High-Angle Radiation Is Not Automatically NVIS
Antenna pattern versus ionospheric critical frequency.
A high-angle antenna can be valuable for short-range HF communication. It can put more accepted power into steep rays and less into low-angle directions that overshoot nearby stations. That is exactly what an NVIS-capable antenna should do.
But an upward lobe is not itself NVIS.
Greg Mihran’s July 2026 antenna primer repeatedly labels high-elevation sectors as “NVIS”, presents a quarter-wave system for “high angle radiation on 40M–6M”, and describes 10–30 MHz high-angle radiation as producing regional skip-zone coverage. A 15 m plot at 21.225 MHz even labels the entire 60°–90° sector “NVIS / Skip Zone”, while the model reports +2.4 dBi near 86°.
Those plots may describe where the modelled antenna launches RF. They do not show that the ionosphere returns it.
The antenna determines where RF leaves. The ionosphere determines whether a usable signal comes back. A high-angle lobe makes an antenna NVIS-capable; it does not make every frequency or every contact NVIS.
Give the high-angle antenna credit for what it really does
Near Vertical Incidence Skywave uses steeply launched HF energy to illuminate a relatively close region after ionospheric return. The precise useful angle range depends on the required distance, layer geometry and environment, but strong radiation at high elevation is normally desirable. A low horizontal dipole is the familiar example; a vertical system deliberately shaped for a high-angle lobe can also be useful.
The primer’s PERformer NVIS model may therefore establish a legitimate antenna result: in the stated model, its pattern has a broad upward lobe and a reported maximum close to 86°. Its comparisons may also show that one radial configuration redistributes more modelled radiation towards the zenith than another.
That is worth knowing. It answers the antenna question:
Gant(f, azimuth, elevation) — how much gain the installed antenna provides in a direction at a frequency.NVIS requires a second, independent answer:
Hionosphere(f, path, place, time) — whether the ionosphere supports that ray, where it returns, and how much absorption and variability it introduces.A strong first term cannot replace a missing second term. Pointing a flashlight at an absent ceiling does not illuminate the room by reflection.
What the primer actually establishes
| Primer section | Published claim or graphic | What that evidence can establish |
|---|---|---|
| Slide 49 | A 15 m plot at 21.350 MHz labels 60°–90° as “NVIS” and 30°–60° as “Regional”. | The model shows angular gain on one frequency. Those geographic labels require a contemporaneous ionosphere and cannot be inferred from elevation alone. |
| Slides 52–54 | The PERformer NVIS is presented for high-angle radiation on 40 m–6 m. Slide 53 separates “traditional NVIS at 10 MHz and below” from 10–30 MHz “high angle radiation” for regional coverage. | The geometry may be broadband and may launch high-angle energy. No fixed frequency boundary guarantees ionospheric return; even 10 MHz can be above the near-vertical usable limit. The later model table starts at 20 m, while 6 m lies outside slide 53’s own 10–30 MHz explanation. |
| Slides 55–56 | A 14.18 MHz model reports +0.67 dBi at 86° with a 78° elevation beamwidth, followed by claimed high-angle improvements of 7.76 dB and 26.4 dB against other radial layouts. | These are comparisons of selected model cuts. The largest number arises because the reference pattern has a deep near-zenith null; it is not a measured 26.4 dB communications improvement. The comparison contains no critical frequency, absorption or received-field result. |
| Slide 57 | At 21.225 MHz, a 60°–90° sector is marked “NVIS / Skip Zone” and the modelled peak near 86° is highlighted. | The plot shows where the model launches energy. Calling it NVIS requires the ionosphere to support a near-vertical 21.225 MHz return at the place and time concerned. |
| Slide 58 | Modelled gains and an 86° radiation angle are paired with field SWR screenshots from 20 m through 6 m. | SWR can validate an input match. It does not identify the propagation mode or prove regional skywave return on any band. |
| Slides 114–115 | Further 15 m graphics attach “NVIS” and “regional coverage” labels to elevation sectors. | Angle is one input to range. Virtual height, frequency, electron density, absorption and the link budget decide whether and where a usable signal returns. |
The problem is not that the model has high-angle radiation. The problem is changing an antenna-pattern label into a propagation result without solving or measuring the propagation path.
