Very Long Distance Propagation in the 144 MHz Band
Discussion of the May 20, 2003 Dx Opening
Feedback, comments, contributions
Volker Grassmann, DF5AI
Udo Langenohl, DK5YA
Last update: September 10, 2003
This webpage discusses the readers' feedback on the article Very Long Distance Propagation in the 144 MHz Band - Discussion of the May 20, 2003 dx opening . In order to stimulate discussion and further investigations, we wish to present the comments and feedback we have received. Please note that initial feedback was received from Väinö, OH2LX and Gene, W3ZZ, you may find their contributions already considered in the above paper.
Email from Graham, F5VHX/G8MBI, August 8, 2003
I still believe that this event is a mixed mode Es+tropo, I cannot give you transition explanations, but I have pondered this some more since corresponding with Volker.....I believe the biggest support for this interpretation is that this path has been worked before _many_ times and was even done again this year after this event at least once.
Based purely on station density and 'probability' it seems very unlikely that we have seen so many 'double hop' events on this path and very few in other directions where station density is so much higher AND geographic possibilities are so much larger also. I am not a statistician, but I cannot reconcile in my mind the pure quantity and frequency of these type of EA8 contacts to a small geographic area with relatively low station density against the possibilities in other directions with higher station densities and larger geographic targets.
That does not mean I do not believe that true double hop (chordal hop, or other mid path reflectors that you speculate on) does not happen/exist, they do, and certainly at least one of this years 'traditional' July russian openings demonstrated that conclusively. It's just that these true double hops appear much rarer (as one would expect) than the long distance ES qso's made where there is a substantial water content path at one end (or both) there are just a disproportionate number of these events when compared to possible double hops over large land masses (given the INCREASED likelyhood of two stations finding each other in the large land mass scenario).
The other explanation for this is that one needs a more regular reflecting surface for the centre point of the double hop scenario to make the math work (reflection losses) ie WATER, which brings me back to my last words to Volker.....I would like to see possible path losses calculated for the various modes you speculate on, as I agreed with Volker in his response at that time, this would _not_ be conclusive because you would end up with a range of overlapping values, it would (might) however add just a little more weight to one method or another and provide a little more 'meat' to the discussions/arguments that you make.
OR... an analysis of one of the July Russian events (one of the two with G at one end and moscow and south at the other, with DL/OK etc in the middle) might provide a comparison model/theory to sit alongside this EA8 one.....personally, without doing all the detailed work, but purely from observation, I see the characteristics of the events as quite different.
Anyway, nice thought provoking work , and a lot of it !!!....thanks guys.
73 Graham F5VHX/G8MBI
Email from Miguel, EA4EOZ, August 12, 2003
I've just read your nice paper, and I would to make some comments.
First of all, I think you have overvalued the number of VHF operators in Spain. The only reason there was no activity that day from Extremadura is because there are almost no VHF operators there. VHF operators are concentrated in large cities. You can find a small group in Madrid, another one in Barcelona, and in other large cities, but it is really hard to find activity outside this areas, even in contests! It would be a nice idea to plot a map with the stations QRV on 2m. You would be surprised! Also, activity outside weekends is almost inexistent in all the country. This activity seems to grow in the Es season, but 20 of May is too early, so there was many few operators on air that day.
Another thing to consider is the tropo between EA8 and EA/CT land. This is a very common mode of propagation here. Stations situated up to 80 or 100 Km inland, can work easily EA8 with very modest FM stations. There have been MANY reports of walkie talkie to walkie talkie QSOs between EA8 and EA7, CT or even EA1 ( La Coruña ) some of them even in 70cm. Stations in that area consider a common thing to heard them each others. The hard thing is to find a SSB active station there, except some good exceptions.
Acording to May 20, there were some QSOs between EA7 (South Spain) and CT with EA8, both on 144 and 432 MHz, You have the logs on your webpage. This proves that day there was a significant tropo over the atlantic sea. This tropo is useable up to 80-100 Km inland in best situations, over that point, no DX is possible. Tropo ducting over EA land is very rare. I only have seen it two or three times in 10 years of VHF work.
For many of us (and me) there was a VHF ducting from EA8 to EA/CT coasts, and from that point, the duct disapeared and the signal followed its way up to the Es cloud. If you compare the logs between a south CT station or a EA7 station with the EB8BTV, you should find many similirities.
