Fix LI L2 accumulated products `'with_area_definition': False` 1-d coordinates computation

ameraner commented 1 month ago

This PR fixes a bug in the li_l2_nc reader when requesting accumulated products as 1-d arrays (by using reader_kwargs={'with_area_definition': False}). Note that it is anyway recommended to request accumulated products as 2-d arrays using reader_kwargs={'with_area_definition': True} (as discussed in https://github.com/pytroll/satpy/pull/2789).

The bug was due to a missing *= -1 on the azimuth geos angle coordinate (used inside the 1-d coordinates calculation), needed to account for a convention difference between MTG data and Proj. Without this fix, the coordinates (particularly the longitude) are computed with fully wrong values.

This PR fixes this issue and adds a test to check the consistency between the 1-d computed coordinates and the according coordinates extracted from the 2-d arrays with AreaDefinition. Before this fix this test would have failed. It also updates the test filecontents for more realistic tests.

[x] Tests added

Checking the reader on a real-life case, comparing the resampled 1-d arrays and the 2-d arrays:

Before this PR, the lightning data is at fully wrong locations (left: resampled 1-d arrays, right: 2-d):

After this PR, the results are comparable (left: resampled 1-d arrays, right: 2-d)::

According code:

from satpy import Scene
from glob import glob
import matplotlib.pyplot as plt
import os
import numpy as np
from satpy.utils import debug_on;

debug_on()
from satpy.writers import get_enhanced_image
import matplotlib

matplotlib.use("TkAgg")

ll_bbox = [33, 13, 60, 27]

# # initialise SEVIRI Scene
sevscn = Scene(
    filenames=['/tcenas/fbf/MSG/in/MSG3/OPE3/SEVI-MSG15/MSG3-SEVI-MSG15-0100-NA-20240501121242.333000000Z-NA.nat'],
    reader='seviri_l1b_native')
sevscn.load(['natural_color'])
sevscn_r = sevscn.resample('mtg_fci_fdss_2km', resampler='nearest')
sevscn_r = sevscn_r.crop(ll_bbox=ll_bbox)
seviri_img = np.moveaxis(get_enhanced_image(sevscn_r['natural_color']).data.values, 0, -1)

# initialise LI L2 acc product scene as 1-d array output
path_to_testdata = '/tcenas/scratch/andream/lil2a/'
scn = Scene(filenames=glob(os.path.join(path_to_testdata, '*AF*---*T_0073_0001.nc')), reader='li_l2_nc',
            reader_kwargs={'with_area_definition': False})
scn.load(['flash_accumulation'])
# resample 1-d array onto grid
scn_r = scn.resample('mtg_fci_fdss_2km', resampler='bucket_avg')
scn_r_c = scn_r.crop(ll_bbox=ll_bbox)

# plot resampled 1-d arrays on top of SEVIRI
crs = scn_r_c["flash_accumulation"].attrs['area'].to_cartopy_crs()
fig, axs = plt.subplots(1, 2, subplot_kw={'projection': crs})
ax = axs[0]
ax.coastlines(resolution='10m')

ax.imshow(seviri_img, transform=crs, extent=crs.bounds, origin='upper',
          interpolation='none')
ax.imshow(scn_r_c["flash_accumulation"], transform=crs, extent=crs.bounds, origin='upper',
          interpolation='none')

# initialise LI L2 acc product scene as 2-d array output (recommended, does not require resampling)
scn_adef = Scene(filenames=glob(os.path.join(path_to_testdata, '*AF*---*T_0073_0001.nc')), reader='li_l2_nc',
            reader_kwargs={'with_area_definition': True})
scn_adef.load(['flash_accumulation'], upper_right_corner='NE')
scn_adef_c = scn_adef.crop(ll_bbox=ll_bbox)

# plot 2d-array on top of SEVIRI
crs = scn_adef_c["flash_accumulation"].attrs['area'].to_cartopy_crs()
ax = axs[1]
ax.coastlines(resolution='10m')

ax.imshow(seviri_img, transform=crs, extent=crs.bounds, origin='upper',
          interpolation='none')
ax.imshow(scn_adef_c["flash_accumulation"], transform=crs, extent=crs.bounds, origin='upper',
          interpolation='none')

plt.show(block=False)

codecov[bot] commented 1 month ago

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 95.95%. Comparing base (3b9c04e) to head (d29268c). Report is 1 commits behind head on main.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #2804 +/- ## ======================================= Coverage 95.95% 95.95% ======================================= Files 379 379 Lines 53888 53908 +20 ======================================= + Hits 51708 51728 +20 Misses 2180 2180 ``` | [Flag](https://app.codecov.io/gh/pytroll/satpy/pull/2804/flags?src=pr&el=flags&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=pytroll) | Coverage Δ | | |---|---|---| | [behaviourtests](https://app.codecov.io/gh/pytroll/satpy/pull/2804/flags?src=pr&el=flag&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=pytroll) | `4.09% <0.00%> (-0.01%)` | :arrow_down: | | [unittests](https://app.codecov.io/gh/pytroll/satpy/pull/2804/flags?src=pr&el=flag&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=pytroll) | `96.05% <100.00%> (+<0.01%)` | :arrow_up: | Flags with carried forward coverage won't be shown. [Click here](https://docs.codecov.io/docs/carryforward-flags?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=pytroll#carryforward-flags-in-the-pull-request-comment) to find out more.

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

coveralls commented 1 month ago

Pull Request Test Coverage Report for Build 9189835061

Details

21 of 21 (100.0%) changed or added relevant lines in 4 files are covered.
No unchanged relevant lines lost coverage.
Overall coverage increased (+0.002%) to 96.034%

Totals
Change from base Build 9157044280:	0.002%
Covered Lines:	51600
Relevant Lines:	53731

pytroll / satpy