Calculating Incoming solar radiation - Error in function ?

[manlaya]

Hello everyone,

I was trying to compute ET with this module and synop data in France and comparing methods (Turc, Hargreaves, Penman-Monteith, FAO56).
When computing Incoming solar radiation with calc_rad_sol_in() I had some differences with the code I used before to do such calculations.

Looking at the code, I saw that the equation used is :
(as1 + bs1 * n / nn) * ra

• with n as duration of sunshine ;
• and nn as maximum possible duration of sunshine

In this calc_rad_sol_in() function, n is optionnal and calculate with daylight_hours() when None, using day_of_the_year, sunset_angle, etc.

Actually, aren't n and nn variables inverted here ?

I mean, not in n/nn term but what is compute by daylight_hours() should be nn variable to correspond to eq 34 and 35 given by Allen et al. (1998). n should then be sunshine duration measured.

Not quite sure if it's an issue or if I am missing something in my use of this module, so I am posting this question here...

Thanks in advance for your answers.

Kind regards,
Adrien

[mvremec]

Dear Adrien,
Thank you very much for raising this issue, and welcome to pyet discussions.
Everything you wrote is correct. There was an error in the code.
I did not test/check this function - I actually also was not using it, as I did not have data on sunshine duration.

nn (maximum possible duration of sunshine) is now computed via Eq. 34 as in Allen et al. (1998),
whereas n (duration of sunshine) is now a positional argument (must be provided).
According to example 10 on p. 50 in Allen et al. (1998), I also added a test for this function.
I corrected/added the function and test in the dev branch. If you have time, you can check.
I will then merge it with the master branch.

Thanks again!
Best,
Matevz

[raoulcollenteur]

I created an issue based on this.

[manlaya]

Thank you both Matevz and Raoul !
I checked your modifications Matevz and did a review, just a minor modification in the docstring and that's perfect (see comment in review) ! :)
Kind regards,
Adrien

[raoulcollenteur]

Thanks Adrien for reporting! Keep 'em coming ;-)
Cheers, Raoul