Polarization cross-talk from non-stationary sources of image jitter

arXiv:2611.08300 · astro-ph.IM, astro-ph.SR · Submitted 2026-11-05 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics. The day in astrophysics: observations, models and what the data forces.

Vera: We're looking at the paper: "Polarization cross-talk from non-stationary sources of image jitter".

Jocelyn: Here's the gist of the paper. We generalize a previous formulation for the estimation of polarization cross-talk induced by atmospheric seeing to the case of image jitter that must be described as a non-stationary, and generally non-isotropic, random process The mean-square error on the demodulated Stokes parameters is still given by Eq.

Vera: (6), but the matrix On is now determined by the twofrequency PSD Pαβ(ω, ω′) of the two-time ACF of the jitter, through Eq.

Jocelyn: (17) (or Eq.

Vera: (20), for a stepped modulator) General Formalism Extension The paper extends a formalism from Casini et al.

Title and authors: Vera: Let's talk about the title and the authors of this paper, "Polarization cross-talk from non-stationary sources of image jitter." It immediately tells you what we're tackling: polarization cross-talk caused by jitter that isn't stationary.

Jocelyn: And the authors are R Casini and A G de Wijn from High Altitude Observatory at the NSF National Center for Atmospheric Research. They’re researchers focused on atmospheric phenomena and how they affect our measurements of light.

Vera: So, what does this title actually mean for us when we look at data? It means we're moving beyond just seeing how atmospheric turbulence causes cross-talk in a steady way.

Subrahmanyan: It suggests that the image motion itself is a complex, time-dependent random process rather than something we can easily average out.

Jocelyn: And the implication is that our current methods might be missing some kind of systematic error introduced by these non-stationary movements.

Vera: They demonstrate how their new formalism converges back to the previous one when the jitter is stationary, which validates that part of their approach.

Subrahmanyan: But they show where the two formalisms diverge significantly when you have a non-stationary process, highlighting those differences in cross-talk estimation.

Jocelyn: The paper uses a specific example of a power spectral density derived from a circulant auto-correlation function to illustrate their point about stationary processes.

Vera: But the real meat is how they introduce off diagonal power spectral features at different frequencies when modeling non-stationary ones.

The paper's summary: Vera: So, to summarize the main points of this paper, they generalize a prior formulation for polarization cross-talk from atmospheric seeing to image jitter that is non-stationary and non-isotropic.

Jocelyn: They show how the new method works by interpreting the Stokes vector and its gradient in Eq. eight of their earlier paper not as ensemble averages, but as the actual values at a fixed point x naught <ref:2611.08300#pg2>.

Subrahmanyan: This means the image motion doesn't have to have zero mean anymore; it can even have a systematic drift.

Vera: Because of this, the two-time autocorrelation function they use corresponds to the second order moment of that displacement from x naught.

Jocelyn: And that second order moment is what allows them to calculate the mean square error for those inferred Stokes parameters, which includes that systematic drift we just mentioned.

Vera: So, essentially they replaced time-independent expectations with quantities directly tied to the instantaneous state of the image motion at a fixed point.

Subrahmanyan: The paper then models non-stationarity by considering things like pointing drifts or thermal transients as sources of this non-stationarity.

Jocelyn: They characterize this by looking at the two time autocorrelation function, which can be found using a corresponding two dimensional power spectral density obtained through a two dimensional Fourier transform.

The paper's improvements: Vera: Now let's look at what they improved in their methodology. They introduced Eq. seventeen or Eq. twenty depending on the modulator, to determine that error matrix On <ref:2611.08300#pg1>.

Jocelyn: This matrix On is now determined by the two frequency power spectral density of the two time autocorrelation function, which they show in Eq. seventeen or Eq. twenty for a stepped modulator version <ref:2611.08300#pg1>.

Vera: This means we are no longer relying on a single one dimensional power spectral density to define this matrix, which is what was done before.

Subrahmanyan: The improvement shows that the one dimensional power spectral density is not enough on its own to capture the full non-stationary character of the random process.

Jocelyn: They show that when power characteristics violate stationarity, there can be significant off diagonal structure in that two dimensional power spectral density.

Vera: And this off diagonal structure can include contributions to cross-talk that a simpler one dimensional PSD model would completely miss.

Subrahmanyan: So, the real improvement here is moving from a single measurement of power to analyzing the full two dimensional spectrum of the autocorrelation function.

Conclusion: Jocelyn: So, wrapping up this discussion on "Polarization cross-talk from non-stationary sources of image jitter," the main conclusion is that the one dimensional power spectral density doesn't have enough information to capture how non-stationary a process truly is.

Vera: The practical implication is that if you want to accurately estimate cross-talk in a real situation with jitter, you fundamentally need access to the two time autocorrelation function itself.

Subrahmanyan: Alternatively, you can generate an ensemble of statistically independent realizations of the jitter and use that to estimate the two time autocorrelation function directly.

Jocelyn: But they also say that even for a simpler stationary model, it can still be useful for identifying critical rates in your modulation cycle and giving you an approximate estimation of their cross-talk effects.

