapping tides produce vertical drafts to flow first up and then down in the E region, at times inhibiting the compression process and at other times not inhibiting the process. The net effect is that when the compression process is trying to work, there is a regular daily variation in the likelihood of Es clouds forming and thus of propagation happening throughout the day. The interested reader is invited to look at the more detailed description of Es in the article by Kennedy in the Fall 2011 issue of CQ VHF.
The F2 region lies from about 250 to over 1,000 km. It is normally ionized by extreme ultraviolet radiation (EUV) from the Sun. The background particle density at these levels is very small, indeed. As a result, the ion lifetimes are fairly long. The F2 layer can remain ionized long after sundown, and there is evidence that ions can actually move along the Earthmagnetic field lines from the dayside to the nightside.
Curiously, although the F2 layer is more directly irradiated by the Sun during the local summer, the electrochemistry of the region is such that free-electron losses (through recombination) are much greater in the local summer months. As a result, the net ionization level (new ions created, minus existing ions lost, per unit time) is actually much higher in the winter months.
The winter anomaly is not the only seasonal effect. Within about ±20° of the Earthmagnetic equator there is a large outward bulge in the F layer along the EarthMagnetic Equator, known as the equatorial anomaly. This bulge follows the Sun around the Earth, but lags a few hours behind it. This bulge is actually a persistent afternoon thickening of the F layer near the equator.
The bulge results from equatorial E-layer ionospheric winds, leading to an afternoon build-up of west-to-east electric fields, called the equatorial electrojet. In combination with the Earthmagnetic field, these fields pump electrons upward from the E and lower F layers into the F2 layer above. This afternoon fountain effect can significantly enhance the F2-layer electron density.
In principle, this leads to two ionization peaks in the F2-layer bulge (not one). One of these peaks is 10° to 20° north of the magnetic equator and the other about the same distance south of the magnetic equator. As a part of the equatorial bulge, these two peaks typically also follow the Sun. Thus, any resulting propagation is generally an afternoon or early evening phenomenon.
Around the two equinoxes, both peak areas are more or less uniformly illuminated by solar radiation and they reach their maximum values. During this time, the two strong peaks can produce Trans-Equatorial Propagation (TEP). However, near the solstices, the winter anomaly tends to support the winter-side peak and strongly diminishes the summer-side peak. Instead of two fairly strong peaks, often there is only one strong peak, in the winter hemisphere (and on the afternoon side). The summer-side peak is often very weak and usually ineffective.
Ne, M Factor, and the MUF
It has already been proposed that chordal mechanisms in the E layer may play a role in EWEE, and it is certainly possible that EWEE plays a role in these long north-south paths. It is possible that F2 processes may also play a north-south role as well. Therefore, lettake a look at the interplay between Ne (the free electron density), something called the M factor, and the MUF (Maximum Usable Frequency) regarding possible path combinations.
If one knows the value of Ne, one can calculate the so-called critical frequency, f0 —the MUF for a signal beamed straight up. However, usually it is the other way around.
The MUF also depends on the angle between the upcoming signal ray path and the ionosphere; the smaller the angle, the higher the MUF.
M (the M Factor) is cosec (a), where a is the angle between the signal path and the plane of the bottom of the layer. What is significant is that by making the angle smaller, one can raise the MUF without changing the free-electron density. As a result, chordal effects caused by tilted or curved reflecting layers can lead to higher MUFs and lower path losses. This is one, but not the only, factor in F-layer TEP and possibly some instances of 2-meter Es (Kennedy 2000 and 2003).
Transequatorial Who Knows What…
Returning to the main subject of this discussion, there were several flurries of the more mysterious southern-summer to northern-winter contacts across the equator between ZL/VK and NA in December 2010, and January and February 2011. Together with the contacts made in the 2009-2010 season, there are now enough of these weak-signal data points to begin to stitch together some ideas.
There were 40 ”or ”database reports, 24 for ZL-NA, and 16 for VK-NA. There were also three KH6 reports and one E51 report. Figure 2 shows the path ranges and the call districts involved. The left-hand side of the figure shows that the path components (W5-KH6, and KH6-ZL/VK) were separately visible at times.
WhatGoing On in the Ionosphere?