There is also an internal contradiction. Slide 53 reserves the term “traditional NVIS” for 10 MHz and below, but slide 57 labels the 60°–90° sector “NVIS / Skip Zone” at 21.225 MHz. Neither fixed label follows from the antenna pattern.
Critical frequency belongs to the ionosphere
An ionosonde transmits a frequency sweep almost vertically and records the returning echoes. From the resulting ionogram, one important parameter is foF2: the ordinary-mode critical frequency of the F2 region. NOAA explains how vertical-incidence ionograms are obtained, and its ionospheric parameter catalogue defines foF2 as the F2-layer ordinary-wave critical frequency.
In simplified language, foF2 is approximately the highest ordinary-mode frequency that the F2 region can return at vertical incidence. Below the applicable critical-frequency condition, a steep ray may be refracted back. Above it, the ray normally penetrates the layer instead.
Nothing in the antenna changes foF2. Not radiator length, not radial count, not SWR, not a modelled 86° peak and not the word “NVIS” printed beside a polar plot. Electron density, geomagnetic conditions, location, time, season and solar activity determine the ionosphere.
This is why the same antenna on the same frequency can provide excellent near-vertical coverage at one time and little or none at another. Forty metres is an amateur band. It is not a propagation mode.
foF2 is specifically an ordinary-mode vertical critical frequency. The extraordinary mode, sporadic E and irregular propagation complicate the real ionosphere, so it should be used as a disciplined planning parameter rather than an absolute declaration that no unusual return is possible.
Why the oblique-path MUF does not rescue a near-vertical ray
Longer HF paths can use a frequency above the local vertical critical frequency because their rays meet the ionosphere obliquely. The familiar simplified secant law is:
MUF ≈ fc sec(i)MUF ≈ fc / sin(α) in a flat, thin-layer sketch,where
i is incidence angle measured from the vertical at the layer and α is launch elevation above the horizon.The closer the ray is to vertical, the closer sec(i) is to one. In other words, the obliquity advantage disappears precisely where NVIS operates.
| Launch elevation α | Simplified MUF / fc
|
Meaning |
|---|---|---|
| 90° | 1.000 | No oblique-frequency advantage. |
| 86° | 1.002 | The primer’s highlighted angle is essentially vertical. |
| 80° | 1.015 | Only about 1.5% above the simplified vertical critical frequency. |
| 75° | 1.035 | Only about 3.5% above it. |
| 60° | 1.155 | A modest oblique advantage, with a longer nominal return distance. |
| 45° | 1.414 | A much larger advantage, but this is no longer near vertical. |
Suppose foF2 is 5.0 MHz. In this teaching model, an 80° ray is supported only to about 5.08 MHz. A 7.1 MHz ray would require an elevation near 45°, before real-layer corrections are considered. Making the antenna radiate more strongly at 86° cannot force a 7.1 MHz wave to return under those conditions.
The secant law is a first-order explanation, not a precision predictor. The original NBS treatment presents it as a rough relationship. The real F2 region is curved, distributed and magneto-ionic, and ITU-R P.533 uses a more complete empirical method for F2 paths. That added sophistication strengthens the central point: the propagation calculation needs ionospheric data that an antenna plot does not contain.
The primer’s band labels fail the critical-frequency test
ITU-R M.1795 describes NVIS as high-elevation single-hop propagation and says that it generally uses frequencies below the critical frequency. For dependable planning, it recommends operating up to about 80% of the critical frequency to provide margin and avoid the region close to critical conditions. It also notes that practical NVIS is below 8 MHz at best and can be below 3 MHz on high-latitude winter nights. A complementary ITU-R report on HF sky-wave systems notes that NVIS at 26 MHz is rarely possible.
