Low-band DX'ing: The high-angle mode hypothesis - part 2

Following are the results of a quick test done this morning on K1JT 80m WSPR beacon. As for the previous post, I was using 2 antennas, my K9AY pointed NW and a magn loop, this time oriented NE, so roughly 90 degrees away from Joe's direction. The magn loop pattern at 90 degrees is particulary interesting because its response at low angle is minimum.
Following is the plot of the measured SNR's around sunrise. Although the mag-loop is 90 degrees away from the TX-direction, it still takes the advantage just after sunrise (SR at 0645z). (you may click on the plots to enlarge them). The above plot is for the SNR's measured on K1JT WSPR-beacon on 2 antennas , a K9AY oriented NW and a mag-loop oriented NE. Despite the limited number of points, and the mag-loop orientation, the trend is clear: Just after sunrise the mag-loop provides slightly better SNR's than the K9AY, which is likely to confirm the "high-angle propagation mode hypothesis".
This plot shows the pattern of the magnetic-loop, for various elevation angles. The front to side ratio goes from 12dB at 5degrees elevation (red pattern) to 1dB at 60 degrees elevation (blue pattern). The orange pattern is for 20 degrees, while the purple is for 40 degrees.
This plot shows the compared patterns of the K9AY and the magn-loop turned 90 degrees away from the transmitter direction. In blue the magn loop pattern a 60 degrees, in light blue at 20 degrees. In red the K9AY pattern at 60 degrees, in purple at 20 degrees. The respective max gains being almost equal, the 2 patterns cross at an elevation angle of approximately 60degrees. Below 60 degrees, the K9AY should provide better SNR's while the Magn-loop takes the advantage at angles higher than 60 degrees.

IMHO, those plots clearly show that a high angle propagation mode appears at sunrise. It would seem, still for this particular path at this particular moment, that the take-off angle is at least 60 degrees, or higher. This explains why stations using low horizontal antennas, get "better than expected" results on a few DX-paths, although a low horizontal antenna (below 1/4 wl) has a poor efficiency due to important ground losses. The little difference shown in the earlier published plot of DG0OPK signal (using a G5RV at 5m) vs mine (using a 1/2 wave vertical over a good soil), as received by VK7KRW on 40m long-path, may also have an explanation when looking at the "high angle mode" hypothesis.

For information I also looked at the SNR's measured on WA6MTZ (CA) 80m signal. On this particular 9300 kms path, there is no obvious benefit provided by the magn-loop. However as I have only a few spots, it might be too early to make definitive conclusions... so, more to follow !

Addendum: I experienced the same phenomena on YA/T61AA signal at sunset..
If you are interested by the topic, I found 4 online articles which mention the phenomena: By ZL1BPU, by G3CWI, by G0KYA and from antennex by ON5AU