A number of concepts have been advanced to explain this kind of propagation. Some have suggested TEP, but this seems very unlikely for two related reasons:
•TEP is vitally dependent on strong, and roughly equal, solar ionization of both the north and south branches of the equatorial anomaly; this only happens near the equinoxes, not near the solstices;
•TEP is an F-layer effect; with the winter anomaly, it will not work in the Southern Hemisphere summer.
It has been suggested that, perhaps, it is a rare linking of the major Es summer peak in the Southern Hemisphere with the minor Es peak in the Northern Hemisphere winter to provide a kind of extraordinary nEs or chordal Es, EWEE”link. Another suggestion is that it could be a linking of southern summer Es to equatorial Es to northern winter Es. These last two suggestions could work on paper. However, both would require six or seven hops. This seems unlikely with one of the hemispheres of season.”However, whatever actually causes the propagation, it is real and it is, in fact, unlikely.
The next step was to take a look at what the ionosphere was doing during the time these contacts occurred. This is a bit of a challenge, because the main route of travel is across the vast open spaces of the north and south Pacific Ocean, where there are very few ionosondes. However, there are several rather sophisticated computer models that can help fill the gaps. Broadly, these programs work by using known physics to calculate how the ionosphere responds to its environment. Then, streams of real-world data are plugged into the model and it produces a simulation of how the ionosphere would behave in response to those values, all as a function of time.
Depending on the specific model, typical inputs might be time series of things such as the time of day, day of year, solar
flux, solar wind speed and magnetic polarity, the Earthvarious magnetic indices, solar X-ray and EUV fluxes, the total electron content (TEC) at many points, etc. While not as good as direct measurements of the actual desired parameters (in this case, Ne), it is the next best thing available.
The Global Assimilation of Ionospheric Measurements model (USU-GAIM) was used for this study. It was developed at Utah State University (Schunk 2004). The ionospheric model was run for the entire day of each of the contact dates in the database, producing a snapshot of the global ionosphere in three dimensions from 92 km to over 1,000 km for every 15-minute period throughout the day. These models were run at the Community Coordinated Modeling Center hosted at NASAGoddard Space Flight Center.
Figure 3 shows a contour-map example of the F2 (upper) and E (lower) layers on the date of the 2010 Vernal Equinox. The Sun is directly above the equator and illuminating both the Northern and Southern Hemispheres equally. This is especially evident in the symmetrical, nearly circular shape of the E-layer ionization. The northern and southern branches of the F2 equatorial anomaly are clearly visible. However, despite the equal solar illumination over the Pacific, the southern F2 branch is smaller and weaker than the northern branch. This asymmetry in the E-layer solar response, as opposed to the E layervery symmetrical response, will play an important role in the discussion that follows.
Reality Meets Theory
The challenge now is to take what happens and see how that might fit into what we think we know about the ionosphere and the underlying physics. Letstart by looking at just two examples of the dozens of documented cases of this extremely long across-the-equator 50-MHz propagation.
February 20, 2010
On February 20, 2010, K6QXY worked ZL3NW at 0108 UTC. Tens of minutes later, he also worked ZL2DX at 0154. So, how did all that happen?
Figure 4 shows the modeled ionosphere for that day at 0115 UTC (the nearest model time point). It shows that the northern E-layer branch clearly dominates the southern one. The northern winter anomaly is strong, and the southern branch is diminished by the summertime electron losses. The F2 northern branch peak Ne is 2.73 x 106/cm3, for an MUF of just over 50 MHz. The model shows that just a little to the east, at the great-circle skip point, the Ne is 2.6 x 106/cm3, corresponding to a ground-level MUF of 49 MHz.
Whether skipping from the first peak point or the slightly weaker second peak, this suggests that marginal skip conditions would exist, even for a non-chordal hop, and marginal conditions are what were observed. On the other hand, TEP wouldnwork because the ionization of the southern branch of the anomaly is too weak at the required skip point.
This 2010 episode is typical of what has been seen for at least the last three years. Generally, what were observed were contacts between ZL, and occasionally VK, and the western U.S. to areas with a good view of the horizon looking to the southwest.
However, the granddaddy of them all (so far) occurred on January 11, 2011.
