Near-field around parabolic antenna?

Having a discussion at work about the $H$-field around a big parabolic antenna.

All of the safety tests done around the antenna only mention $E$-Field. They state in the radiating near-field the $E$ and $H$ are related so no need to measure them both. On the other hand all of the wire antenna have $H$-field included in their measurments.

I would have thought this was the wrong way around. Simple wire antenna have near-fields extending only a couple of wavelength but paraboic near-field can extend to a few $km$.

Only thing I can think of is that the near-field definition only applies to the front of the antenna - the back and surrounds are shielded by the parabola.

Is there a $H$-field, independent of the $E$-field in the immediate vicinity of a parabolic antenna?

-
As soon as you do not touch the wire (the source), you are in a free space where the relationship between $E$ and $H$ is unique. It does not mean there is no $H$. The magnetic field can be independent if it is static. The static part (if any) should be measured separately. – Vladimir Kalitvianski Dec 21 '11 at 10:47

I think the reason you are confused is because of the different notion of "near field". The near field that extends for kilometers in the parabolic antenna case is not a "near field" like an electrostatic field, it is just a radiative field which is collimated and hasn't travelled far enough to diffract over an appreciable area.

If you have a wire whose voltage alternates, you have an electrostatic field near the wire which changes with time, which follows the electrostatic law of $1/r^2$ decay, so $1/r^4$ energy density. The same holds for currents which change with time. Near enough, the law of the magnetic field is just by magnetostatics. This field falls away fast, and further away, you have a radiation field which falls off as 1/r, so that the energy density goes as $1/r^2$. The two cross over at a certain scale.

The notion of "near field" you are using in your case of a parabolic antenna is different--- it is not a local electrostatic or magnetostatic field, it is just the idea that the field is collimated, and so all the energy is going in the same direction, and it hasn't had time to spread away. This field is still purely radiative, just directional, and E and B are proportional. There is a crossover to a different power-law for falling off, but this crossover is purely in the radiation field, and has nothing to do with the non-radiative components which are important only very close to the antenna itself.

This radiative "near field" is only going forward from the antenna, because it is reflected by the metal. Only radiative fields are reflected by a metal, non-radiative electric fields are screened, while magnetic fields just go through, (changed perhaps by the magnetic characteristics of the metal).

So you will see some residual magnetic field near the antenna, but it will be due to either ferromagnetic materials, or the currents driving the antenna itself. It will fall off as $1/r^3$, and it will be the same as for a short wire antenna ignoring the parabolic reflector entirely (unless you have an iron reflector!).

-

So reactive fields are present around a parabolic antenna, not just in front of the antenna. In this reactive region E and H are independent. Is there a boundary where this reactive component can be ignored?

It looks like my concerns about not measuring H in the immediate vicinity of the antenna are valid.

-