Excluding areas above/below certain latitude

[JannisHoch]

For 'simple' methods like Hamon, the resulting PET estimates can exclude areas above and below a certain latitude (i.e. returns NaNs), even though the input data ranges between -90 and 90.

The area excluded changes depending on the date of the input data.

So probably that has to do with sunlight hours derived from latitude at a given date? Not per se wrong, but maybe good to either return 0 instead of NaNs (if that makes sense) or throw a warning so the user knows why this happens.

Also, probably good to specify this behaviour in the documentation somewhere.

[mvremec]

Hi Jannis,

This is a good point! Thank you for bringing this up!
We currently use the FAO-56 method to estimate daylight hours based on latitude.
However, as you wrote, it turns out this method returns NaN values for the subpolar/polar belt for the polar night period.

I will correct this in the new release.

Thanks,
Matevz

• [ correct daylight_hours for subpolar/polar belt]

[artessen]

Dear Matevz,

Thank you for your reply.
I believe the issue is still present in the current version, but in case you have not already fixed, here is a possible solution:

• Substitute lines 338-342 in pyet/meteo_utils.py with the following:

    if isinstance(lat, DataArray):
        lat = lat.expand_dims(dim={"time": sol_dec.index}, axis=0)
        return arccos(clip(-tan(sol_dec.values) * tan(lat).T, -1, 1)).T
    else:
        return arccos(clip(-tan(sol_dec) * tan(lat), -1, 1))

• Substitute lines 5-6 in pyet/meteo_utils.py with the following:

from numpy import tan, cos, pi, sin, arccos, mod, exp, log, nanmax, isnan, \
    where, clip

This essentially sets the range of values to be between -1 and 1 for the arccos computation and avoid the NaN issue.
Hope it helps
Best
Arthur

[mvremec]

Dear @artessen,

Thank you very much for this suggestion, and apologies for the late response—I somehow missed this comment.
I have implemented your suggestion in the dev branch, and it will be integrated into the v1.3 PyEt version.

Thank you again and all the best,
Matevz