Hi Folks, Here is (P.) Smith's list/stream of consciousness/fantasy for SN2011fe. These items are not necessarily in order of importance. * The primary polarization constraints for the system are the sudden 90 deg shifts relative to the surrounding continuum in the linear polarization PA that are associated some absorption features, particularly the strong, isolated Si feature. * This behavior is reminiscent of BALQSOs, where relatively strong polarization is seen in the absorption troughs. This is generally interpreted as the absorption trough blocking the unpolarized glare of the QSO and the flux in the absorption troughs being light scattered into our line of sight. The clear implication is that there are sight lines to the QSO that avoid the absorbing clouds. That is, from a different vantage point, we would see a "normal" QSO spectrum. * SN2011fe shows a very mild form of non-isotropic emission. * The 90 deg shifts in polarization PA drive one to the conclusion that there are 2 orthogonal polarization components competing. The constraint that the components are orthogonal is pivotal, as otherwise it is very hard to explain the sharp changes from one PA to the other and back (see epochs 2 & 3 for the best examples associated with Si, but epoch 1 also has a polarization feature in the blue that is orthogonal to the polarization seen in the red and in Si). * We are seeing the effects of the close competition between the 2 polarized components. Largely, they cancel each other leading to the generally low polarizations observed (0.1-0.4%). However, relatively small variations in the spectrum incident on the scattering material (electrons) lead to either one or the other component dominating in a spectral region. All of this evolves with time as well as spectrally. * It is important to remember that with the scattering components being orthogonal, when component 1 even slightly dominates, we observe its position angle in the polarization. When component 2 gets the upper hand by even the barest of margins, we observe a position angle 90 degrees from that of the first component. Transitions are "smeared" only by the instrumental resolution and noise. P, of course, goes to zero when the polarized fluxes from the 2 source are exactly equal. * The orthogonality of the scattering regions suggests an underlying structure to the early explosion/environment that may be a signature of Type 1as. My guess is that we have small imprints of the orbital/accretion disk plane and the poles of the system at early epochs of the event, but this is just rampant speculation. * The spectropolarimetry also implies that the absorption of light from the photosphere is not uniform in all directions (a mild version of the situation for BALQSOs). In other words, from another vantage point, we would measure different equivalent widths (EWs) of the absorption features at a given epoch. *** AN ILLUSTRATIVE EXAMPLE*** The polarization spectrum of SN2011fe during Epoch 3 can be reproduced quite well, especially for the Si absorption line, in the following manner: The total flux spectrum is decomposed into 2 components with orthogonal linear polarizations. Component 1 is given a Si EW that is only 0.8 (arbitrary choice) of that observed. Its FWHM and central wavelength is left identical to those of the observed (total light) spectrum. Outside of 6000-6350 Angstrom, the spectrum of component 1 is left identical to that of the observed spectrum. The spectrum for component 2 is then simply the total light spectrum - the spectrum of component 1. Therefore, it may be useful to remember that component 1 is the low-EW component, while component 2 is the high-EW component. Except in the region of Si, both components contribute an equal amount of flux to the observed spectrum (this, of course, is an arbitrary choice and simply eliminates one free parameter). Component 1 **MUST** have a polarization PA of about 150.5 (-29.5) or one will never get a good fit the observed PA within the Si absorption. Component 2 **MUST** have a polarization PA (60.5) orthogonal to component 1 because the data demand it. The free parameters to fiddle with are the intrinsic degree of polarization (P) and the Si EWs for one of the components. Of course, once P is chosen for one of the components, a true fitting procedure will find the best P value for the other component. The total flux spectrum constrains the EW of the other components once the EW is chosen for the first component. Epoch 3 is ideal for this exercise because the peak in P in the Si absorption line is intimately tied to the differences in EW and P between the two components. I haven't done Epoch 2 yet, but the lack of a strong peak in P for the Si line core probably means that P for component 1 is much lower than in Epoch 3. ***Notes concerning the model * A unique model is not possible given the large number of free parameters (P, EW, FWHM and the component fluxes/continuum shape, etc. can all be adjusted to come up with reasonable solution), though the polarization PAs are SET IN STONE. Because of this, the orthogonality of the 2 polarization components is the hook into the physics of the system. * Detailed modeling should deal with the entire observed spectrum. It would be interesting to apply the same treatment to the complex (and largely blended) line system to the blue of Si, but this would be grad student/theorist work, not someone with an observing schedule like mine. * The polarization of the continuum at the red end of the spectrum is reasonably stable throughout the 4 epochs. This could indicate that an ISP-like component (transmission), but it would be incredibly fortuitous to have it be exactly orthogonal to the "line" polarization. For this reason, the polarization of Component 2 has to be intrinsic to the source regardless of the polarizing mechanism. Also, subtracting this component out would lead to a significant blue polarized continuum during the first 2 epochs (i.e., no subtraction of a continuum component leads to a removal of all continuum polarization and a simplification of the situation). * It needs to be remembered that in epochs 1 & 4 the polarization of Si is essentially the same as for the surrounding continuum (Component 2 is dominant). However in epoch 1, component 1 shows itself in the polarization feature at about 4600-4900 Angstroms (and maybe in the blue wing of Si). By epoch 4, component 1 seems to have disappeared. So, component 1 may exist only during early times. Paul