Beyond First-Order Correction: Understanding Higher-Order Ionospheric Effects in GNSS

The ionosphere is one of the largest natural error sources affecting GNSS positioning accuracy. For decades, researchers and engineers have focused primarily on correcting the first-order ionospheric delay, which represents the dominant portion of ionospheric signal errors.

With the development of dual-frequency and multi-frequency GNSS technology, more than 99% of the first-order ionospheric effect can be removed under typical conditions, enabling centimeter-level positioning performance.

However, eliminating the dominant error does not mean the ionosphere has completely disappeared from the GNSS error budget.

As GNSS applications continue moving toward millimeter-level precision in areas such as geodesy, scientific research, autonomous systems, and advanced surveying, the remaining higher-order ionospheric effects become increasingly important.

In our previous article, How Multi-Frequency GNSS Eliminates Ionospheric Delay, we explained how multi-frequency observations remove the first-order ionospheric delay, which is primarily dependent on signal frequency and electron content.

This article goes one step further by exploring the physical mechanisms, mathematical models, and practical correction strategies for the remaining second-order and third-order ionospheric delays.

The total ionospheric delay can be represented as a series expansion:

I=I⁽¹⁾+I⁽²⁾+I⁽³⁾+⋯

where:

• $I⁽¹⁾$ represents the first-order ionospheric delay
• $I⁽²⁾$ represents the second-order ionospheric delay
• $I⁽³⁾$ represents the third-order ionospheric delay

While the first-order term dominates under normal conditions, higher-order terms become relevant when the highest levels of GNSS accuracy are required.

Fig1. Atmospheric Refraction Effect of Satellite Signals

1. Why Higher-Order Ionospheric Effects Matter

The first-order ionospheric delay is the largest component because it is inversely proportional to the square of the carrier frequency:

$I⁽¹⁾$∝ 1/$f²$

This relationship makes it the dominant ionospheric error source at GNSS frequencies, which typically range from approximately 1.1 to 1.6 GHz.

After applying the ionosphere-free (IF) combination using dual-frequency observations, the majority of this error can be removed.

However, several smaller residual effects remain:

• Second-order ionospheric delay: proportional to 1/f³, typically ranging from millimeter to centimeter levels.
• Third-order ionospheric delay: proportional to 1/f⁴, generally below the millimeter level but potentially larger under extreme ionospheric conditions.
• Signal bending effects: caused by geometric changes in the signal path, mainly affecting low-elevation observations.

Although these errors are much smaller than the first-order term, they can still influence:

• Precise Point Positioning (PPP)
• Long-baseline GNSS processing
• Geodetic reference stations
• Scientific GNSS applications

During periods of strong solar activity or ionospheric storms, especially near the geomagnetic equator, the second-order ionospheric delay can exceed 1 cm and become a meaningful component of the positioning error budget.

2. Second-Order Ionospheric Delay: The Influence of Earth's Magnetic Field

The second-order ionospheric delay is caused by the interaction between GNSS signals and Earth's geomagnetic field.

In plasma physics, this phenomenon is known as the magneto-ionic effect.

Unlike the first-order delay, which mainly depends on the total electron content (TEC), the second-order term depends on:

• Electron density distribution
• Geomagnetic field strength
• The angle between the signal path and Earth's magnetic field

The geomagnetic field causes the ionosphere to become anisotropic, meaning the signal propagation characteristics depend not only on the electron density and frequency but also on the magnetic field orientation along the propagation path.

The group delay form of the second-order term can be expressed as:

I₂=7527∫ NₑB cosθ ds /f³

where:

• $Nₑ$ is the electron density (electrons/m³)
• $B$ is the geomagnetic field strength (Tesla)
• $\theta$ is the angle between the geomagnetic field vector and signal propagation direction
• The integral is calculated along the GNSS signal path

2.1 Why TEC Alone Is Not Enough

For first-order ionospheric correction, TEC information is usually sufficient because the delay is directly related to the total number of electrons along the signal path.

However, the second-order ionospheric delay is different.

It depends not only on the amount of ionization but also on how the ionized medium interacts with Earth's magnetic field.

Global Ionospheric Maps (GIMs), such as those provided by the International GNSS Service (IGS), provide essential electron density information through TEC measurements.

