High-Angle Radiation Is Not Automatically NVIS

Greg Mihran, KJ6ER, presents a deliberately high-angle PERformer configuration in his July 2026 antenna primer. That is an interesting antenna idea. My objection is the next step: putting a geographic coverage label on an antenna-pattern sector does not establish the propagation path.

On slide 57, the 15 m sweep labels the 60°–90° sector “NVIS / Skip Zone”. Slide 58 gives the associated 21.225 MHz model result: +2.38 dBi at 86°. An upward lobe can be real in the model while the proposed near-vertical radio path is unavailable. Pointing a flashlight at an absent ceiling does not create a reflection.

Give the Antenna Result Credit, Then Ask the Propagation Question

A low horizontal dipole is the familiar way to obtain useful high-angle radiation. It is not the only possible way. A vertical-and-wire system can redistribute radiation upward too. If the wanted regional path supports that launch angle and frequency, more gain in that direction is a genuine advantage.

The primer's acknowledgement that propagation matters is important. The problem is not that it forgot the ionosphere exists; it is that the fixed coverage sectors invite a conclusion the antenna calculation cannot deliver. High-angle gain is the antenna's contribution to NVIS, not the definition of a completed NVIS path.

The Radio Circuit Has Two Halves

Near Vertical Incidence Skywave uses a high-elevation single-hop path to reach stations beyond direct or useful ground-wave coverage without the long skip distance of a lower-angle HF path. A useful circuit requires all of the following:

• transmitter power at a declared reference plane and controlled feed-system loss;
• transmit gain or realised gain used consistently with that power reference;
• an ionospheric mode that returns the frequency over the path;
• survivable D- and E-region absorption and other propagation loss;
• receive realised gain and polarization coupling; and
• enough received SNR for the mode, bandwidth and reliability target.

A normalized polar plot shows relative shape. A coverage claim needs absolute gain with its matching power reference, ground and environmental assumptions, receive antenna data, propagation loss, noise and a success criterion.

Critical Frequency Is an Ionospheric Measurement

An ionosonde sweeps HF nearly vertically and records returned echoes as an ionogram. The NOAA/NCEI ionogram explanation describes the measurement, and the NOAA ionospheric-parameter catalogue lists the standard scaled characteristics.

foF2 is the ordinary-mode critical frequency of the F2 region: approximately the highest ordinary-mode frequency returned at vertical incidence under the observed conditions. A frequency above the relevant critical condition normally penetrates that region instead of returning vertically.

The antenna cannot raise foF2. Radiator length, height, radial count, match and zenith gain do not set electron density. foF2 changes with location, local time, season, solar illumination, solar cycle and geomagnetic disturbance.

The GIRO Digital Ionogram Database provides real-time and retrospective ionograms from participating observatories. GIRO’s ionogram-data guide explains trace polarization, automatic ARTIST scaling and why interpretation remains a difficult feature-recognition problem.

An ionosonde is a vertical measurement at its own location, not a direct measurement of every point along a nearby radio path. Use the nearest suitable stations, examine spatial agreement and trends, and keep the observation close in time to the radio test.

Oblique MUF and the Secant-Law Limit

A lower-elevation path meets a horizontally stratified layer more obliquely, so it can often use a frequency above the vertical critical frequency. The simplified secant law is:

If foF2 is 5.0 MHz, the simplified 80° result is only about 5.08 MHz. More antenna gain at 80° cannot make a substantially higher frequency return through an F2 region that does not support that ray.

Apply the same reasoning to the 21.225 MHz example. At 86°, the multiplier is only about 1.0024: ordinary-mode return in this teaching model would need a relevant critical frequency near 21.17 MHz even before adding an operating margin. This is not an observation that such conditions existed. It shows why invoking the higher MUF of a long oblique DX path does not rescue an almost vertical ray.

The NBS secant-law treatment presents the relationship as an approximation. Earth curvature, distributed refraction, layer tilt, the geomagnetic field and path control points limit the flat-layer sketch. The in-force ITU-R P.533-14 method predicts HF-circuit frequency availability, field strength, received power, SNR, lowest usable frequency and reliability from path, time, solar and system inputs.

Use a Frequency Window, Not a Permanent Band Label

ITU-R M.1795 describes NVIS as high-elevation single-hop propagation and states that it generally uses frequencies below the critical frequency. For its land-mobile planning context, it recommends operation up to about 80% of critical frequency to allow for short-term variation and absorption close to the critical condition.

That 80% value is a reliability guideline, not a new physical reflection threshold. The useful operating frequency still has a lower and an upper boundary:

The upper boundary is set by ionospheric support for the required steep path. The lower usable frequency is a link-budget result: D-region absorption, natural and man-made noise, antenna and feeder loss, transmitter power, receiver bandwidth, required SNR and target reliability all contribute.