The 80% figure is an operational reliability guideline, not a new law of nature. Using foF2 as the ordinary-mode F2 reference and applying that margin to the frequencies printed in the primer nevertheless provides a useful audit:
| Frequency shown or implied |
foF2 needed for the 80% planning margin |
What the high-angle plot permits us to say |
|---|---|---|
| 7.1 MHz / 40 m | At least about 8.9 MHz | The antenna may be well suited to NVIS when the short-path usable window includes 7.1 MHz. It is not automatic. |
| 14.175 MHz / 20 m | At least about 17.7 MHz | A modelled 86° lobe does not demonstrate return. This lies far above the range the ITU identifies for normal dependable NVIS planning. |
| 18.118 MHz / 17 m | At least about 22.6 MHz | Near-vertical ordinary F2 return would require an unusually high critical frequency; the pattern alone supplies no evidence that it existed. |
| 21.225 MHz / 15 m | At least about 26.5 MHz | The slide’s “NVIS / Skip Zone” label is not established by its modelled +2.4 dBi at 86°. |
| 24.940 MHz / 12 m | At least about 31.2 MHz | High-angle launch remains an antenna result, not proof of dependable near-vertical return. |
| 28.850 MHz / 10 m | At least about 36.1 MHz | Occasional short-skip mechanisms may occur, but routine F2 NVIS cannot be inferred from an upward lobe. |
| 52 MHz / 6 m | At least 65 MHz | A matched 6 m antenna with zenith gain is not thereby an NVIS antenna. Any unusual E-region or scatter event must be identified and measured separately. |
The table does not claim that a higher-band short-range contact can never occur. It shows why one cannot promise it from the antenna pattern. A model that contains wires, radials and soil but no time-varying ionosphere cannot turn +2.4 dBi at 86° into 15 m NVIS coverage.
“10 MHz and below” is not a universal NVIS switch
Slide 53 calls 10 MHz and below “traditional NVIS”. That is too coarse in both directions.
- A frequency below 10 MHz can still be above the current near-vertical usable limit and penetrate overhead.
- A frequency below the critical limit can still be too badly absorbed or too noisy for the required circuit.
- Under favourable conditions, a frequency somewhat above a habitual rule of thumb may return, but that result belongs to the conditions, not to a permanent antenna label.
- The proper operating frequency changes with place, hour, season, solar activity and disturbance level.
A responsible description is conditional: “This antenna provides substantial high-angle gain and is suitable for NVIS operation on frequencies supported by the current ionosphere.” That claim gives the antenna full credit without pretending it controls propagation.
There is a lower boundary too: absorption and noise
Staying below the near-vertical upper limit is necessary, but it is not sufficient. A usable circuit normally needs its operating frequency inside a window:
LUF < foperating < MUFnear-vertical pathThe lowest usable frequency, LUF, is a link-budget concept. D-region absorption rises strongly towards lower HF frequencies; atmospheric noise and man-made noise also matter. Transmitter power, antenna gain, bandwidth, required signal-to-noise ratio and interference determine whether a returned signal is useful. ITU-R P.533 consequently treats received power, noise, time variability and circuit reliability as separate parts of the prediction.
NOAA identifies the D region as the principal ionospheric region responsible for HF absorption. Its D-Region Absorption Predictions show how solar X-rays and energetic particles can sharply increase that loss.
This explains a familiar operational reality. Moving down in frequency may restore ionospheric return but worsen absorption and noise. Moving up may reduce absorption but cross the near-vertical upper limit. NVIS frequency selection is the search for a usable window, not the selection of an antenna whose product name contains “NVIS”.