However, the second-order correction requires more than simply applying slant TEC values.

The correction process must also consider:

• The satellite-receiver geometry
• The variation of the magnetic field along the signal path
• The interaction between electron density and geomagnetic field

2.2 The Role of the IGRF Geomagnetic Model

To calculate the magnetic field contribution, a reliable geomagnetic reference model is required.

The International Geomagnetic Reference Field (IGRF) is the standard model used to describe Earth's magnetic field.

The IGRF represents the geomagnetic field through a spherical harmonic expansion based on satellite observations and ground-based measurements.

For higher-order GNSS ionospheric correction, the IGRF provides:

• Three-dimensional magnetic field vectors
• Magnetic field strength
• Signal-field alignment information

This enables the calculation of the $B\cos\theta$ weighting factor required for second-order correction.

Without a geomagnetic field model, accurate second-order ionospheric correction cannot be obtained from ionospheric TEC information alone.

Under typical mid-latitude daytime conditions, the second-order delay is usually several millimeters on L1 signals. Near the geomagnetic equator or during disturbed ionospheric periods, it can increase to approximately 1–2 cm.

3. Third-Order Ionospheric Delay: The Effect of Electron Density Distribution

The third-order term originates from the next term in the expansion of the Appleton–Hartree refractive index. Physically, it reflects the nonlinear response of the ionospheric plasma at high electron densities. Its group delay form is:

I₃ = 2437 ∫ Nₑ² ds / f⁴

Notice that this term depends on the square of the electron density integrated along the path, not on the magnetic field. Because  $$Nₑ²$$grows rapidly in the F2-layer peak (around 300–400 km altitude), the third-order term is highly sensitive to the shape of the vertical electron density profile.

Fig2. Total Field Intensity(left) and Annual rate of change of Total Field Intensity(right)

In absolute terms, I₃ is small — typically less than 0.1 mm under standard conditions, and rarely exceeding 1 mm even during extreme storms. For this reason, it is often neglected in routine PPP processing. Nevertheless, for the most demanding geodetic applications and for completeness in theoretical modeling, it represents the next frontier of ionospheric error budget characterization.

4. Practical Correction Strategies for Higher-Order Ionospheric Effects

Correcting higher-order ionospheric terms in real time or post-processing requires external data products that go beyond the raw dual-frequency observables. The following resources are essential:

• Global Ionospheric Maps (GIM): Available from IGS analysis centers with 1-hour or 2-hour temporal resolution and 2.5° X 5° spatial resolution in latitude and longitude. These maps provide the TEC values needed to scale the second-order integral. They are typically distributed in IONEX format.
• IGRF Model: The International Geomagnetic Reference Field is updated every five years and is freely available from NOAA and the British Geological Survey. Because the geomagnetic field changes slowly, a single epoch’s coefficients are usually sufficient for GNSS correction purposes.
• Ray-Tracing Integration: Unlike the first-order term, which uses a simple mapping function, the second-order term ideally requires numerical integration along the true ray path through the ionosphere, using the IGRF field evaluated at each step. Simplified slant-to-vertical mapping approximations exist but introduce modeling errors of 10–20%.

5. Summary

The transition from meter-level to centimeter-level GNSS accuracy was made possible by the elimination of first-order ionospheric delay through multi-frequency combinations. The journey from centimeter to millimeter precision, however, demands that we confront the higher-order terms.

The second-order term, governed by the interplay between electron density and the geomagnetic field, is the largest residual and the most important to model. Its correction depends fundamentally on two external data sources: global ionospheric maps for the electron content, and the IGRF geomagnetic model for the field geometry. The third-order term, while smaller, completes the theoretical picture and depends on the detailed squared electron density along the path.

For everyday dual-frequency positioning, these corrections remain optional. But as the next solar maximum approaches and as geodetic science continues its march toward ever-tighter error budgets, understanding and applying higher-order ionospheric corrections will become an increasingly standard part of the precision GNSS tool kit.

📘 Recommended Reading

How Multi-Frequency GNSS Eliminates Ionospheric Delay

Learn how ionospheric delay affects GNSS signals, why frequency dependency creates positioning errors, and how dual-frequency and multi-frequency GNSS technologies eliminate the dominant first-order ionospheric effect to achieve centimeter-level positioning accuracy.