Moving lower may restore near-vertical return while worsening absorption and noise. Moving higher may reduce those penalties but cross the path’s usable ceiling. The best NVIS frequency therefore changes during the day, across seasons and through solar and geomagnetic conditions.

For scale, applying an 80% planning margin to 21.225 MHz would require a critical frequency of about 26.53 MHz. That calculation is not a claim about today's ionosphere or a ban on unusual high-band short skip. It is a warning against treating an upward 15 m lobe as dependable NVIS coverage. Likewise, “below 10 MHz” is not an automatic pass: a lower frequency can still lie above the current steep-path ceiling or below the usable absorption/noise floor.

D-Region Absorption Can Close the Lower Side

The NOAA/NCEI ionospheric-region guide identifies the D region as the principal HF-absorption region. Solar X-rays can rapidly increase dayside D-region ionization, and solar energetic particles can produce strong high-latitude absorption.

The current NOAA SWPC D-Region Absorption Prediction uses GOES X-ray and proton data to provide operational guidance on affected frequencies and regions. It is an absorption product, not a complete NVIS predictor: pair it with ionograms and a path calculation.

Solar and geomagnetic events do not produce one universal “bands up” or “bands down” response. A flare can raise D-region loss on the sunlit side within minutes; energetic particles can impair polar paths; geomagnetic storms can restructure or depress F-region ionization in ways that vary with latitude and phase. Check current NOAA space-weather alerts and scales and then confirm the local result with current ionosonde data.

High-Angle Pattern, Ground and Environment

A low horizontal dipole often provides a broad high-angle lobe because antenna height and ground reflection reshape its free-space pattern. The result depends on height in wavelengths, soil conductivity and permittivity, ground loss, terrain and nearby conductors. A simple vertical monopole normally has a zenith null, although arrays or mixed vertical-and-horizontal current paths can produce different high-angle patterns.

For link work, use the installed realised pattern: mismatch, conductor and ground loss are included, and gain is resolved by azimuth, elevation and polarization. A feedline that carries common-mode current, a conductive support or nearby building can tilt a lobe, fill a null or add an unintended polarization.

The ionosphere is magneto-ionic. Ordinary and extraordinary modes can acquire different phase, polarization and absorption; Faraday rotation and ground reflection further change polarization coupling. A high total-gain number is therefore incomplete unless the transmit and receive polarization components and loss assumptions are stated.

Skip Zone: Geometry After Path Support

A skip zone is a region where the ground wave is no longer usable but the first returned skywave has not yet become usable. NVIS is valuable because a supported steep ray can reduce that close-in gap.

In a deliberately simplified flat-layer sketch, one-hop span d for effective height h and launch elevation α is:

This is geometry, not a coverage promise. It assumes the ray returns at that height. Virtual and true height differ, refraction is distributed, Earth curvature matters, ground wave may cover part of the region and received SNR determines the usable boundary.

When frequency is above the near-vertical usable limit, steep energy can penetrate while a lower-angle component still returns farther away through its larger obliquity advantage. The result is a skip zone despite strong high-angle antenna gain.

Short contacts on higher HF or VHF do not identify the path by themselves. Sporadic E, ordinary or extraordinary modes, sidescatter, backscatter and irregular structures can all shorten apparent range. Mode identification needs ionograms, path geometry, timing, bearings and observations from several locations. The ITU-R report on 26 MHz Digital Radio Mondiale and skywave interference provides a higher-HF system example.

What Each Piece of Evidence Can Prove

Choose the Frequency for the Path, Then the Antenna for the Angle

My practical recommendation is straightforward: for dependable regional HF, start with a frequency the ionosphere supports, then choose an antenna that puts useful power into the required high angles. Do not reverse that order and expect an antenna label to make an unsupported band work.

Nearby quality-checked ionograms, absorption information and a path prediction establish the operating context. Rapid antenna A/B observations at equal accepted power then help establish the directional advantage on a supported path. A complete station-to-station comparison can instead use equal power before the feed systems, counting their different losses explicitly. These answer different but useful questions; neither needs a universal antenna winner.

Keep the Upward Lobe; Drop the Automatic Coverage Claim

The PERformer high-angle idea can have a useful job. The model's upward radiation is not the objection, and a different low-angle antenna is not automatically better for regional work. The objection is turning that radiation direction into an unconditional propagation label.

The antenna determines where the power leaves. The ionosphere and the link budget determine whether useful power comes back. That is the distinction I want the reader to carry from the plot to the field. Improve the antenna when its directional gain is the limitation. Change frequency when the required skywave path is unavailable.

Questions, antenna-factor records or height trials to share? Contact RF.Guru.