Skip distance is not printed into the antenna pattern
Slides that label one elevation sector “NVIS”, another “regional” and another “DX” hide an essential conditional. Elevation helps determine where a ray would return if a layer supports it. It does not guarantee that return.
In a deliberately simplified flat-layer sketch, a ray reaching an effective height h at elevation α has a one-hop ground span of approximately:
d ≈ 2h cot(α)For an illustrative 300 km effective height, that geometry gives roughly 106 km at 80°, 161 km at 75° and 346 km at 60°. But those distances exist only after assuming that the ray returns at that effective height. Real ionospheric height is distributed and frequency-dependent; refraction bends the path; terrain and ground wave affect close coverage; absorption and antenna beamwidth affect the usable footprint.
When the frequency is too high for steep return, a lower-angle component may still return at a greater distance because it gains more from obliquity. The result can be the very skip zone that a genuine NVIS path is meant to fill. Increasing the 86° gain then sends more power into a direction the ionosphere does not return to the intended region.
Short skip on a high band is not automatic proof of NVIS
Sporadic E, extraordinary-mode propagation, scatter, irregular layers and other mechanisms can produce surprisingly short contacts on 15 m, 10 m or even 6 m. Amateur radio is interesting partly because the ionosphere does not behave like one perfect mirror.
Those exceptions do not validate a permanent 40 m–6 m NVIS label. A short contact might involve a low E-region layer, an oblique path, sidescatter, backscatter or a combination. Establishing the mode requires evidence: ionograms, timing, geometry, bearing, multiple receiving locations and consistency with a propagation model.
One contact proves that communication occurred. It does not by itself identify the path, measure antenna gain or establish a general coverage claim.
SWR and a NEC lobe answer different questions
The primer pairs modelled high-angle gains with analyser screenshots. Both forms of evidence can be useful:
- The NEC plot predicts the angular distribution for the mathematical model and assumptions actually entered.
- The analyser shows the impedance or SWR at its calibration plane.
Neither contains electron density, critical frequency, D-region absorption, virtual height, time variability, receiver noise or required reliability. A low SWR proves that the input can accept power at the measurement plane. A high-angle gain number predicts how the model distributes some accepted power. The received NVIS signal still needs a propagation path and a link budget.
Expressed broadly in decibels:
Received signal = accepted transmit power + directional antenna gains − system losses − propagation lossesIf the ionosphere does not return the selected ray, a favourable antenna term cannot repair the missing path.
What would demonstrate the claimed regional coverage?
A reproducible NVIS coverage claim needs antenna evidence and propagation evidence collected together. Publish:
- The antenna result. Absolute realised gain versus elevation and azimuth, frequency, accepted power, efficiency, ground model and installation details—not only a normalised lobe.
- The path. Transmitter and receiver coordinates, path lengths, bearings, terrain, date, UTC, duration and every test frequency.
-
The ionosphere. Contemporary nearby ionograms,
foF2, relevant E-layer parameters, minimum observed frequency and an estimate of the near-vertical MUF. - Absorption and disturbance. D-region conditions, solar-flare or particle-event status and geomagnetic context.
- The link budget. Accepted transmitter power, feed losses, receiver bandwidth, noise floor, antenna gains and the signal-to-noise criterion used to define successful coverage.
- Comparative measurements. Rapid A/B tests against a reference antenna at equal accepted power, preferably over several distances and azimuths.
- Repetition. Tests across several hours, days and frequencies, including cases above and below the observed near-vertical limit.
- Raw data and uncertainty. Signal levels, calibration records, failed paths and the uncertainty large enough to cover equipment and propagation variation.
The most persuasive experiment would switch quickly between the proposed antenna and a known high-angle reference while multiple regional receivers record calibrated signal levels. Simultaneous ionosonde data would show whether the frequency was actually supported. Repeating the test as foF2 crosses the operating frequency would separate antenna gain from propagation availability.
