Soil Conductivity Maps: Inputs, Not RF Performance
Soil Conductivity Maps: Inputs, Not RF Performance
A conductivity colour can be a useful model input. It is not, by itself, antenna efficiency, ground-wave loss, skywave gain or “dB left on the table.” Each of those outputs needs a different electromagnetic calculation.
A geological resistivity map may be rigorous, useful and correctly calculated while an “RF performance” legend built from it is still wrong. The error occurs when a material property is transformed into a logarithmic index and the index is presented as though an antenna or propagation model had produced an RF result.
The central distinction: conductivity describes a material. Antenna loss, field strength and transmission loss describe a complete electromagnetic problem. A logarithm does not create the missing physics between them.
1. The Logarithm Is Not the Main Problem
Suppose a map assigns each region the index:
CdB = 10 log10(σ/σref)
where σ is conductivity and σref is a stated reference conductivity. Because the ratio is dimensionless, this is mathematically legitimate as a defined logarithmic conductivity index. For example, changing from 50 mS/m to 5 mS/m produces −10 dB on that chosen index.
What it does not show is that received signal, radiated power, antenna efficiency or ground-wave field strength fell by 10 dB. No frequency, distance, antenna geometry, relative permittivity or path has entered the calculation. The result says only that one conductivity value is one tenth of another.
Engineering principle: a declared logarithmic conductivity index is mathematically valid. Label it only as conductivity. RF loss, gain, efficiency and field strength require an electromagnetic model that calculates those outputs for stated conditions.
The familiar 10-versus-20 rule also needs care. Decibels express a ratio; the multiplier follows the physical relationship being represented. Power ratios use 10 log10. Voltage or field-amplitude ratios use 20 log10 only when the relevant power relationship and impedance conditions justify it. Choosing 20 instead of 10 for conductivity would merely define a different index—it would not turn conductivity into field strength.
| Expression | What it can mean | What it cannot establish alone |
|---|---|---|
| 10 log10(P2/P1) | A power ratio in dB | Where the power went without a system model |
| 20 log10(E2/E1) | A field-strength ratio in dB | Power ratio when impedances or wave conditions differ |
| 10 log10(η2/η1) | An efficiency comparison | Absolute gain or path loss |
| 10 log10(σ/σref) | A declared conductivity index | Antenna loss, propagation loss or signal level |
2. Conductivity Is Only Part of the RF Ground Model
For a simple, homogeneous and isotropic medium, conductivity σ is measured in siemens per metre. Resistivity ρ is its reciprocal:
σ = 1/ρ
That conversion is valid only when both values describe the same material and conditions. Real soil can be layered, anisotropic, non-uniform and dependent on moisture, temperature, salinity and frequency. A geological or low-frequency resistivity estimate does not automatically equal the effective electrical constants sampled by a particular RF field.
In a common time-harmonic convention, conduction and displacement response can be combined in a complex permittivity:
εc = ε0εr − jσ/ω
tan δ = σ/(ωε0εr)
The sign of the imaginary term reverses if the opposite time convention is used; the physics does not. The ratio σ/(ωε) shows why “conductive” is not an absolute RF category. The same soil can behave more conductor-like at one frequency and more dielectric-like at another. Relative permittivity, frequency and layering therefore belong beside conductivity.
ITU-R P.527-6, currently in force, models the electrical characteristics of water, sea water, ice, soil and vegetation through complex relative permittivity up to 1,000 GHz. It explicitly treats conductivity as frequency- and condition-dependent rather than as one universal colour.
3. Four “Ground” Questions That Need Different Models
Amateur discussions often merge several distinct mechanisms because each involves earth. Keeping them separate immediately clarifies what a map can predict.
| Question | Relevant ground interaction | Required output |
|---|---|---|
| Ground-wave coverage | Propagation along a path over a lossy boundary | Field strength or basic transmission loss for frequency, distance and path sections |
| Ground-mounted antenna efficiency | Near fields and return currents dissipating power locally | Accepted-power split between radiation and loss |
| HF elevation pattern | Amplitude and phase of the reflected field | Far-field pattern for antenna height, polarization, angle and ground model |
| Electrical earthing | Current from electrodes into soil | Electrode-system resistance or impedance under stated conditions |
The same conductivity can be favourable in one row and irrelevant or unfavourable in another. “Good ground” has no complete technical meaning until the objective is named.