January 11, 2011
At 0037 UTC, VK3DUT reported hearing the K6FV beacon in the Bay Area. For the next hour or so, 19 worked or heard events were recorded, followed by a couple of stragglers in the two hours following that. Remarkably, 14 of the total of 21 (so far identified) were between VK and NA. The remaining seven were between ZL and NA. In the south, VK2, 3, 4, and 9, and ZL1 and 2 were worked. In the north, it was W5, 6, 7, 9, and 0, and XE2. Table 2 shows the south and north participants found in the database.
Figure 5 shows that the F-layer and E-layer configurations were essentially the same as during the February 20, 2010 event. This dual E and F pattern is commonly seen in all the events of this sort studied.
In an earlier section, a few processes were suggested as possible candidates to explain these recurring phenomena. All seemed problematic in one way or another. Yet, it is clear that whatever is happening is a statistically improbable event.
There is another possibility, but to see that more clearly, first letconsider the previous review of Es. It was said that normally E-layer ionization is pretty low, and more or less follows the Sun around the Earth, with the help of meteors. This modest level of ionization, spread over a large vertical distance, is a reservoir of free electrons.
Of course, there will be no VHF propagation unless conditions are right for the wind-shear process to come into play and produce sufficient vertical ion compression to develop the necessary thin, high-density, sporadic-E ionization sheets.
However, the computer model is only calculating the ionization level of the E-layer reservoir. It does not predict when or where, within the reservoir, Es will occur. Thus, the f-layer diagrams in figures 4 and 5 do not show the Es locations, but they do show a quite adequate reservoir (an MUF of about 18 MHz) for wind shears to compress and build into Es.
The next thing to note is that in both figures 4 and 5 the E-layer reservoirs are shaped roughly like inverted Ts. The east-west part of the T shows a good reservoir extending from east of ZL westward to the eastern half of VK. This is the normal summer-day starting point for Es between ZL and eastern VK. As the Sun moves farther west later in the day, there will be Es extending to western VK.
The north-south branch of the inverted T is really quite extensive, reaching well north of the equator. Since there is a reservoir, itquite reasonable to think that Es will happen there at some point, and due to the geometry it might favor a generally north-south direction.
At the north end of the reservoir things get a little tricky. Getting quite close to the magnetic equator (±10°), Es takes on a somewhat different character than the temperate-zone variety. The influence of the equatorial electrojet is thought to be the dominant factor (Whitehead 1997a).
This Es is normally a daytime effect, without an evening peak. However, Es may not be needed at all for the next hop. There will be more on this shortly. E-Layer Valley The previous article in this series (Kennedy 2011) pointed out that the presence of elevated ionization in the F layer above the E-layer electron reservoir has the capacity to squeeze a second region of ionization down into the E-layer area from the top, where the wind-shear compression mechanism normally operates to produce Es. This provides a site for at least a second layer of Es clouds to develop, which raises the probability of ordinary Es skip, and also chordal hops or bottom-top skip ducting. Figure 6 shows this same process was at work in the Southern Hemisphere E-layer, establishing a channel along the north-south line of the path.
Plausibly, the ZL/VK and NA stations were in the right place at the right time to catch a southern summertime EWEE path (nEs or chordal) up the Southern Hemisphere channel in the inverted TE-layer structure that carries the signal to a point somewhat south of the magnetic equator by Es mechanisms. From there, it skips across the equator directly to the northern F-layer anomaly, which then completes a north-side F2 hop that feeds a Northern Hemisphere wintertime 1Ee or 2Es link (or chordal Es equivalent) on into NA.
The Great Circle Path
Figure 7 shows some typical great circle paths between the U.S. and ZL/VK. From the north, they all go southwest, passing a bit east of KH6 near the northern F2 peak. The path goes progressively southward, into the extended T-shaped E-layer reservoir region and stays within it until arriving at ZL or VK. Those paths all overlay the northern F2 branch, and also the T-shaped E-layer ionization patterns seen in figures 4 and 5. (They are also essentially parallel to the Earthmagnetic field lines, which may or may not be a factor.)
Wintertime Es in the North
The wintertime Es peak is small and very unpredictable. Nevertheless, winter Es happens. Its ionization reservoirs are either very weak or very small. Letassume there is winter Es south and west of W6, and then look at how the rest of the path lines up. The location is W6, and the time is about 1700 LST.
Wintertime Equatorial Anomaly in the North
Figures 4 and 5 both showed that the northern-branch of the F2 equatorial anomaly forms a long, thin strand at about 20°N, extending east and west of Hawaii. The time here is about 1300 LST, near the peak of the daytime F2 anomaly. The local MUF is greater than 50 MHz.
