Reflection refraction or scattering?

Last Updated on May 6, 2025 by John Berry

There’s huge ambiguity across the radio amateur books, magazines, and sites when it comes to describing how it is that the ionosphere returns radio waves to Earth. The ambiguity is this. Do we call it reflection, refraction, or scattering? This ambiguity is also prevalent in writings on propagation in other media such as meteor trails and the aurora.

Many writers routinely talk of reflection. Others often call it scattering. And some refer to the mechanism as refraction. So, which is it? Here’s an attempt at clarity.

Nature of signals

First, we need some context.

An antenna has a broad response in both the vertical and horizontal plane. As a result, the transmission from an antenna is not a single pencil beam. The wave emits from all parts of the response. And it spreads and hence can be considered in a ray-tracing sense to be multiple pencil beams – within the main lobe and the various side lobes. This occurs in three dimensions. So, it’s a complicated picture.

And the signal received at the receiving antenna has traveled via many paths. Those conceptual pencil beams from the transmitter are acted on in the ionosphere. Many will arrive at the receiver. Some will be attenuated by the poor receiving antenna response at their angles of arrival. And some will be lost to space. The resulting energy at the receiver is the vector sum of all discrete rays – of all energy received from all paths. So, receiving is equally complicated.

Let’s investigate each possible mechanism – reflection, refraction, and scattering.

Reflection

Considering reflection, and using a single ray-tracing model, but accepting the complicated multiple ray model described above, the wave is launched from Station A towards the ionosphere (below).

The wave front aa-bb advances toward the ionosphere. In specular reflection, Snell’s law is obeyed, and the angle of arrival equals the angle of departure. The situation is described below. Snell’s law is obeyed for each point on the wave front and each point is turned such that the wave front departs as described by the reflected wave front cc-dd. One of the many rays is shown dashed.

This would require the ionosphere to be a flat reflecting surface like a lake. We do know that the ionosphere comprises several stratified layers of ionised gases or plasma. For reflection to apply, it would be necessary for the ionosphere to become opaque (lake like) somewhere in these layers.

Reflection is clearly a possibility, but we need to know more, and it’s not the whole story.

Refraction

So, what about refraction?

We do know that the ionosphere varies in density. It reduces in density with height above the main D, E and F regions. And there are several peaks in electron density coinciding with the main regions.

Refraction is where the wave front is turned as it propagates the ionosphere. The one part of the wave front is slowed relative to the other. For refraction to give a complete description, we would need the wave front to enter the ionosphere as aa-bb and exit as cc-dd. That would be a neat description – turned and heading back to Earth.

But Snell’s law (n₁*sin θ₁ = n₂*θ2 where n₁ and n₂ are the refractive indices of adjacent regions and θ₁ and θ2 the angles of incidence and refraction) forbids the angle of refraction to be >90 degrees. That is to say that the wave front can never be turned toward Earth. It will only ever be turned through some other angle(s) resulting in aa’-bb’ or the like.

The refraction idea seems useful, were it not for Snell’s law. So there must be something else going on that will explain.

Scattering

That brings us to scattering.

Many hams use scattering to describe the way in which a wave is returned from patches of the ionosphere, from ionised meteor trails and from the ionised aurora. So, is that correct?

Scattering is the case where there’s a plurality of discrete reflections. This might be in the ionosphere where there are refractive index mismatches or from particles. Backscatter would describe the case where the aggregate of returns proceed in the direction from which the wave came. Forward scatter describes the case where the aggregate of reflections is forward in the same general direction as it was travelling before experiencing scatter. This state is described in the figure below.

For scattering, there would need to be reflections either from particles or from many layers in the ionosphere. And reflections from layers would require proof of such Fresnel reflections during refraction.

Scattering criterion

So might the effect be caused by particle reflection? The empirical limit of particle scatter is where the size of particle doing the scattering is about one tenth of a wavelength across. Above that wavelength (below that in frequency) the scattering efficiency is just too low. Rain drops will scatter frequencies in the GHz amateur bands. So, the tiny particles found in the ionosphere won’t scatter HF, VHF, UHF or SHF waves.

Reflection refraction or scattering?

So is it reflection refraction or scattering? It’s complicated. The missing link in these mechanisms is the ionosphere’s behaviour as regions of plasma.

The Sun creates an electrostatically neutral plasma comprising ions and free electrons. Electrons are then added (from meteors) and these cause collective electron oscillation within the plasma. This resonant frequency of the plasma results in a frequency term in n, the plasma’s refractive index [n=√(1-ΚN/f²), where N is the electron density, Κ a constant, and f the frequency of incident wave].

When the electron density is low (weak ionisation) n=1 and the wave goes straight through. When n<1 (moderate ionisation) there is both refraction and modest layer junction reflection. When ΚN/f² >1 the refractive index is imaginary, the ionosphere appears metallic to the incident wave, and the wave is reflected.

Reflection occurs at a critical (heightened) electron density (as occurs with E-layer ionisation resulting in Es propagation) and a critical frequency (defining the frequency above which the wave is lost to space.

So, the answer is that all three mechanisms occur, but it’s reflection that gives rise to the efficient MF/HF/VHF ionospheric propagation that many hams exploit.