4. What ITU-R P.368 Actually Predicts
ITU-R P.368-10 is the in-force ground-wave propagation prediction method for 10 kHz to 30 MHz. It predicts field strength—or corresponding basic transmission loss—for defined source, frequency, distance and Earth parameters. It does not turn one conductivity value into generic “antenna performance.”
For a homogeneous path, the calculation needs at least:
- frequency and distance;
- conductivity and relative permittivity;
- the recommendation’s reference-source and field definitions; and
- conditions within the method’s stated range of applicability.
P.368’s reference case uses a short vertical monopole on a perfectly conducting plane and specified radiated power. The antennas are on or near the surface and the method concerns the vertical ground-wave field component. Those reference conditions matter: the output is not the gain of the operator’s real antenna.
A route crossing different ground types is a mixed path. The P.368 method uses ground-section lengths, conductivity and permittivity, including the Millington procedure for mixed paths. A colour under the transmitter cannot represent the whole transmitter-to-receiver route, and swapping the order of land and sea sections need not produce the same result.
A legitimate dB map is possible. Run complete P.368 calculations at a stated frequency and distance, with documented path sections and electrical constants, then plot predicted field-strength difference or basic transmission-loss difference. The dB values then belong to that scenario—not to conductivity alone.
5. Local Antenna Ground Loss Is Not Path Loss
Near a ground-mounted vertical, strong reactive fields and return currents interact with lossy soil. A radial system changes where those currents flow and how much field penetrates the ground. The useful engineering result is radiation efficiency:
ηrad = Prad/(Prad + Ploss)
In a deliberately simplified series model this may be written as Rrad/(Rrad + Rloss), but only when both resistances are referred to the same current and reference plane. Real ground loss is distributed. Radial number, length, height, soil contact, conductor resistance, feedpoint geometry and feedline common mode all affect the result.
This is why a soil colour cannot state the loss of every vertical in that area. Two antennas on the same plot can have very different efficiencies because one has a dense radial screen while the other forces substantial current through earth.
Nor does feedpoint resistance solve the problem. A measured 50 Ω can be radiation resistance, ground loss, conductor loss, transformer loss, common-mode participation—or a combination. Match is not an efficiency meter.
6. Skywave Reflection Is a Third Problem
For an elevated HF antenna, the direct and ground-reflected fields combine. The reflection coefficient is complex and depends on polarization, incidence angle, conductivity, relative permittivity, frequency and the structure of the surface. Antenna height then sets the phase difference between direct and reflected paths.
More conductive ground can strengthen low-angle reflection in some geometries, but it does not add one universal number of decibels. Terrain slope, roughness, vegetation, water, layering and the finite area illuminated by the field can all matter. A far-field elevation pattern requires an antenna-plus-ground model, not a conductivity ratio.
Receive antennas provide a useful warning against the slogan “higher conductivity is always better.” A Beverage and related travelling-wave antennas depend on their interaction with lossy ground. Very conductive ground can change wave tilt, coupling, termination requirements and pattern. It is safer to say the response can degrade or change substantially over highly conductive terrain—not that every Beverage literally “dies at the beach.”
7. What the BGS Resistivity Dataset Really Contains
The British Geological Survey Resistivity dataset is a legitimate engineering-geology product. BGS describes a 1:50,000-scale spatial model for Great Britain, intended for uses including the earthing characteristics of the ground.
The underlying model classifies geological units and assigns ranges of saturation, porosity, clay content and pore-fluid resistivity. For each geological classification, thousands of parameter realizations produce a statistical distribution. The product reports values such as median, 20th and 80th percentiles for the expected resistivity of the upper 3–5 m.
