It looks like Edgar J. Twining may occasionally be getting a glimpse from DCF39 via the long path to Orford/Tasmania, about 23.6 megameters! Please have a look into the folder 
 df6nm.bplaced.net/LF/longpath
There is a DCF spectrogram from yesterday morning (14-12-28) which contains a little dash between 7:44 and 7:47 UT, only three minutes long. It is showing on exactly the same frequency as the usual nighttime trace (138830.3 Hz on the scale, slightly below the real transmit frequency 830.6). The dash occured 23 minutes after sunrise in Magdeburg (7:22), and about two hours before sunset in Tasmania. Searching my harddisk, I happend to find one other saved screenshot which possibly contains a similar event at Oct 4th, 5:54 to 6:00 UT.

 

I have to admit that I still have some doubts as the two candidate dashes are pretty close to the noise level. Anyway these were just serendipitous observations, as I normally did't think of watching Edgar's grabber
 https://dl.dropboxusercontent.com/u/101251787/testIDC.html
before our darkness. After informing Edgar, he has offered to start saving an additional screenshot at 9 UT, covering the period of potential long path reception. As you can guess, I was eagerly awaiting the spectrogram this morning, but unfortunately the noise background from thunderstorms happened to be prohibitively high.

 

On VLF frequencies around 20 kHz, long path propagation is not really unusual. In Europe, spectrogram ripples on NWC 19.8 kHz caused by interference between short and long path components are easily visible every day. The effect is emphasized by daytime reflections at the lower D-layer boundary, a predominantly oceanic path across the Pacific and Atlantic, and geomagnetic nonreciprocity favouring west-east direction.

 

But on LF frequencies this is a very different story. The daytime D-layer behaves as an effective absorber rather than a reflector. But (by definition) the long-path can never be completely in darkness. Thus such a LF long-path detection would at least be very unusual, and might even be the first one ever reported. The Orford receiver on the Tasmanian east cost seems to be in a very favourable location for this purpose, with a mostly oceanic path from Europe, including the chance to bring home the signal across the last couple of megameters in daylight by seawater groundwave.

 

73, and best wishes,
Markus