Find your next learning step
A Kennaugh matrix is a real representation of how a polarimetric radar target transforms an incident Stokes description into a received one. Huynen parameters reorganise the same monostatic reciprocal information into physically motivated real quantities. Some entries reveal orientation, while others remain unchanged by a roll of the polarization basis. Reliable reading begins by separating invariant evidence from convention-dependent coordinates.
Open the chapter contents · Radar root · Gold Standard of Education · How X Works
Chapter contents
Choose the route closest to the decision, or read from Chapter 1. The chapters remain open; these grouped routes are a navigable index.
Route 1 · Define the decision
Route 2 · Work the ledgers
Route 3 · Transfer the method
Route 4 · Check limits and evidence
Route 5 · Ask and continue
CHAPTER 1 OF 12 · DEFINE THE DECISION
1. Locate the Kennaugh matrix in the measurement chain
A calibrated polarimetric radar begins with complex scattering channels. From a stated Sinclair, covariance or coherency convention, one can construct a real 4×4 Kennaugh matrix that acts on Stokes vectors up to the chosen normalisation.
The exact signs, factors and channel order vary among radar texts and software. A matrix copied without its convention cannot safely be compared entry by entry with another product.
CHAPTER 2 OF 12 · DEFINE THE DECISION
2. Separate coherent targets from averaged scenes
A deterministic reciprocal target can be represented by a complex symmetric 2×2 Sinclair matrix after calibration. Its Kennaugh representation obeys constraints inherited from that coherent target.
A multilooked distributed pixel is commonly described by a covariance or coherency matrix averaged over scattering realisations. Its Kennaugh-equivalent representation need not behave like one deterministic target, so a point-target parameter story may not transfer unchanged.
CHAPTER 3 OF 12 · WORK THE LEDGERS
3. Work a roll rotation on a parameter pair
A rotation of the linear-polarization reference axes mixes selected real parameter pairs. In one common two-component convention, a pair (u,v) rotates through twice the physical roll angle ψ.
| Physical roll ψ | Coordinate angle 2ψ | Rotated pair from (4,0) | Invariant norm |
|---|---|---|---|
| 0° | 0° | (4.000, 0.000) | 4 |
| 15° | 30° | (3.464, 2.000) | 4 |
| 30° | 60° | (2.000, 3.464) | 4 |
| 45° | 90° | (0.000, 4.000) | 4 |
CHAPTER 4 OF 12 · WORK THE LEDGERS
4. Distinguish Huynen parameters from invariants
Under monostatic reciprocity, a 3×3 Hermitian coherency matrix contains nine real degrees of freedom. Huynen’s parameterisation assigns nine real target descriptors to that information using a stated basis and normalisation.
Several descriptors are orientation dependent. A physical name attached to one parameter does not make its numerical value roll invariant. Invariant combinations must be derived from the transformation law, not guessed from the label.
CHAPTER 5 OF 12 · WORK THE LEDGERS
5. Change the scene from pure to mixed
Consider a coherent point target with a dominant rank-one coherency matrix, then an averaged pixel whose normalised eigenvalue shares are 0.55, 0.30 and 0.15. The second scene retains 45% of its span outside the dominant eigenchannel.
| Scene | Eigenvalue shares | Deterministic-target reading | Safer conclusion |
|---|---|---|---|
| Calibrated trihedral-like reference | 1, 0, 0 | Coherent rank-one model plausible | Use as a convention check |
| Distributed pixel | 0.55, 0.30, 0.15 | Single generator incomplete | Report mixed scattering |
| Near-degenerate pixel | 0.36, 0.34, 0.30 | Axes unstable | Prefer invariant spectrum summaries |
CHAPTER 6 OF 12 · TRANSFER THE METHOD
6. Build a roll-invariant reading sequence
First verify calibration, reciprocity and matrix positivity. Next record total span and the eigenvalue spectrum, which remain invariant under a consistent unitary change of polarization basis. Then inspect orientation-dependent coordinates after stating whether deorientation was applied.
Finally compare the invariant quantities before and after compensation. If span or eigenvalues change materially, the processing was not merely a basis rotation and needs investigation.
CHAPTER 7 OF 12 · TRANSFER THE METHOD
7. Connect decomposition and orientation routes
Use the Cloude eigenchannel guide and entropy–alpha guide for spectrum-based mixed-scene checks. The Pauli guide separates scattering coordinates from RGB display choices.
Use polarimetric deorientation and orientation angles to test basis-roll effects, and polarimetric coherence to keep channel consistency and calibration drift visible.
CHAPTER 8 OF 12 · TRANSFER THE METHOD
8. Treat roll compensation as a documented transform
An estimated orientation angle is itself uncertain and can be unstable in low signal-to-noise or mixed pixels. Applying compensation does not recover a unique physical pose when several mechanisms occupy the resolution cell.
Archive the pre-compensation matrix, estimated angle, rotation convention and post-compensation matrix. A deoriented colour composite without that lineage cannot show which changes came from geometry and which came from display scaling.
CHAPTER 9 OF 12 · CHECK LIMITS AND EVIDENCE
9. Know the limits and common misconceptions
A real Kennaugh matrix does not remove phase information; it repackages polarimetric second-order information according to a convention. It also does not make every matrix entry a basis-independent physical observable.
Huynen parameters are not a universal material dictionary. Speckle, averaging support, orientation mixtures, non-reciprocity, Faraday rotation, channel imbalance and model mismatch can all change or invalidate a simple target interpretation.
CHAPTER 10 OF 12 · CHECK LIMITS AND EVIDENCE
10. Use an evidence and source routine
Record sensor, frequency, incidence geometry, scattering convention, Stokes convention, normalisation, calibration, reciprocity test, averaging support, matrix construction, positivity check, Huynen convention, roll transform, deorientation angle, invariant quantities and uncertainty.
ESA’s SAR Polarimetry Basics develops Sinclair, Kennaugh, covariance and coherency representations, identifies the nine real Huynen parameters and states the associated pure-target constraints. ESA’s 8th Advanced Training on Radar Polarimetry provides the course context and linked theory material for current polarimetric processing and interpretation.
CHAPTER 11 OF 12 · ASK AND CONTINUE
11. Frequently asked questions
Is total span roll invariant?
Yes under a consistent unitary basis rotation and fixed normalisation. A processing change that alters calibration or scaling is a different operation.
Do all Kennaugh entries rotate at twice the roll angle?
No. Different blocks can be invariant or mix through different multiple-angle terms, depending on convention.
Can one Huynen parameter name identify land cover?
No. It is a model-derived descriptor whose meaning must be checked against geometry, scale, rank and independent evidence.
CHAPTER 12 OF 12 · ASK AND CONTINUE
12. Continue with a convention transfer check
Rotate the pair (3,4) through 30° in coordinate space and verify that its norm remains five. Then reverse the sign convention for the second coordinate and explain why an entry-by-entry comparison fails while the invariant norm survives.
Continue through the Radar root, the Gold Standard of Education and the How X Works hub.
Top · Contents · Previous · How X Works