Fitting a power-law model to log(I) versus log(ν) at each pixel location and then converting this to an I(λ2) model

Dividing the Stokes Q(λ2) and U(λ2) values by the I(λ2) model

pFDS: the maximum amplitude of FDS. This provides a measure of the band-averaged fractional polarization (plus Ricean bias, which is insignificant here

see

peak: the peak Faraday depth—i.e., the Faraday depth at which FDS is maximized. This represents the Faraday depth from which the bulk of linearly polarized emission has emerged from along a particular sight line

The noise level in the FDS

The interfaces connect up smoothly to broader structures in the FD map of like sign, which extend over regions much larger than a synthesized beam area (see Figures, 6: #apjaaaec0f6 and, 21: #apjaaaec0f21)

The peak and pFDS morphological structure does not appear to change significantly with a factor of ~4 improvement in spatial resolution in both R.A. and decl. (see, Appendix: #apjaaaec0app1, Figures, 20: #apjaaaec0f20 and, 21: #apjaaaec0f21)

Crossed magnetic fields should generate frequency-independent depolarization, whereas we show in Section, 5.3: #apjaaaec0s5-3 that this is not generally what is observed

Stokes q and u generally show considerable nonsinusoidal frequency-dependent structure (see Figures, 13: #apjaaaec0f13 and, 14: #apjaaaec0f14)

Multiple emission components are generally required in/detected in the model FDF/reconstructed FDS. Typically, each component strongly influences the observed spectropolarimetric behavior

The components often emit over a substantial range in Faraday depth, based on the widths of components required in the best-fit model FDF and the location of rmclean components in the reconstructed FDS

Stokes q and u remain greater in magnitude across the entire λ2 range

While some degree of nonsinusoidal frequency-dependent behavior in Stokes q and u remains, the polarization behavior is generally dominated by single emission components with narrow -widths and low absolute peak Faraday depths

Where additional components are present in the model FDF, they are comparatively narrow in -space and contribute a comparatively small amount of polarized flux

Strong, oscillatory (as a function of λ2) depolarization is often observed toward the low-p patches. This is not consistent with pure Faraday rotation, or with frequency-independent depolarization. It implies that a Faraday-rotating plasma exists along the LOS that possesses a comparatively complicated magneto-ionized structure

It is rarely the case that σRM is well constrained to have a high value while Δ is well constrained to have a low value, or vice versa. Thus, in general, the {\boldsymbol{P}}({\lambda }^{2}) behaviors cannot be uniquely attributed to either external Faraday dispersion or internal differential Faraday rotation—again, the depolarizing medium must have rather more complex structure

That the low-p patches are (a) spatially resolved, (b) centered around enhancements in the value of | {\phi }_{\mathrm{peak}}| of typically up to ~50 rad m-2 (and sometimes greater), (c) centered around reversals in the sign of peak, and (d) located in regions where the sky-projected magnetic field orientation exhibits complicated eddy-like structure (though see our caveat in Section, 5.2.2: #apjaaaec0s5-2-2)

That complex polarization models are generally required to reproduce the detailed depolarization and repolarization behaviors observed toward the low-p patches, including the need for multiple Faraday-rotated emission components, possessing differential Faraday depths and dispersions of typically several tens, and up to roughly hundreds, of rad m-2 (Section, 5.3: #apjaaaec0s5-3)

That the low-p patches exhibit very low values of fractional polarization, implying that the depolarizing medium must either reside at or extend to within a small distance of the lobe surface (as opposed to being limited in extent to a small volume deep within the lobe

see Section, 5.1: #apjaaaec0s5-1)

In regions where the crest of a K-H wave is viewed from above (i.e., the part of the flow feeding into the vortex), the direction of {B}_{| | } will show a rapid reversal

A complex mixture of Faraday-active material will form in the K-H vortex, potentially Faraday-depolarizing emission from the background. Such regions will be spatially correlated with the regions described in the previous point

Waves should appear on the lobe edges as well as on the face, and in the former case, the characteristic projected scale height of the resultant peak depolarization structure should be roughly one-third of the inter-low-p-patch distance on the lobe face