Limb Sunspot Region Perspective Enhancement

[Drax Spacex]

Solar flares sometimes occur off the east limb of the solar disc before an active region (AR) number has been assigned.  Sunspots may barely be seen, but the magnetic classification cannot yet be ascertained because of the foreshortening of the image in that region.  Solar flare probability has increased, but that increase has not been accounted for in the daily predictions.

An image transformation can be applied to enhance the imagery in those compressed foreshortened regions by stretching them out in a deterministic manner using a polar to rectangular coordinate transformation.  Analysts could then better discern the detail and structure of the region.

As an example, here is AR2824 on the east limb, the day before the region was defined, from which a C1.14 originated.  This region of the image is foreshortened, and its structure is difficult to ascertain.

Performing a polar to rectangular transformation on the full disc, and then zooming in on that region of interest, the geometry and structure of the region is easier to see.

e.g. using imagemagick
convert latest_4096_HMIBC_202105180245.jpg -distort DePolar 0 -rotate 90 out.jpg

This deterministic transformation and resulting imagery could perhaps supplement existing data and analysis methods, enabling earlier definition of an AR emerging from the east limb and more accurate daily solar flare predictions.

Edited May 19, 2021 by Drax Spacex
HMIBC image timestamp 20210518024500

[Christopher S.]

This same technique is applied here: https://weather.cod.edu/satrad/?parms=global-halfdiskeastnorth-13-24-0-100-1&checked=map&colorbar=data and here: https://weather.cod.edu/satrad/?parms=global-northernhemi-13-24-0-100-1&checked=map&colorbar=data just as one example. It would be useful for getting a day or two ahead for prediction's sake.

[Drax Spacex]

Cool - thanks for those links.  Yes, there are scores of different types of map projections.  There's nothing really novel here, except that I have not seen it applied to the solar disc for this particular purpose.

This polar to rectangular transformation does work to elongate the content along the limb.  An advantage of using it is that it is a transform that is available in many graphics programs.   But physically it doesn't maintain a consistent scale and rotation for a zoomed area.   It compresses the height to width ratio by a factor of 2, and the rotation varies as a function of latitude, and scale varies as a function of longitude.

A better transform could be constructed, requiring some custom programming, that would present the solar disk as an elliptical shape which would preserve the same rotation and scale across any zoomed area for any location, such that any area of interest would appear as though it were being observed from a normal viewpoint along the equator with all lateral foreshortening removed.

Edited May 19, 2021 by Drax Spacex
lateral

[Christopher S.]

2 hours ago, Drax Spacex said:

Cool - thanks for those links.  Yes, there are scores of different types of map projections.  There's nothing really novel here, except that I have not seen it applied to the solar disc for this particular purpose.

This polar to rectangular transformation does work to elongate the content along the limb.  An advantage of using it is that it is a transform that is available in many graphics programs.   But physically it doesn't maintain a consistent scale and rotation for a zoomed area.   It compresses the height to width ratio by a factor of 2, and the rotation varies as a function of latitude, and scale varies as a function of longitude.

A better transform could be constructed, requiring some custom programming, that would present the solar disk as an elliptical shape which would preserve the same rotation and scale across any zoomed area for any location, such that any area of interest would appear as though it were being observed from a normal viewpoint along the equator with all lateral foreshortening removed.

If the images could be taken at a much higher resolution, this method of re-scaling the image could work very well. There would be less detail lost and therefore the compensations made in the transformation would be reduced. It doesn't have to be perfect, just clear enough to identify the magnetic structure. That's the whole point, isn't it?

[Patrick P.A. Geryl]

In the forum Coming Sunspots I explained that if there are no Triple alignments with Mercury or Venus, then there are no complex sunspots. The incoming sunspot on the northern limb is dead on arrival because the Triple with Mercury ended on the 21!

[Vancanneyt Sander]

Stretching the map won't do the trick indeed because sunspots look like a depression (optically lower then the surface itself) so when it's near the limb you don't see a good picture because of that, so you can't see the magnetic layout because of that.

More explanation can be found here: http://ephemeris.sjaa.net/1209/b.html 

[Drax Spacex]

I can certainly see in the AIA 093 - AIA 335 imagery what can appear to be depressions.   At the limb you may see a curved notched-out region (like the Apple logo).   But that's probably more the light/dark contrast of magnetic loops that's giving that impression.  For the actual sunspot,  I don't see these depressions in the AIA 1600, 1700 and magnetogram images (unless that Pac-Man curvature of the large blue spot is evidence of a depression on its left side).  This Wilson effect and the opacity of the penumbra may be dependent on the wavelength of the image.

The article describes sunspots as "optical sinkholes where we see down about a thousand kilometers lower than the surrounding photosphere."  Well, one thousand kilometers is a extremely shallow depression, virtually negligible in solar dimensions.  

In any event, it may take 5 degrees or so of rotation beyond the east limb for the magnetic layout to form up with sufficient detail around a sunspot for initial classification to be made.   But still I think some enhancement is useful to remove the foreshortening to more clearly see the features that are visible.

Edited May 23, 2021 by Drax Spacex
Pac-Man