Centroid definition for FES2022

[eagle383838]

Hi all,

My kernel keeps crashing when trying to run the compute_tide code from the FES model. The errorlog can't give a clear answer what happened. I searched online and it appears a new version of pyfes came out in February 2026 (https://cnes.github.io/aviso-fes/changelog.html). Could this have broken the code of CoastSAT?

[kvos]

@eagle383838 it's like to be an "out of memory" issue as for compute_tide you need to load the netcdfs with the tidal components for the whole world. If you look in the documentation there is a way to clip the tidal component files to only the area you need to reduce the RAM usage when loading those files.
I haven't check the new pyfes version, is the old one still working?

[eagle383838]

@kvos Thanks for the quick reply. I reverted to your source code, and then pyfes worked fine. It was probably something to do in my fork.

However, reverting back to your code I found maybe another issue i want to address: The centroid coordinates. The cell says the following:

# get polygon centroid, coordinates to get tides from
centroid = np.mean(polygon[0], axis=0)
print(centroid)
# if longitude is negative add 180 (longitudes are from 0 to 360 in fes)
if centroid[0] < 0: centroid[0] += 360

But I found that pyfes2022 can handle negative longitudes, therefore i think adding 360 (also not 180 as the comment says) would load your tidal component at a different location than desired.

[kvos]

can you test if it works with negative longitudes?

[eagle383838]

@kvos It does work with negative longitudes!

[kvos]

but does it work with both? if you run a negative longitude and then the same negative longitude + 360, do you get the same tide time-series?
It would be great if you can check this. If it's not necessary I don't want to remove that line which adds 360 as users that are running an older version of pyfes (like me) will then have compatibiltiy issues. Thanks