Vera: This approach lets you get a baseline understanding while still being able to pinpoint the important timing parameters.

Subrahmanyan: To conclude, this paper gives us a way to quantify how image jitter, when it's complex and non-stationary, introduces errors in our polarization measurements by using the two dimensional power spectral density.

Jocelyn: It’s a solid framework for handling more realistic observational noise sources in imaging.

R. Casini, A. G. de Wijn

High Altitude Observatory · National Center for Atmospheric Research

astro-ph.IM, astro-ph.SR

Submitted: 2026-11-05

Updated: 2026-11-05

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 80/100

The gist: We generalize a previous formulation for the estimation of polarization cross-talk induced by atmospheric seeing to the case of image jitter that must be described as a non-stationary…

Key concepts

Two-time ACF
This is a statistical measure that describes how the image position at one time relates to its position at another time. For non-stationary jitter, this function captures time-dependent statistical properties, such as pointing drifts or thermal changes, which are not constant over time.
Two-frequency PSD
This is a 2D power spectral density derived from the two-time ACF. It maps how the variance of image motion changes across both spatial and temporal frequencies. Analyzing this 2D map helps identify non-stationary features that simpler models miss.
Stationary Limit
When image jitter is stationary, its statistical properties are time-invariant. In this case, the 2D PSD simplifies to a diagonal form, allowing the use of a simpler 1D model that recovers previous results. This limit is useful for identifying critical modulation rates.
Matrix On
This matrix determines the mean-square error of estimated Stokes parameters. It is calculated using the two-frequency PSD derived from the two-time ACF, ensuring that systematic drift contributions are correctly accounted for in the final error estimate.

Terminology

Summary

We generalize a previous formulation for the estimation of polarization cross-talk induced by atmospheric seeing to the case of image jitter that must be described as a non-stationary, and generally non-isotropic, random process The mean-square error on the demodulated Stokes parameters is still given by Eq. p. 0

Where the paper says it

Box 3000, Boulder, CO 80307-3000, U. Version November 6, 2026 ABSTRACT We generalize a previous formulation for the estimation of polarization cross-talk induced by atmospheric seeing to the case of image jitter that must be described as a non-stationary, and generally non-isotropic, random process.

(6), but the matrix On is now determined by the twofrequency PSD Pαβ(ω, ω′) of the two-time ACF of the jitter, through Eq. p. 7
Where the paper says it

The mean-square error on the demodulated Stokes parameters is still given by Eq. (6), but the matrix On is now determined by the twofrequency PSD Pαβ(ω, ω′ ) of the two-time ACF of the jitter, through Eq.

(17) (or Eq. (20), for a stepped modulator)

General Formalism Extension

The paper extends a formalism from Casini et al. (2012) to image motion that is not stationary and possibly non-isotropic The key change involves interpreting the Stokes vector and its gradient in Eq. p. 0

Where the paper says it

In particular, the analysis in Paper I rederived those previous results by interpreting the seeing displacement of the image at the detector as a stationary and ergodic stochastic process with zero mean amplitude, and isotropic over the detector plane. With this paper we want to extend such a formalism to the case of image motion that is not stationary (hence, non-ergodic), and possibly non-isotropic (i.

(8) of Paper I as the unperturbed Stokes vector and its gradient at the fixed position x0, rather than ensemble averages This means that the image motion is no longer required to have zero mean, and instead, the two-time ACF of Eq. p. 1
Where the paper says it

(8) of Paper I, which no longer correspond to ensemble (long-time) averages over the realization of the image motion, but are directly the unperturbed Stokes vector and its gradient at the fixed position x0. As a consequence, the image motion is no longer required to have zero mean.

(12) corresponds to the second-order moment of that displacement from x0 The resulting error estimate yields the mean-square error of the inferred Stokes parameters, which includes any systematic drift of the image

Modeling Non-Stationarity

Non-stationarity in this study arises from contributions to image jitter that are not invariant under a time shift of the polarization modulation sequence, such as pointing drifts, mechanical and thermal transients, and periodic components The expectation values introduced in Sect. 2 must be interpreted as direct unperturbed Stokes vector and its gradient at the fixed position x0 For a generally non-stationary random image motion, the expectation value in the integral of Eq. p. 1

Where the paper says it

Therefore, what changes in the current study is the interpretation of S˜(x0) and ∇S˜(x0) in Eq. (8) of Paper I, which no longer correspond to ensemble (long-time) averages over the realization of the image motion, but are directly the unperturbed Stokes vector and its gradient at the fixed position x0.

(11) corresponds to the two-time ACF Γαβ(t, t′) = E(xα(t) xβ(t′)) This two-time ACF can then be characterized by a corresponding 2D PSD Pαβ(ω, ω′), which is obtained via 2D Fourier transform

Stationary Limit and Comparison

In the limit of a stationary random process, the 2D PSD is non-vanishing only along the diagonal, Pαβ(ω, ω′) = 2π δ(ω − ω′) Sαβ(ω), and Eq. (17) reduces identically to Eq. p. 7

Where the paper says it

In the limit of a stationary random process, the 2D PSD is non-vanishing only along the diagonal, Pαβ(ω, ω′ ) = 2π δ(ω − ω ′ ) Sαβ(ω), and Eq. (17) reduces identically to Eq.