There are other minor points, like the area you call Extremadura. Extremadura in fact is smaller than the area you draw on the page 31 and is located north west the painted area, in fact, those squares (IM77, IM78, IM79, IM88 and IM89) are "Castilla La Mancha" and north Andalucia a "desert like" area without big rivers or lakes... and of course, VERY LOW or almost inexistent VHF activity, in fact, IM77 and IM78 are "rare" squares here.
Anyway, good work! It's a nice paper. Congratulations!
Miguel Angel, EA4EOZ
Email from Peter, PA3BIY, August 19, 2003
First of all: Congratulations on a very fine report that you and DF5AI wrote! Splendid work! I read it with great interest, but I have some comments. Please refer to them as "comments", not critic!
1. You made an elaborate study of the different types of propagation combinations that may have been involved in the May 20 opening. The Es + tropo is (partially) disregarded as the link between Es and Tropo was unlikely. I think that is true, but only if the Es link ends on top of the tropo duct (where it will bounce off again). You said that the circumstances were not favourable to couple between Es and tropo duct, but there was a possibility present. The tropo duct between EA8 and CT/EA7 ends at the coast line (or perhaps a little over the coast line), as during day time the warm convecting air will destroy the low altitude sea duct. hence, an opening will appear, that will direct the radio wave straight out, or even upwards, due to the "negative duct" that will appear over land. I have more reason to believe that ionospheric and tropo propagation may link: The recent long distance MS QSO's between G4LOH and RU3ACE and RW3PF (2500 resp. 2600km!). Both QSO's were only possible due to the presence of a sea duct across the north sea between IO94 (G4LOH) and the OZ/DL coast (were the duct probably changed shape, releasing the tropo signals). Since the signal strength nature of MS is at least one order of magnitude less than Es, the coupling mechanism must have worked rather well. There is another evidence of this type of QSO, as a northern DL station managed to work CN8LI last august 13, over a QRB of 2600km (Sri, I don't know his call....). At that particular time a very strong tropo (duct?) was present over western Europe towards Spain. In other words, tropo and Es might certainly link up, though that does not necessarily may have been the link on May 20th as well.
2. I find your evidence for water scatter opposed to ground scatter a bit weak. Indeed, EME operators mention good results from ground gain (ie: solid ground), and HF back scatter is often the results from mountain back scatter. Well, if mountains can scatter back, why should not flat ground scatter forward? The idea of a tropo duct enabling scatter up is an attractive one, but I doubt if the difference in loss is more than say 10 dB between ground and the "top duct" scattering.
Well, just a few of my comments. As I said before, a splendid article, and I hope it will help resolving the mysteries around (double hop) Es a little, hi!
Email from Günter, DL6WU, August 20 and 27, 2003
Habe mit großem Interesse die Ausführungen über den vermutlichen Mechanismus der Verbindungen gelesen. Vielleicht ist dazu ein Logauszug einer ähnlichen Öffnung von statistischem Nutzen. Vorweg: ich habe beidesmal ea4-stationen gehört und ich gehe davon aus, dass der fm-Taxifunk am 20. 05. auch aus der Ecke (Madrid) stammte. Ich glaube fest an double-hop. Interessant auch, dass alle Signale etwa gleich laut waren. Ich habe damals erst beim Eintragen der qths ins Log gemerkt, dass da ein "dicker Fisch" dabei war.
05.08.1987: 1605ut ea4bpa in80dj 59/58, 1612 ea7uh im66rx 59/59, 1615 ea7dzi im66um 59/59, 1617 ea4afi in80cl 57/59, 1620 ea4qr in80hl 59/59, 1625 ea7wm in67mi 59/59, 1635 ea8xs il28ga 59/59
Nachtrag: mein Locator ist jn49ht. Ergänzung zum Log vom 20.05.03: 1759z hrd ea4of, im89at, 59. Also keine totale Abschirmung. Interessanterweise war in beiden Fällen ea7 aus der Südwestecke Spaniens zu erreichen. Ein "aufgesplitteter" Strahl?
Es ist voller Rätsel.
Email from Dom, F6DRO, August 26, 2003
tnx for the very nice and interesting article about EA8 spectacular opening. I was there that day, and could'nt get anything from EA8 here in JN03TJ. No sign of ES in this direction for me, but I got several times bursts from EA8, as I get very often when MUF is high.
