How to Select the Right Ground Plane Model

Created OnMarch 13, 2026

Updated OnMarch 13, 2026

By:Tony Golden

In computational electromagnetics, ‘one size does not fit all.’ The transition between a ground acting as a good dielectric versus a good conductor is governed by the Loss Tangent. This article provides a practical guide with embedded calculation and selection tools to select the most appropriate real ground plane model below an antenna elevated or connected to ground in AN-SOF.

Introduction

Configuring a realistic ground plane is a critical step in high-fidelity antenna modeling. While it may be tempting to always default to the most rigorous mathematical model, the optimal choice depends on a specific balance of numerical accuracy, simulation speed, and physical constraints such as antenna height and soil parameters.

In computational electromagnetics, “one size does not fit all.” A model that is accurate for a high-frequency dipole may produce instabilities for a low-frequency buried wire. This guide provides a systematic roadmap for selecting the most suitable real ground model in AN-SOF. We will use the Loss Tangent as a primary metric to classify your environment and compare the three core formulations available:

The Loss Tangent: Classifying the Medium

In antenna simulation, the ground is rarely a perfect conductor. To achieve high-fidelity results, AN-SOF utilizes advanced formulations to account for the finite conductivity ($\sigma$) and relative permittivity ($\varepsilon_r$) of the Earth. Understanding the electromagnetic behavior of the ground allows you to select the most computationally efficient and physically accurate model.

The behavior of the ground at a specific frequency is determined by the ratio of conduction current to displacement current, known as the Loss Tangent ($\tan \delta$).

Mathematically, it is expressed as:

$\displaystyle \tan \delta \,=\, \frac{\sigma}{\omega \varepsilon} \,=\, \frac{\sigma}{2 \pi f \varepsilon_r \varepsilon_0}$

Where:

$\sigma$: Conductivity (S/m)
$f$: Frequency (Hz)
$\varepsilon_r$: Relative Permittivity (Dielectric Constant)
$\varepsilon_0$: Permittivity of free space ($\approx 8.854 \times 10^{-12}$ F/m)

Classification Boundaries

The ground transitions through different states based on the value of the loss tangent:

Good Dielectric ($\tan \delta \leq 0.1$): Displacement currents dominate. The ground behaves like a low-loss insulator.
Poor Dielectric ($0.1 < \tan \delta < 1$): A transition phase where losses begin to significantly impact the phase and magnitude of reflected waves.
Transition ($\tan \delta = 1$): The point where conduction and displacement currents are equal.
Poor Conductor ($1 < \tan \delta < 10$): Conduction currents begin to dominate, though dielectric properties still influence the results.
Good Conductor ($\tan \delta \geq 10$): The ground acts primarily as a conductor (e.g., seawater or highly saturated, mineral-rich soil at lower frequencies).

Understanding the Ground Models

AN-SOF provides several methods to solve the electromagnetic interaction between the antenna and the ground. While there are “gray zones” where models overlap, the following principles guide the selection:

1. Reflection Coefficients (RC)

This method simplifies the ground interaction by multiplying the free-space field by a plane-wave reflection coefficient.

Strengths: Extremely fast and computationally efficient.
Limitations: It assumes the antenna is far enough from the ground that the wavefronts hitting the surface are essentially planar. It does not account for the complex interactions of near-field induction.
When to use: Wires at heights $h \geq \lambda/2$. It may be used for vertical wires below $h < \lambda/2$, provided there are no lossy connections to the earth.

2. Sommerfeld-Norton (SN)

This model is the industry standard for antennas operating in proximity to the earth’s surface.

Strengths: Highly accurate for horizontal and vertical wires at moderate heights.
Limitations: It may become numerically unstable or inaccurate when wires are extremely close to the ground (typically $h < 0.005\lambda$) or when wires are physically bonded to the soil.
When to use: General antenna structures at low-to-moderate heights without direct ground connections.

3. Sommerfeld-Wait (SW)

Based on the formulations developed by James R. Wait, this model is specifically designed for wires in close proximity to a lossy interface.

Strengths: It excels at modeling complex geometries, including oblique (slanted) wires, wires virtually touching the surface, and physical connections, whether lossy or lossless, directly to the earth.
Limitations: This model is intended for non-dielectric soils (poor to good conducting grounds). Its accuracy improves as the loss tangent increases.
When to use: This is the preferred choice for ground-connected wires, low-hanging horizontal wires, and antennas positioned over poor-to-good conductors.
Wire Screens: The “Radial Wire Ground Screen” model in AN-SOF utilizes the SW formulation by incorporating the surface impedance of the screen. This approach is significantly more rigorous than the simplified reflection coefficient approximations found in legacy software like NEC.

Ground Model Selection Guide

This summary helps you match your specific environment to the most appropriate ground model. First, use the Loss Tangent Calculator to determine your ground state. Then, refer to the Ground Plane Selector Tool to finalize your selection based on antenna height and ground plane connections.

Classification	Loss Tangent (tanδ)	Primary Behavior	Recommended Model in AN-SOF
Good Dielectric	$0 \le \tan \delta \le 0.1$	Insulating	SN / RC (based on height)
Poor Dielectric	$0.1 < \tan \delta < 1$	Lossy Insulator	SN / RC (based on height)
Transition	$\tan \delta \approx 1$	Balanced	SN or SW
Poor Conductor	$1 < \tan \delta < 10$	Lossy Conductor	Sommerfeld – Wait (SW)
Good Conductor	$\tan \delta \ge 10$	Conductive	Sommerfeld – Wait (SW)

AN-SOF Loss Tangent Calculator

Calculate the loss tangent of your ground plane to classify its electromagnetic properties and determine the best modeling approach.

Frequency (MHz):

Conductivity σ (S/m):

Relative Permittivity ε_r:

AN-SOF Ground Plane Selector

Navigating the “Gray Zones”

In computational electromagnetics, selecting a ground model is often a balance between mathematical rigor and numerical stability. “Gray zones” occur where the physical conditions sit on the boundary between two formulations, requiring an iterative approach to validation.

Recommended Workflow

Calculate the Loss Tangent: Determine the electrical state of your soil at your specific operating frequency.
Use the Ground Plane Selector Tool: Identify the suggested starting model based on your antenna’s geometry and height.
Cross-Validation: If your design falls within a transition zone (for example, at the edge of the SN/SW boundary), run the simulation using both models to observe how the results behave.
Empirical Refinement: Ultimately, no mathematical model is “perfect.” With the modern availability of affordable tools like NanoVNAs, it is highly recommended to measure the VSWR curve of the physical antenna. Comparing measured data with simulated results allows you to refine your model’s parameters, such as adjusting the local soil conductivity, to accurately predict real-world behavior.

Conclusions

Selecting the optimal ground plane model in AN-SOF is not merely a matter of choosing the most complex algorithm, but of matching the mathematical formulation to the physical reality of the environment. By utilizing the Loss Tangent to classify soil behavior and considering the antenna’s proximity and connectivity to the earth, antenna designers can ensure both high accuracy and computational efficiency. While the Sommerfeld-Wait (SW) model provides unparalleled rigor for ground-connected and low-hanging structures in conducting soils, the Sommerfeld-Norton (SN) and Reflection Coefficient (RC) models remain indispensable tools for elevated-antenna analysis. Ultimately, the integration of simulation results with empirical measurements enables the most reliable refinement of real-world antenna performance.