Summertime North-South Branch of the Southern Hemisphere E-layer Reservoir
Figures 4 and 5 also showed that the north end of the north-south branch of the Southern Hemispheric E-layer reservoir is reasonably well aligned with the great circle path, going roughly southwest from the vicinity of KH6 towards ZL. In ZL, the time is 1230 LST —at the end of the local morning summer Es peak —and a contact ensues.
This would constitute a Transequatorial E to F to E path, where the E segments were each just a few hop equivalents —call it TEFE, for short. Figure 9 shows that even the longest observed paths can be explained by one or two 2.000-km winter Es hops in the north, a 4.000-km F2 hop across the magnetic equator, and two or three summer Es hops in the south.
Why ArenThere Other TEFE Paths?
Whatever the mechanism, one has to ask why isnthis same kind of path being seen when the seasons are reversed, and why arenpaths like this being seen in other parts of the world? Specifically, if the ZL/VK-NA path happens in the paired southern-summer and northern-winter, shouldnit happen during the paired northern-summer and southern-winter? Likewise, why isnthere, say, a NA-ZS path in one or both of the seasonal pairs?
Was Anyone Talking and Listening?
One reason could be that no one has spent enough time trying to work those paths, from both ends at the same time, during the right seasons and times of day. Figure 8 shows the LST early-on-late diurnal plot for the database ZL/VK-NA events. Broadly, what it shows is that the east-end stations are rather consistently in their afternoon/ evening Es peak time period.
In this specific case, the east-end stations are in their winter Es season. It may be that the wintertime stations need the most ideal Es conditions in order to find the rarely-available necessary winter Es to make the link to the F2.
The west-end station times are less systematic. Much of the propagation does occur in their morning peak period. However, many do show up in the normally lower-probability dip between the morning and the rise of the afternoon-evening peak. On the other hand, they have good summer conditions on their side.
If one were to systematically search for an NA-ZS path, or the combination of reversed-seasons variations, figure 8 suggests when the east-end stations should be on and, thus, the times for which schedules should be made.
What Does The Ionosphere Say?
If the propagation is really TEFE, with winter Es linking to the winter-side F2 equatorial anomaly, and then to the summer side of the equator, and summer Es to the far end, then the ionospheric conditions should also show why these other variations are not seen.
ZL/VK-NA in the Northern Summer
The first of these ”three path possibilities is by just reversing seasons for the known ZL/VK-NA path. Figure 10 shows the F-and E-layer configurations for a good day in the northern summer on June 20,2010 at 2200 UTC. There were excellent JA-NA conditions with many contacts being made around the same time. One can see that the E-layer reservoir is very nicely positioned to facilitate Es (EWEE) between JA and NA. It is also clear that the northern summer ?-layer ionization has very nearly the same shape as that seen in the southern summer, except, of course, in the north the T is upright, not inverted.
On the other hand, as figure 3 shows, even during the Vernal Equinox, the north-side branch of the F2 anomaly was noticeably stronger than the south, during this same time. Though its (summer) electron density was less than half its winter value, figure 10 shows that the winter southern branch is very much weaker still (MUF « 29 MHz). The main thing to note is that during these observed seasons, the:
E layer was symmetric across the equator with the reversal of the seasons, and the
F layer was not symmetric across the equator in any season during the date range of the study.
If the propagation seen in the northern-winter/southern-summer is truly TEFE, then the reason the same path across the Pacific isnseen in the reversed season, northern-summer/southern-winter, is that there is no southern F2 hop across the equator to the northern summer E reservoir (and the otherwise necessary Es, even if itthere).
NA-ZS in the Northern Summer
While NA-ZS may look like an eastern-hemisphere analog to the ZL/ VK-NA path, itnot that good a fit, even to begin with. Figure 11 shows that the necessary southern F2 branch MUF, needed to link to the northern Es, is well he low 50 MHz (about 33 MHz).
In addition, the TEFE hypothesis depends on Es processes for most of the path. The probability of Es is known to be dependent on the value of the horizontal component of the Earthmagnetic field, which is at its worldwide lowest over ZS. As a result, Es itself is fairly rare over ZS even in the local summer (Whitehead 1997b), and thus it is far less likely in the winter.
