Dipoles and Doublets Above Seawater: What Lower Heights Change

The practical maritime question is irresistible: if seawater is such a good RF ground, how low can I hang a dipole or doublet and still make useful contacts? The honest answer is more interesting than a height recipe. Salt water can reduce dissipative ground loss and produce a strong coherent reflection, so a modest-height antenna may behave very differently from the same wire above poor soil. But the sea is not a perfect mirror, and “lower” has meaning only after it is expressed in wavelengths.

Seawater Is a Complex Boundary, Not a Perfect Mirror

A field reaching the water produces transmitted and reflected components. The reflection coefficient is complex: its magnitude and phase depend on frequency, incidence angle, polarization and the water’s complex permittivity. Conductivity and permittivity in turn depend on salinity, temperature and frequency.

ITU-R P.527-6 provides the current model for complex electrical properties of seawater and other Earth surfaces. It treats conductivity as part of a frequency-dependent complex permittivity rather than as a single timeless “good ground” number.

At HF, representative seawater is usually much more conductive than typical dry soil, so reflection can be strong and surface loss can be lower. Strong is not perfect. Some energy enters and is dissipated in the water; phase is not fixed for every angle and polarization; waves and surface roughness perturb the boundary; and a shoreline is a transition between at least two different grounds.

The image-antenna picture remains useful if it is treated as a limiting model. For a horizontal electric dipole over a perfectly conducting plane, the image current has the opposite sign. At very small electrical height, direct and reflected fields therefore tend to cancel toward the horizon rather than guarantee a low-angle lobe. Real seawater changes the magnitude and phase of that reflection, but it does not repeal the interference geometry.

Translate Every Height into Wavelengths

A fixed 3 m or 5 m support does not represent one electrical height across 40–10 m. Using nominal amateur-band wavelengths:

Those are starting ratios, not performance ratings. Wire sag means different points can have different heights. On a boat, heel, pitch, tide and wave state move the geometry. On shore, the relevant surface beneath the near field may include beach, rock, quay, seawall and water at the same time.

For each frequency, record the height of the feed point and wire ends above the local water surface, the wire orientation relative to the shoreline and target bearing, and the distance to the water/land transition. A dipole broadside direction on the drawing can be reshaped by that mixed boundary.

Lower Ground Loss Does Not Guarantee Lower Takeoff Angle

Ground loss and pattern are different results. More conductive water can reduce power dissipated in the surface field and can make the reflected wave stronger. The same stronger reflection can also make height-dependent cancellation and reinforcement more pronounced.

A very low horizontal antenna commonly retains a strong high-angle response. That can be useful for short-range ionospheric work when the ionosphere supports the path, but high-angle radiation is not automatically NVIS and seawater is not an automatic DX lens. Low-angle response appears or disappears according to the direct/reflected phase relationship, which changes with exact height, frequency, angle and polarization.

Likewise, “half a wavelength is ideal” is not a universal rule. At one height the desired bearing may fall in a lobe; at another it may approach a null. A chosen height is an operating compromise among elevation pattern, azimuth pattern, structure, feed impedance, ground loss, safety and the bands that matter.

Reflection Redistributes Field; It Does Not Create Efficiency

Radiation efficiency is radiated power divided by accepted power. Directivity describes how that radiated power is distributed. Gain combines efficiency and directivity; realized gain additionally includes mismatch at the declared reference plane. A reflective surface can redistribute field into some directions without increasing total radiated power.

Seawater may reduce one loss term, especially compared with poor dry ground close to the antenna. The complete loss budget still includes:

• wire and joint resistance, including wet or corroded connections;
• insulator leakage and salt contamination;
• balanced-line or coax loss under the actual complex load;
• tuner, balun and common-mode choke loss and heating;
• current induced in a lossy hull, rigging, rails, mast or nearby wiring;
• unintended feed-line or station return currents; and
• transmitter foldback and mismatch at the chosen power reference plane.

A stronger field at one bearing can result from lower loss, changed directivity, changed polarization or changed common-mode participation. Measure enough of the system to separate them.

A Balanced Name Does Not Guarantee a Symmetrical Installation

A centre-fed dipole or doublet has a two-terminal port that can support equal-and-opposite differential current. On a boat or shoreline, the two sides seldom see identical environments. One end may approach a mast, stay, hull or shore while the other hangs above open water. Feed-line route, tuner case, protective bonding and control cables can add another return path.

Open-wire line is not immune. Its low differential loss can be an advantage under high SWR, but nearby metal, unequal spacing, wet salt deposits and asymmetric routing can convert energy between differential and common modes. A “balanced tuner” label also does not prove equal terminal voltage or negligible common-mode current under every load.

Map net feed-line current at several positions with a calibrated current probe. Record both conductor currents separately where practical. Measure the complex feed-point or tuner-plane impedance on each band, tuner settings, accepted power and component temperature. A clean match at the transmitter is not a current-balance measurement.