(21), recovering the results of Paper I Numerical tests using a circulant 2D ACF show that cross-talk estimates from the full 2D model and from the 1D model applied to the diagonal PSD agree within typical numerical noise However, when power characteristics violate stationarity, significant off-diagonal structure in the 2D PSD may include a contribution to cross-talk that is entirely missed by the 1D PSD model

Practical Implications

The paper concludes that the 1D PSD does not carry enough information to recover the possible non-stationary character of the associated random process The application of the present formalism fundamentally relies on the availability of the two-time ACF, or which is equivalent, of the time series of the jitter from which Γαβ(t, t′) can be estimated An alternative is to utilize realistic physical models to generate an ensemble of statistically independent realizations and hence Γαβ(t, t′) The simpler stationary model can still be useful to correctly identify the critical rates of the modulation cycle and provide an approximate estimation of their cross-talk effects

References

Casini, R., De Wijn, A. p. 0

Where the paper says it

Casini & A. de Wijn High Altitude Observatory, NSF National Center for Atmospheric Research, P.

G., & Judge, P. p. 0
Where the paper says it

(2012) (hereafter, Paper I; see also Erratum) we analyzed the role of atmospheric seeing in the production of polarimetric cross-talk during a polarization modulation sequence. That study allowed us to cast into a general formalism, and a set of physical assumptions, prior analyses of the seeing effect on polarization observations (Lites 1987; Judge et al.

G. 2012, ApJ, 757, 45

Gray, R. M. 2006, Toeplitz and Circulant Matrices: A Review, Foundations and Trends in Communications and Information Theory, 2, 155

Judge, P. G., Elmore, D. p. 7

Where the paper says it

1561/0100000006 Judge, P. , Elmore, D.

F., Lites, B. p. 0
Where the paper says it

(2012) (hereafter, Paper I; see also Erratum) we analyzed the role of atmospheric seeing in the production of polarimetric cross-talk during a polarization modulation sequence. That study allowed us to cast into a general formalism, and a set of physical assumptions, prior analyses of the seeing effect on polarization observations (Lites 1987; Judge et al.

W., Keller, C. p. 7
Where the paper says it

, Lites, B. , Keller, C.

U., Rimmele, T. p. 7
Where the paper says it

, Keller, C. , Rimmele, T.

2004, Appl. p. 7
Where the paper says it

, Rimmele, T. 2004, Appl.

Opt., 43, 3817

Lites, B. W. 1987, Appl. p. 7

Where the paper says it

, 43, 3817 Lites, B. 1987, Appl.

Opt., 26, 3838

Papoulis, A. 1965, Probability, Random Variables, and Stochastic Processes (New York:McGraw-Hill) p. 7

Where the paper says it

, 26, 3838 Papoulis, A. 1965, Probability, Random Variables, and Stochastic Processes (New York:McGraw-Hill) This paper was built using the Open Journal of Astrophysics LATEX template.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved AI system can estimate polarization cross-talk for non-stationary image jitter by utilizing a two-time autocorrelation function instead of a single power spectral density. This allows the system to capture time-dependent statistical properties that are not invariant under time shifts, such as pointing drifts or thermal transients.

  2. The improved system can distinguish between stationary and non-stationary jitter processes during cross-talk estimation. It can do this by comparing predictions from two formalisms: the one based on a single 1D PSD, which is only valid in the limit of stationarity, and the extended formalism using the two-frequency PSD of the two-time ACF.

  3. The system can identify spectral features that violate stationarity in image jitter and quantify their contribution to polarization cross-talk. This is achieved by analyzing how off-diagonal power in a 2D PSD map affects the estimation, which is missed by simpler 1D models.

  4. The improved system can use a stationary model of image motion for qualitative assessment while still being able to identify critical modulation cycle rates. It can do this because the stationary model can be useful to correctly identify the critical rates of the modulation cycle, as well as providing an approximate estimation of their cross-talk effects.

  5. The system can be designed to directly estimate the mean-square error of inferred Stokes parameters, which includes systematic drift contributions. This is achieved by calculating Eq. (6) using a matrix On determined by the two-time ACF, rather than relying on time-independent expectation values as in the previous formulation.

Abstract

We generalize a previous formulation for the estimation of polarization cross-talk induced by atmospheric seeing to the case of image jitter that must be described as a non-stationary, and generally non-isotropic, random process. We demonstrate how the newly derived formalism converges towards the previous one in the limit of stationary random image jitter. In the case of non-stationary processes, we compare the predictions of the two formalisms, and illustrate the salient differences. For the sake of demonstration, we adopt for our analysis a specific example of power-spectral density derived from a circulant auto-correlation function, which describes a class of stationary random processes periodic over the length of the time series, and which we successively modified by the insertion of off-diagonal power spectral features at different frequencies in order to test the extended modeling framework for the estimation of polarization cross-talk.

Related papers