From my QTH EA8 is what I call a "strange" direction, with sometimes 144 openings when there was absolutely no indication (50 MHZ, Band 1 TV transmitters) that the path is open, making me often wondering : was this tropo or Es? But as signal level is high, I believe it is Es.
Another common phenomenon is that very often when path to EA8 is open, at the same time, path to 9H/I8 is open too, sometime exactly at the same time, sometimes one after the other.
I think I told you already , that during EA8 openings, very often the EA8 guys are working into JN03 and JN36/26 at the same time, making single ray calculation doubtfull.
About long distances qsos done from here, I worked several times stations around the 2800-3000km ( 5B-TA deeply inside the country) limit and heard once UD6DE, close to 4000km!
Tnx very much for your major contribution to Es understanding.
Email from Leigh, VK2KRR, August 26 and September 6, 2003
I have attached a graph of the ducts over Canary Islands at 12z on the 20th may 2003 for your reference.
The pink layers are ducts and if you look across you can see their height above sea level.
I have studied a lot of your thoughts on the May 20th 2003 contacts (boy that day must have been exciting!) by reading the PDF file. I checked vertical radiosonde readings from other locations along the radio path, but I found many to have very poor data readings. The balloons appear to have been raised in the lower altitudes to quickly, providing little information for the lower levels which most affect ground based radio paths.
I have just had to rebuild my computer and lost the data I had for the other areas along the path so I cant give an example at this stage, sorry. From memory for instance, some readings did not start until high above the ground and critical sampling in the 0 to 3000 m above ground level area was very sparse, maybe only 7 to 10 samples. Not know your area, but if I was using this data for checking stations in my area I would discard it as I would consider it to inaccurrate to draw conclusions from.
Not living in the northern hemisphere, so I wouldn't know, but from some checks I have done, it would seem that some claims of E skip etc on 2m (I am not saying this in your case) as cause of some long distance contacts is incorrect and its actually by ducting, though I could be wrong. It could also be easy to confuse E's with Ducts, as from my experience here, I could make a contact via a duct to a station 900 km away and stations at say the 300 to 500 km mark can be totally missed as the duct rises and falls along its path.
Keep up the good work on your website, your doing a great job.
PS- Check the Southern Ducting Notes on http://www.users.bigpond.com/vkvhffmdx/
Email from Keith, G4FUF, September 7, 2003
The case for a 2 "hop" Es link to EA8 is well founded. Firstly, there has been no evidence of MUF soundings over the path being quoted (other than in personal correspondance) and secondly the signal levels experienced in the opening are not consistent with either a tropo ducting scenario or any reflection scenario at the path midpoint or any location in between.
On this particular opening the path as experienced from G-land appears to have been confined to G7RAU in IO90, there are no other reports from the south coast of England that I am aware of. This is fairly consistent with 2 "hop" paths, they are generally very localised and of very short duration at ANY given QTH (often appear random locations). I use the phrase "hop" with tonge in cheek, I personally believe that these are all chordal modes with an optimum "jump" point along a "MUF ridge" at 1600-1800km.
I can only provide information and observations relevant to the G end of this path. Lets cover the first point. At the time of the QSO to G7RAU the band was observed as "open" from here (JO01gn) to IM67/68/85/86/87. This was consistent with my backscatter soundings at 1400-1800 km with broad temporal spread on the returns (=huge patch area). I would define "open" as the ionospheric loss is within some 30dB of the free-space loss, that is signals are "loud" and the band is open for even modest equipped stations at both ends of the path.I could also determine the second return at approximately 3180km, that is 2m was "open" for me at this distance, unfortunately further than the EA8 group. This situation is fairly "common" in the 2 "hop" scenario. (for the transatlantic guys, I have measured, but not recorded, returns at 3 "hops" on 22 July 2003 to ->UA9) (highlighted by the ed.)
Lets cover the second point. There is a lot of talk about sea reflections. Little published information is available at LF (<2GHz) since virtually all serious research work in this field remains classified. What information that can be freely found is consistent. If we take an ionospheric "mirror" reflection as unity (=1) then a sea reflection at 144MHz even at the critical angle is around 10E-7, i.e. around 70dB straight loss and add to this an additional 8-10dB scattering loss. Note this is a far field phenomena and not to be confused with near-field ground gain and completely ignores the effects of dielectric "lensing".