Boat, Shoreline and Open-Sea Models Are Different

An infinite homogeneous seawater plane is a useful reference case. It is not a marina, beach or yacht. The installed model should include everything large enough or close enough to carry relevant current or scatter the field:

• exact wire coordinates, diameter, conductivity, insulation and sag;
• feed point, balanced-line spacing and route, coax exterior and tuner network;
• water conductivity and complex permittivity at the modelled frequency, salinity and temperature;
• water depth or layered ground where penetration makes it relevant;
• finite shoreline, beach, seawall, dock or quay geometry when it occupies the near field;
• hull material and shape, mast, rigging, rails, lifelines and large cables; and
• actual height, heel, heading and target bearing.

LLNL’s NEC-5 description confirms that the code can model wires and conducting surfaces, homogeneous ground, loads, networks and transmission lines and can output currents, near fields and patterns. The NEC-5 validation manual specifically includes horizontal wires lowered toward a ground interface.

Use that homogeneous-ground case as a baseline. A finite shoreline or complex vessel may require a solver that can represent inhomogeneous material boundaries and conducting surfaces without forcing the scene into one infinite ground. Whatever solver is used, declare segmentation or mesh, ground formulation, losses, convergence tests and the reference used for gain.

Run sensitivity sweeps. Change height, water parameters, shoreline distance, feed-line route and hull/rigging conductivity one at a time. If a claimed advantage vanishes with a plausible tide, salinity or cable-route change, publish the range—not one flattering trace.

Compare Complete Systems at Equal Accepted Power

A horizontal doublet, vertical and end-fed wire are different complete systems. None wins by name. The vertical needs a defined return-current system; the end-fed wire needs a defined transformer and counter-current path; the doublet needs a defined balanced feed and tuner. Their patterns, losses, support needs and common-mode behaviour vary with installation.

For a controlled comparison:

Use a calibrated local field-strength receiver or stable remote sources at several bearings. One local point is not a 3D pattern. IEEE 149-2021 treats pattern, gain, impedance, range, instrumentation and uncertainty as one antenna-measurement problem.

For receive comparisons, record wanted signal and same-bandwidth noise separately and keep AGC, gain, attenuation, detector and bandwidth fixed. Seawater does not guarantee lower RFI. A maritime installation can be quiet, or it can couple strongly to chargers, inverters, shore power, navigation electronics and unintended feed-line paths.

Saltwater Hardware and Safety Are Part of the RF Design

Saltwater is a strong electrolyte. NAVSEA’s boats and small-craft manual documents the galvanic-corrosion risk created by dissimilar metals in electrical contact in seawater. Choose compatible marine hardware, isolate dissimilar metals where the approved design requires it, seal and strain-relieve connections, and inspect after salt exposure.

Do not improvise a connection between the antenna system and a vessel’s protective, lightning, DC-negative or shore-power bonding merely to obtain an “RF ground.” Those systems have safety and corrosion functions governed by the vessel design and applicable rules. Current U.S. Coast Guard small-vessel electrical guidance, for example, shows how an incorrect neutral/ground arrangement can create parallel current, shock risk and danger to people in the water. Local maritime requirements control.

Keep antenna wires, feed points and balanced lines out of reach; wet surfaces and salt deposits reduce insulation margins. Allow for vessel motion, rigging loads and emergency lowering. De-energize before touching or changing the system, and do not deploy or handle temporary wire antennas when thunderstorms threaten.

Assess RF exposure and contact-current risk for the actual duty cycle and power. In the HF near field, E and H fields are not necessarily related by the free-space impedance; ICNIRP’s RF guidelines require both components to be considered from 100 kHz to 30 MHz. A favourable far-field model is not an access-control plan.

The Useful Maritime Conclusion

Seawater can absolutely make a low horizontal antenna worth trying. Compared with poor soil, it can lower surface loss and change the reflected field enough to produce a materially different installed pattern. That is the opportunity.

The boundary is just as important: the same strong reflection remains height-, angle- and polarization-dependent. Three metres is extremely low on 40 m and much less so on 10 m. A balanced wire can acquire common-mode paths through its installation. A low loss line can still feed a lossy tuner or wet insulator. A clean local field reading can still represent one lobe rather than higher total efficiency.

Model the declared scene, sweep the uncertain parameters, map the currents and compare complete systems at equal accepted power. Then “works low above seawater” becomes an engineering result instead of another maritime proverb.

Primary and authoritative technical sources

• ITU-R P.527-6—frequency- and temperature-dependent complex electrical properties of seawater and other Earth surfaces.
• Lawrence Livermore National Laboratory: NEC-5—wire, surface, ground, network, current, field and pattern modelling scope.
• LLNL NEC-5 Validation Manual—validation cases including horizontal wires close to a ground interface.
• IEEE 149-2021—antenna pattern, gain, impedance, range, instrumentation and uncertainty practice.
• NAVSEA Boats and Small Craft Manual—marine galvanic-corrosion mechanisms and material boundaries.
• U.S. Coast Guard small-vessel electrical guidance—grounding, shore-power and parallel-current safety.
• ICNIRP RF Guidelines—RF-exposure quantities and HF near-field E/H requirements.

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