coordinate transformation issue

[Shilpi96]

Hi @samaloney I am trying to make images for 2024-10-24 data. I get this error when I try to run the python script. Attached is the code.

stix_img_plt.py

[samaloney]

Yea, so this is because the source is off-disk. When the source is on disk, we promote the 2D coordinates to the full 3D coordinates based on the assumption that the feature is on the surface of the Sun (as defined in the header), but when off-disk, you need to make other assumptions, and I don't think it’s possible to make a generally correct assumption.

• Should absolutely do a better job of detecting this and raising a more informative and useful error at the least

What should and seems to work, but I will have to test more, is the same approach as used in sunpy. Using a screen PlanarScreen would probably be best for this, but that will depend on your use case. So compare

with PlanarScreen(solo):
     max_stix.transform_to(Helioprojective(obstime=vis_tr.center, observer=solo))
<SkyCoord (Helioprojective: obstime=2024-10-24T03:36:53.206, rsun=695700.0 km, observer=<HeliographicStonyhurst Coordinate (obstime=2024-10-24T03:36:53.206, rsun=695700.0 km): (lon, lat, radius) in (deg, deg, km)
    (-16.1103309, 7.99395634, 81465601.79941675)>): (Tx, Ty, distance) in (arcsec, arcsec, km)
    (-1383.15172938, -1460.89839817, 81469476.84406196)>

to

max_stix.transform_to(Helioprojective(obstime=vis_tr.center, observer=solo))
<SkyCoord (Helioprojective: obstime=2024-10-24T03:36:53.206, rsun=695700.0 km, observer=<HeliographicStonyhurst Coordinate (obstime=2024-10-24T03:36:53.206, rsun=695700.0 km): (lon, lat, radius) in (deg, deg, km)
    (-16.1103309, 7.99395634, 81465601.79941675)>): (Tx, Ty, distance) in (arcsec, arcsec, km)
    (nan, nan, nan)>

So if you change the script towards the like this it should work

#### Create a sunpy map for the clean image in Helioprojective
with PlanarScreen(solo):
     ff_cord = max_stix.transform_to(Helioprojective(obstime=vis_tr.center, observer=solo))

header = make_fitswcs_header(
    clean_map.data,
    ff_cord,
    telescope="STIX",
    observatory="Solar Orbiter",
    scale=pixel,
    rotation_angle=90 * u.deg + roll,
)

[Shilpi96]

But the plot I get from IDL does not match the one from python. The position of the sources are different.

[samaloney]

Ok but that a different problem and might not be "wrong". The IDL code uses vis_fwd fit to locate the flare so perhaps they are finding different flare. I'd suggest trying to use the location from the IDL code and manually use that for the flare location in the python code see if that works better.