The observed EA8 signals are within 20dB of the free-space loss. This causes a serious headache when added to the analysis. As "hams" we are seriously underestimating our measurement capabilites.To get a better grip it is imperative that the signal strengths of stations is recorded accurately. Yes, we all know that "59" is exchanged on 99% of Es QSO's but in reality that is far from the truth. Simply plotting the geographical distribution of QSO's tells us little about the MUF behavior. ( I would add that potential 2 "hop" paths are identifiable, more later)
The FMDXer's and TVDX'ers monitoring Band I and Band II are in a better position to help. Not only do they appear to "log" signals somewhat more accurately, they also have the benefit of a wealth of data regarding the ERP capabilities of the signals they are actually monitoring. Likewise, it becomes common knowledge, or folklore as to what most stations are running on 144MHz.
Thats the easy part of the equation, what we desperately need is accurate signal reporting. Its difficult to give a simple example of why this is so important in solving paths. Needless to say some 20dB variation in observed signal strength at 50MHz can give rise to some 120dB change in ionospheric attenuation at 144MHz. We have all "experienced" the "S9" 1km away and "nil" here situation.
I am not attempting to add any controversial input to the discussion over and above the basic observation data experienced on the day.
73' Keith G4FUF
More information on Keith's radar experiments (email from September 7, 2003):
Editor's comment: You certainly have the same interest in Keith's radar experiments than I have. Therefore, I asked him to provide further information on this exciting subject. Keith writes:
My setup is somewhat similar to that of Leif SM5BSZ and is described on his web pages associated with auroral backscatter measurements. I first used this on 6m for finding the direction and distance for F2 but was later to discover some guys in the USA had been using this for a couple of decades.
No reflector is perfect, not even optical ones. Therefore there appears scatter at the reflection point(s) and not absorbtion of the signal energy. The backscatter signal is weak and on many occassions may not be detectable. On strong openings the backscatter can be quite loud. First I should say that to actually hear the return in the audio usually requires a "third party" observer nearby since the TX signal and return are very short. I use a TS850 and HB transverter (actually built inside the ATU bay). The filters are in the 12kHz position to accomodate the shortest pulse length. The TX pulse is injected into the audio. Both TX, PA and antenna C/O require electronic synched switching. The RX audio can be fed to a scope but sound cards are more favourable since the TX pulse can be generated also. Of course the range resolution is relatively crude if the pulse is too wide and is a basic limit unless more sophisticated techniques are employed. The limiting factor is RX recovery time in just about all measurements. I have attempted to use some DSP radios such as the TS570 with poor success. There is too much "preprocessing" taking place. Essentially the older the radio is the simpler it is to get a system up and running. Basically this is like a personal iono sounder except you are testing the actual MUF versus distance function and not the critical frequency.
Obviously, if you are just checking at one single frequency (144MHz) you generally know if the band is open or not when there is a complete absence of actual stations to make a QSO (as is the case a significant amount of the time). If you observe the return its possible to integrate the signal over several TX pulses. From the spread of the signal it is possible to make a guess at the "reflector" diameter. In major openings like the 22 July 2003 this can be really some 500-700km across (thats around 2ms of spread backscatter signal) at the first "hop" zone. When it is that large you can actually hear the backscatter spread on stations from behind or in front of you in the direction of the Es point. Of course this is not a definative measurement since the presence of Es formation is already known. However a fully automated system could be foreseen that checks over all azimuths and logs data versus time.
Western Europe is not well placed to determine most multi "hop" situations, from here they are mainly into desert or sparsely populated areas. The two most favourable locations from here are of course EA8 and UA6 ( and probably 4L). UA9 is also favourable but the ham concentration seems to be just around the main heavy industrial cities of the old soviet union and not in the countryside. This all leads to a deception as to what really seems to be taking place. The same situation of course applies in the USA. Regarding beam headings. I have noted on more than one occassion Es signals arriving off the direct path. They are usually weak and they may actually be a scatter component. We had such an observation this year with 9H1XT coming in at 190deg. As I write this I would have to check the date and time. From memory it was some 30 minutes or so after a main Es event to I8 area.
73's Keith G4FUF
Editor's comment: Interested readers are requested to view Leif's (SM5BSZ) webpage too, see http://www.antennspecialisten.se/sm5bsz/aurora.htm
Thank you very much to all fellow hams sending information and material.