PFSSpy not tracing field lines

[Richard Morton]

Hi,

I am having an issue with PFSSpy not tracing field lines, and was hoping for some pointers.

I can run the following example fine (https://pfsspy.readthedocs.io/en/latest/auto_examples/using_pfsspy/plot_aia_overplotting.html).

However, I am trying to do this using a different date and HMI data. No matter what I set as the seed points, the field line tracing just returns `fields lines` with the same coordinates for each entry. No warnings given.

The tracing works when I create the seeds with a reference frame that refers to the pfss data, i.e., seeds = SkyCoord(x, y, r, frame=pfss_out.coordinate_frame).

Full code I am using is below.

Thanks,

   Richard

-----------------------------------------

# Load AIA data

file_path='/Users/richardmorton/google_drive/sdo_cor_hole_data/2010-05-23/aia_lev1_193a_2010_05_23t01_00_06_63z_image_lev1.fits'
aia = sunpy.map.Map(file_path)

# Set seed locations

hp_lon = np.linspace(-100, 100, 5) * u.arcsec
hp_lat = np.linspace(-930, -900, 5) * u.arcsec
lon, lat = np.meshgrid(hp_lon, hp_lat)
seeds = SkyCoord(lon.ravel(), lat.ravel(), frame=aia.coordinate_frame)

# Download and import HMI Carrington map

car_rot = int(carrington_rotation_number(aia.meta['date-obs']))

time = a.Time('2010/01/01', '2010/01/01')
series = a.jsoc.Series('hmi.synoptic_mr_polfil_720s')
crot = a.jsoc.PrimeKey('CAR_ROT', car_rot)result = Fido.search(time, series, crot,
                     a.jsoc.Notify('X'))
files = Fido.fetch(result)

hmi_map = sunpy.map.Map(files[0])

hmi_map = hmi_map.resample([720, 360] * u.pix)

# Calculate pfss extrapolation

nrho = 25
rss = 2.5
pfss_in = pfsspy.Input(hmi_map, nrho, rss)

pfss_out = pfsspy.pfss(pfss_in)

# Trace field lines

tracer = tracing.PythonTracer()
flines = tracer.trace(seeds, pfss_out)

[Nabil Freij]

Hi Richard,

It's hard to say for sure what is going on and currently PFSSpy has been archived and we are actively working on creating a new version (with the original authors approval)  under the SunPy Project. 

Could you check the "r_sun" value in the AIA image and magentogram?

One potential problem is that the seed point transform from AIA to magnetogram is not behaving.
One avenue would be to try and debug the seed point coordinates to make sure they are sensible when transformed to the frame of the magnetogram. If you try manually transforming the seed points from one frame to another and see what the result is, hopefully that might shed some light on the problem.

Kind regards,

Nabil