Very interesting..
Isn't that double exponential also called a Chapman layer?
https://heliophysics.ucar.edu/sites/default/files/heliophysics/resources/presentations/2014_Lab_4.pdf
In the work I was doing a few years ago, we were using the NRL model (SAMI3,
IIRC) They have a new one called NIMO, which extends higher. And also regrids.
https://www.swpc.noaa.gov/sites/default/files/images/u97/McDonald-Sarah.pptx_.pdf
One thing that's interesting is that there's a LOT more data now to measure the
upper parts of the ionosphere - GPS radio occultation from COSMIC 2 data for
instance.
We do a lot of GNSS reference station processing at JPL (and do the Radio
Occultation (RO) too), and it's always getting better. The challenge is that
multi frequency GNSS observations kind of only get you a "total electron
content (TEC) along a line" so you have to do something akin to tomography to
turn that into a 3D model.
It's pretty cool - you have a "well surveyed" ground station, so you get a TEC
along a line from station to GNSS satellite that sort of sweeps across the sky.
Then, with a LOT of processing you can "invert" those measurements to a model.
The RO satellite measurements get you lines from satellite to satellite,
which kind of sweep down from high to low altitudes, so they get much better at
high altitudes.
On Thu, 16 Apr 2026 17:27:28 -0400, Mike H <mph@sportscliche.com> wrote:
I'm giving a talk on this subject tomorrow (April 17) at the SEVHFS
Conference. A preprint of the Proceedings paper is here:
https://www.sportscliche.com/wb2fko/E1.pdf
Although this paper focuses on 50 MHz, the relevant physics is
applicable across the radio spectrum. It also shows how the
Appleton-Hartree equation can be applied to ray-tracing in the
ionosphere and the implications for polarization ellipticity.
There is a superb, open-access paper that describes experiments
measuring reflected beam ellipticity with links to real-time videos of
the effect:
https://www.mdpi.com/2073-4433/13/12/2027
Mike WB2FKO
>
>
>
> That *is* a rabbit hole..
>
> Get yourself a copy of PHARLAP from Australia and you can go wild ray tracing
> the ionosphere.
> https://www.dst.defence.gov.au/our-technologies/pharlap-provision-high-frequency-raytracing-laboratory-propagation-studies
>
> Yes, the mode splitting is super important when it comes to things like the
> "bright line" at the distant edge of the skip zone. And of course, the
> polarization sense flips with every sea/ground surface reflection so multi
> hops gets more confused.
>
> There are a fair number of people who have done polarization diversity
> receiving. A big X is popular (because it's easier to build than a big cross,
> and the two sloped elements have similar ground effects)
>
>
>
>
> On Thu, 16 Apr 2026 11:06:23 -0700, David Gilbert via TowerTalk wrote:
>
> Exactly so. That's a great comparison. That diffuse refractive
> ionospheric volume is also variable in density and position ... maybe
> not as extreme as the my local terrain or the big rocks below the
> surface of my lot, but the analogy holds.
>
> Your comment on polarization brings up an interesting question. I know
> that both vertically and horizontally polarized waves end up being
> elliptical as they get refracted by the ionosphere, but is one or the
> other treated more efficiently as it passes through? I asked that
> question to ChatGPT and got the answer below. I know nothing about
> O-mode versus X-mode, so I guess I have some studying to do.
>
> But maybe when we think of using both horizontal and vertical
> polarization to better receive what really is an elliptical signal, the
> difference in the path absorption is also having an effect for
> transmit. I separately asked ChatGPT what the difference in ionospheric
> coupling efficiency was IGNORING GROUND LOSSES, and it replied that the
> difference was typically about 3 dB but could sometimes be a bit more
> than that.
>
> 73,
> Dave AB7E
>
> From ChatGPT:
>
> When an HF wave enters the ionosphere, it doesn’t just gradually become
> elliptical. It *splits into two characteristic propagation modes*:
>
> * *Ordinary wave (O-mode)*
> * *Extraordinary wave (X-mode)*
>
> These modes are defined by the plasma + Earth’s magnetic field
> interaction (Appleton–Hartree physics). Each mode:
>
> * Has a *different refractive index*
> * Follows a slightly *different path*
> * Experiences *different absorption*
>
> Both modes are generally *elliptically polarized*, regardless of what
> you launched from the ground.
>
> ------------------------------------------------------------------------
>
>
> Why launch polarization still matters
>
> Even though everything becomes elliptical up there, your *initial
> polarization determines how efficiently you excite those two modes*.
>
>
> Horizontal polarization (typical dipole)
>
> * Couples *strongly into both O and X modes*
> * Tends to produce *more balanced mode excitation*
> * Results in *better overall returned power*, especially at mid-latitudes
>
>
> Vertical polarization (typical vertical antenna)
>
> * Often couples *less efficiently into one of the modes*
> * More energy can be lost due to:
> o *D-layer absorption*
> o Ground losses on launch
> * Can result in *lower skywave efficiency*, especially at lower
> takeoff angles over average الأرض
>
> ------------------------------------------------------------------------
>
>
> The big practical effect: absorption differences
>
> The *D-layer* (especially during daytime) is where much of the loss happens.
>
> * The *O-mode is usually more strongly absorbed*
> * The *X-mode penetrates better*
>
> If your antenna launches energy in a way that favors the more absorptive
> mode, you lose signal.
>
> 👉 Horizontal antennas tend to distribute energy in a way that:
>
> * Ensures *some energy survives in the lower-loss X-mode*
> * Reduces total absorption loss
>
>
>
>
> On 4/16/2026 9:08 AM, Jim Lux wrote:
>> In some sense, it's like modeling ionospheric "reflection" as a single
>> height, when in reality it's a diffuse refractive effect spanning many
>> 10s-100s of km.
>> That said, I think one can come up with a frequency specific "single useful
>> number" given an assumed soil profile. You'd get an "effective depth" and a
>> rolled up reflection coefficient (which would be angle dependent).
>>
>> I suspect, also, that for H-pol, it's WAY simpler than for V-pol, just
>> because the "first surface" is so reflective at low angles.
>>
>>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>
>
>
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|