The antenna is a "horizontal end-grounded line current" antenna; commonly referred to as a "horizontal end-grounded electric dipole". The word "dipole" is often used a bit too loosely, although that is a moot point. If you think "horizontal electric dipole" you get the general idea: its an electric dipole, not a loop antenna.
The end-grounding is a method of making the current flow in the full length of the antenna since, at LF, in a non-resonant structure, it is otherwise very difficult to get the current to go anywhere useful.
This is all as I stated in my 2003 article in the CREG journal, although at that time I was confused by some of the detail. The salient point is that the antenna does not *depend* on the current flowing in the ground. If you disconnect the antenna from the ground it stops working, but that is not because there is no current in the ground, its because there is no current in the antenna wire. It is possible that this could be considered a pedantic detail but it is not; it is an important distinction between what the currents actually "do". The current in the ground does not contribute to making a "large loop". The antenna is expressly not a loop antenna, it is identical to an electric dipole.
The point is quite difficult to explain without recourse to maths, and Im still wondering how to do so in an easy-to-understand way. An amusing aspect of it is that, to an enlightened electromagnetics expert, it is "obvious". This is presumably why it is not mentioned by, e.g. Hill & Wait, Burrows, etc in their papers on end-grounded electric dipoles. This bugged me for a long while - I just couldnt understand why they did not make allowances for the current flowing in the ground, till I realised the "obviousness" of it all. I have now justified it to myself: I just had to take a deep breath and a large pad of blank paper and start from Maxwell's equations. After a couple of days, I proved it, and ... yes it *was* obvious all along :-)
One application of this enlightenment concerns transmission of signals to cavers. The idea is to make as much current as possible flow in the antenna wire. Grounding is clearly a lot better than non-grounding but, even so, grounding is not very efficient since it is difficult to get an inter-electrode resistance of less than a few hundred ohms. Clearly a better way is to connect one end of the electric dipole back to the other end. But, of course, the problem with this is that it makes a loop, and the return current generates a field which almost cancels the field in the "outward" wire, so this is no good. Except that...
a) if the antenna diameter is a lot larger than a skin depth, and it is positioned correctly, then the receiver underground will not be able to "see" the far side of the loop so, in effect, it does not exist and you have a very efficient electric dipole!
And b) even a large loop is likely to be more efficient than an end-grounded dipole if the trade-off of lower resistance v. field cancellation is advantageous. I have yet to do the calculations but it could well be that if you are prepared to walk 200m to lay out a 100m grounded-dipole, then you might be better off if you are prepared to walk a 200m perimeter and lay out a 64m single turn loop.
Of course, this doesnt really help your RSGB friends, John, who are looking at above-ground transmission. For them, they just need to think that an end-grounded wire is precisely the same as a horizontal electric dipole.
The problem here is now that a HED does not radiate very well over a conducting half-plane. A horizontal magnetic dipole (i.e. a vertical loop) does radiate over a conducting half-plane but, that's not what you have got here. Any vision anyone had that a grounded wire formed a vertical loop in the ground, which is therefore "good" is, unfortunately, wrong - its not a loop!
I think Im probably far too busy to participate in collaborative projects. I dare not subscribe to the e-list or I will just get too involved!