The BGS user guide also explains that field soundings and airborne electromagnetic estimates support verification. That is not the same as measuring every garden. The companion limitations state that the 1:50,000 product must not replace site assessment and that local conditions can vary.
Map scale is not point certainty. Fill material, drainage, seasonal moisture, buried services, coastal salinity and water-table depth can make a station site differ from its representative geological polygon. A percentile range is more honest than one supposedly exact site value.
The BGS data can be an excellent starting input. Converting its resistivity to conductivity is straightforward. But the conversion does not invent relative permittivity, frequency dependence, antenna geometry or a propagation route.
8. There Is No Universal “Better Ground for RF”
- For a defined surface-wave path: higher conductivity often reduces attenuation, but frequency, distance, permittivity and mixed terrain determine the actual difference.
- For a ground-mounted vertical: reducing local loss usually improves efficiency, while a sufficiently dense radial screen can make performance less sensitive to bulk soil.
- For an elevated horizontal antenna: ground constants alter reflection amplitude and phase, so the useful result is an elevation pattern.
- For a Beverage: excessively conductive ground can alter the mechanism that produces the desired travelling-wave response and directivity.
- For safety earthing: lower resistivity may help an electrode system, but electrode geometry, layering and applicable safety rules remain essential.
“Better” is therefore an optimization statement with the objective missing. The question should be “better for which antenna, which mode, which path, which frequency and which metric?”
9. A Reproducible RF Map Workflow
- Name one output. Examples: predicted vertical ground-wave field at 1 MHz and 100 km, or radiation efficiency of a defined quarter-wave vertical.
- Select the applicable model. Use P.368 for a suitable surface-wave problem; use a validated full-wave or network model for antenna loss and pattern.
- Specify the complete input set. Include conductivity, relative permittivity, frequency, geometry, distance, path sections and reference conditions.
- Represent variability. Propagate plausible percentiles, seasons or measured site values instead of hiding uncertainty behind one colour.
- Keep local and path effects separate. Do not call surface-wave transmission loss “antenna efficiency,” or local radial loss “propagation loss.”
- Label the legend precisely. “Predicted field-strength difference at 1 MHz, 100 km” is defensible; “RF quality” is not.
- Publish model version and assumptions. Another engineer should be able to reproduce the colour assigned to each location.
- Validate. Compare predictions with calibrated field-strength, accepted-power, current or pattern measurements appropriate to the claimed output.
A geological-only map can remain useful if it is titled honestly—perhaps “Modelled Near-Surface Geological Conductivity”—and its legend stays in S/m or mS/m with uncertainty bands. It becomes an RF performance map only after a declared RF model produces an RF output.
Bottom line: conductivity is an input. RF loss is an output. Between them sit frequency, permittivity, geometry, distance, path composition, boundary conditions and a model suitable for the exact question.
Primary technical references
Mini-FAQ
- Can conductivity be expressed on a logarithmic scale? Yes. A dimensionless ratio may be assigned a logarithmic index. It must not be relabelled as signal loss or antenna performance without a model connecting conductivity to that output.
- Does doubling conductivity improve a signal by 3 dB? Not as a general rule. The result depends on the mechanism, frequency, distance, relative permittivity, geometry and initial ground conditions.
- Does P.368 predict antenna efficiency? No. It predicts ground-wave field strength or corresponding basic transmission loss for defined conditions from 10 kHz to 30 MHz.
- Is the BGS map wrong? No. It is a modelled geological resistivity dataset. The misuse begins when its values are presented as direct RF performance measurements.
- Is conductivity alone enough for an RF ground model? No. Relative permittivity, frequency and often layering, moisture and temperature are also required.
- How should I test my site? Measure the quantity relevant to the claim: site resistivity for a ground model, antenna accepted power and loss for efficiency, or calibrated field strength for propagation.