A practical NVIS operating workflow
- Check a nearby ionogram and note
foF2, the minimum recorded frequency and any sporadic-E trace. - Check current D-region absorption and geomagnetic conditions.
- Select a frequency below the near-vertical usable ceiling with a reliability margin, but high enough to avoid excessive absorption and noise.
- Use an antenna with useful absolute gain over the elevation angles required for the intended distances.
- Control mismatch, feedline loss and common-mode current so the predicted pattern is relevant to the installation.
- Verify coverage with stations or beacons at several known distances, and log time, frequency and signal-to-noise ratio.
- Change frequency when the ionosphere changes. Do not expect an antenna adjustment to repair an unavailable path.
This approach replaces a permanent band label with a measurable engineering process.
Takeaways you can trust
- NVIS is a propagation path, not a shape on an antenna plot.
- High-angle gain is useful and often necessary for NVIS, but it is not sufficient.
-
foF2is an ionospheric property; an antenna cannot raise it. - Near-vertical rays receive almost none of the MUF increase available to oblique paths.
- A frequency above the near-vertical usable limit generally penetrates overhead even when the antenna radiates strongly there.
- A frequency below that limit may still fail because of D-region absorption, noise or insufficient link margin.
- Ten megahertz is not a universal boundary between NVIS and non-NVIS operation.
- Labels such as “NVIS”, “regional” and “DX” cannot be assigned from elevation angle alone.
- SWR establishes a match, not a propagation mode.
- A NEC lobe predicts antenna behaviour under its assumptions, not the time-varying ionosphere.
- Exceptional short skip on higher bands must be identified; it does not make routine F2 NVIS automatic.
- A credible coverage claim combines absolute antenna data, ionospheric data, a link budget and repeated field measurements.
In Summary
The primer contains a potentially useful antenna idea: shape a vertical-and-radial system so that it launches substantial RF at high elevation. On a frequency supported by the current ionosphere, that pattern may contribute to effective short-range skywave coverage.
The overclaim begins when the high-angle sector itself is named NVIS, when an 86° model peak on 15 m is presented as regional coverage, or when one antenna is called an NVIS system from 40 m through 6 m without any critical-frequency or propagation analysis.
An antenna plot tells us where power is launched. foF2, path MUF, LUF, absorption and the link budget tell us whether useful power returns. Both halves are required.
A high-angle lobe makes an antenna NVIS-capable. Only the ionosphere and a successful link budget make the path NVIS.
Mini-FAQ
- Does high-angle radiation automatically produce NVIS? No. An antenna pattern describes where energy is launched. NVIS occurs only when the ionosphere can return that near-vertical energy at the operating frequency, time and location.
- What determines whether a near-vertical HF signal returns to Earth? The important factors include electron density, ionospheric-layer height, operating frequency, incidence angle, absorption and current geomagnetic conditions. The antenna supplies the launch angle; it does not control the ionosphere.
- What is the ionospheric critical frequency? The critical frequency, commonly expressed as foF2 for the F2 layer, is approximately the highest frequency that the layer can return at vertical incidence. A signal above it normally passes through rather than returning as NVIS.
- How is critical frequency different from MUF? Critical frequency applies to vertical incidence. Maximum usable frequency depends on the oblique path and is normally higher. Because NVIS is near vertical, its upper usable frequency lies much closer to the relevant critical frequency than a long oblique DX path does.
- Can 15, 12 or 10 metres provide NVIS if the antenna radiates straight upward? Usually not. On those bands the operating frequency is commonly above the near-vertical critical frequency, so much of the upward energy passes through the ionosphere. A high-angle lobe alone cannot make the path close.
- How can an operator determine whether NVIS is likely to work? Check nearby ionosonde foF2 data, select a frequency above excessive D-layer absorption but below the near-vertical usable limit, and verify the path with regional stations or controlled beacon measurements.
Want more technical RF content? Subscribe for new deep dives and lab notes.
Have a question or field observation? Contact RF.Guru.