|4.5 metre high UK acoustic mirror used as early warning system|
during WW1. Not so relevant for Long Delayed Echos, but something
which definitely reflects well. Wikipedia Commons, Paul Glazzard
It is likely that there isn't just one explanation, but several. I found it helpful some years ago to list all 15 possible hypotheses and then narrow them down to The Five Most Likely Explanations for Long Delayed Echoes. This list is now being referenced on the RSGB site for Long-delayed echoes (LDE) as well as in the Wikipedia article on the subject.
In this blog post there are examples of multisecond echoes on 21 and 27 MHz.
Echo on 21.590 MHzThis winter Thierry, F4EOB from the Paris region got in touch because he heard interesting LDEs on the VOA Chinese program on 21590 kHz. On February 12, he wrote "This transmitter is on from 9 to 11h UTC, transmitting from a small island in the Philippine sea. The echo is around 2.07 second delayed and it seems to me really strange. I get it during a few weeks so I decided to make some recordings. Last week I decided to tell it to my friend F6AGR here in Paris and since yesterday we have together maybe some more recordings but the LDE have disappeared, just the normal around the world echo remains!"
Here is his recording from February 4, 2012. I noted the curious coincidence that as the travel time around the world is 138 ms, 2.07 seconds happens to be exactly 15 times that number. However, I have always been skeptical to claims about many times travel around the world (number two of five on the list above), but here it seems to be one possible explanation.
Thierry, F4EOB then wrote on May 17: "I have used a preamplified tuned loop antenna. But I have to tell you that I'm not really sure about this LDE because I get it a lot of time. At this time on 15250 kHz, I also have this kind of LDE, and it's again with the same transmitter location, and the VOA in Chinese."
One could wonder if VOA broadcasts the same program from different sites and at the same frequency and that for some reason the distribution system sometimes adds a delay. In today's digital world, such delays are very easy to generate. This is even more likely for my next example.
Echo on 27.025 MHzThis is an echo recorded by CB station 20CB110 in Norway (Radøy 50 km North of Bergen on the West coast) on frequency 27.025 MHz on 11 May 12.21 UTC, 2008. The main echo is delayed by as much as 7.5 seconds. The FM-signal was emitted from a 5/8 wavelength vertical antenna with a radial system. Notice the double echo (easy to hear initially on the word ‘test’). The second delay comes about 0.45 seconds after the first one.
At this time there was sporadic E propagation to central Europe, but sun spot activity was at its minimum. This is one of 8-9 echoes heard at this time, selected because it is double. The delays were in the 7-7.5 second range except for one echo which was delayed by almost 13 seconds. The utterance which is spoken is: Test, signal rundt hele verden (2x) [Test, signal around the whole world (2x)]. The cause of the long delayed echo here is not known, but it could just be a simplex repeater (with delay) which can be found in the CB band in Europe.There is of course also a remote possibility that it could be due to e.g. mode-conversion in the ionosphere (number three of five on the list above).