Delete unnecessary coordinates in tropomi reader

[zxdawn]

The coordinates of longitude and latitude shouldn't be linked to themselves again.

And this PR can fix this error of scn.save_datasets() when loading longitude or latitude:

  File "test_save.py", line 77, in <module>
    scn.save_datasets(filename='./output/test.nc')
  File "/public/home/zhangxin/new/miniconda3/envs/s5p_tobac/lib/python3.8/site-packages/satpy/scene.py", line 1065, in save_datasets
    return writer.save_datasets(dataarrays, compute=compute, **save_kwargs)
  File "/public/home/zhangxin/new/miniconda3/envs/s5p_tobac/lib/python3.8/site-packages/satpy/writers/cf_writer.py", line 721, in save_datasets
    res = dataset.to_netcdf(filename, engine=engine, group=group_name, mode='a', encoding=encoding,
  File "/public/home/zhangxin/new/miniconda3/envs/s5p_tobac/lib/python3.8/site-packages/xarray/core/dataset.py", line 1689, in to_netcdf
    return to_netcdf(
  File "/public/home/zhangxin/new/miniconda3/envs/s5p_tobac/lib/python3.8/site-packages/xarray/backends/api.py", line 1107, in to_netcdf
    dump_to_store(
  File "/public/home/zhangxin/new/miniconda3/envs/s5p_tobac/lib/python3.8/site-packages/xarray/backends/api.py", line 1142, in dump_to_store
    variables, attrs = conventions.encode_dataset_coordinates(dataset)
  File "/public/home/zhangxin/new/miniconda3/envs/s5p_tobac/lib/python3.8/site-packages/xarray/conventions.py", line 806, in encode_dataset_coordinates
    return _encode_coordinates(
  File "/public/home/zhangxin/new/miniconda3/envs/s5p_tobac/lib/python3.8/site-packages/xarray/conventions.py", line 765, in _encode_coordinates
    written_coords.update(attrs["coordinates"].split())
AttributeError: 'list' object has no attribute 'split'

[djhoese]

My long term goal (a wish) has been to make the base reader support this type of logic (a coordinate variable depending on itself for coordinates). I don't expect you to fix it and I don't expect myself to have the time to do it right now either so I am OK with this PR if it fixes all of this CF writer stuff.

That said, I'm still not sure I understand why the CF writer chokes on this.

[codecov]

Codecov Report

Merging #1589 (2f5aeef) into master (d7fbbdf) will increase coverage by 1.22%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #1589      +/-   ##
==========================================
+ Coverage   91.62%   92.84%   +1.22%     
==========================================
  Files         246      251       +5     
  Lines       36034    37193    +1159     
==========================================
+ Hits        33016    34532    +1516     
+ Misses       3018     2661     -357     

Flag	Coverage Δ
behaviourtests	4.47% <ø> (+0.05%)	⬆️
unittests	92.98% <ø> (+0.89%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
satpy/readers/file_handlers.py	95.77% <0.00%> (-1.41%)	⬇️
satpy/readers/satpy_cf_nc.py	97.56% <0.00%> (-1.02%)	⬇️
satpy/tests/test_yaml_reader.py	99.53% <0.00%> (-0.31%)	⬇️
satpy/tests/writer_tests/test_awips_tiled.py	98.23% <0.00%> (-0.26%)	⬇️
satpy/resample.py	89.50% <0.00%> (-0.20%)	⬇️
satpy/readers/nucaps.py	89.14% <0.00%> (-0.05%)	⬇️
satpy/readers/__init__.py	97.42% <0.00%> (-0.04%)	⬇️
satpy/tests/test_scene.py	99.72% <0.00%> (-0.01%)	⬇️
satpy/plugin_base.py	88.00% <0.00%> (ø)
satpy/readers/ami_l1b.py	96.99% <0.00%> (ø)
... and 72 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d7fbbdf...2f5aeef. Read the comment docs.

[zxdawn]

@djhoese If these coordinates are not deleted, the problem is caused by https://github.com/pydata/xarray/blob/50d97e9d35bac783850827fa66ff5eb768e62905/xarray/conventions.py#L764-L765.

attrs["coordinates"] is a list ['longitude', 'latitude'] which doesn't support .split()

[djhoese]

🤔 So maybe the CF writer should check for bad coordinates and remove them? But then again, the CF writer shouldn't have to fix every piece of metadata...

[djhoese]

Looks good enough to me. This is a complicated topic that has a lot of parts (as far as the problem that is actually trying to be solved). This seems like the simplest fix for now until we get better coordinate handling in the base reader class.