Therefore, for two unrelated reasons this path is very unlikely. Since the absence of the path could just be a matter of no winter Es in the south, nothing conclusive to be said about the TEFE hypothesis one way or the other.
NA-ZS in the Southern Summer
Here the situation is somewhat different. It is summer in ZS, so the probability of Es is much higher than in the winter, but it is still only about one third that of either ZL/VK or NA.
Coming from the north, the TEFE hypothesis holds that winter Es feeds the signal to the northern-winter F2 branch, which then provides a link across the equator to the southern-summer Es (if there is any).
Figure 12 shows that the northern F2 branch MUF is near 48 MHz, but itin the wrong place. This peak is 3,000 km east of the great-circle F2 skip point, where the actual ground-level MUF is no more than 33 MHz and probably less. In addition, the F2 skip point is about 1,200 km farther from NA than the F2 skip point for the ZL/VK-NA path (requiring at least one additional winter Es hop).
It is also clear that the southern tip of Africa is actually north of the most intense part of the summer E reservoir (whereas in the ZL/VK case, they sit right under the peak). Finally, it appears that ZS and its environs may be just a bit too far east to line up well with NA and the northern branch, in any case.
Thus, like the ZS winter case, there are too many reasons this path would be very unlikely. Again, it provides no conclusive evidence one way or the other about TEFE. Whatever happens in the F layer, the E layer is a problem.
Discussion and Conclusions
The objective of this study was to understand the behavior of the very long summer-to-winter north-south paths across the equator, taking advantage of the additional data that became available as a result of the 2010-2011 southern summer season.
So far, this effect is only observed in the southern-summer/northern-winter and over the Pacific path. It has not been observed over the Atlantic and has not been observed with the seasons reversed.
The propagation mechanism that currently best fits all the examined data is one in which the path arises from some combination three-hop equivalent sporadic-E (be it nEs, chordal, etc.) in the Southern Hemisphere, going northeast toward the equator, following the persistent bulge of the summer E-layer ionization reservoir toward the Northern Hemisphere.
At some point near the equator, the path passes through the E layer and goes off toward the F layer. Farther north and east, it finds the northern branch of the equatorial anomaly with an MUF near 50 MHz, which the model reports frequently existed even for normal ground-level angles of incidence. (If the incoming path angle arrives from a chordal Es hop, the F2 MUF would be even higher.) The subsequent F2 hop, now moving more east-northeast, finds one-or two-hop equivalent wintertime Es propagation that carries it to NA.
Note that the F2 classic ground-level MUF need not be all the way up to 50 MHz in some of these scenarios. Coming out of the E layer, the skip angle at the northern F2 skip point could be smaller than the ground-level value, and if so the ground-level MUF could be a few MHz less than 50 MHz.
As might be expected on such a long and torturous path, the signal strengths observed are generally very weak, requiring great skill and good equipment on the part of the radio operators. On rare occasions, stronger signals have been noted for brief periods of time.
Seasonal and Longitudinal Symmetry
To date, the only documented paths occur in the southern-summer/northern-winter seasonal combination, and only between ZL/VK (and nearby islands) and NA. In a symmetric ionosphere, one would expect that if the seasons were reversed between Northern and Southern Hemispheres, the same propagation would occur semiannually. However, this is not observed. Similarly, one would expect that the propagation would move around the world on a more or less daily basis and be seen to occur at many different longitudes. However, so far this is not observed.
The seasonal and geographic variations of the potential propagation paths, which are not observed, form a fairly strong argument supporting, but itself not fully conclusive of, the TEFE scenario.
Es-Only Mechanisms are Excluded; An F-layer Role is Required
In the case of the E layer, the model and observations show that the north-south seasonal symmetry is quite good. Thus, if the mechanism were purely an Es process over the whole the path, then that symmetry should work well with the reversal of the seasons over the same path. The absence of paths between ZL/VK and NA in the southern-winter/northern-summer appears to exclude known Es-only mechanisms.
If there must be something in addition to Es, then the models show that all that seems to be left is the winter-side branch of the north-side F2 anomaly. Itin the right place and has the MUF capacity to fill the role.
So far the TEFE mechanism is the only one of the several mechanisms offered that coincides with all the known vagaries of the observed temporal and seasonal propagation patterns. Further observations may refine or change this conclusion.