Criteria 3 in BrightSun2020CSDs.m

[cwhanse]

The high zenith limit for the line length criterion (c3) is 0.5 in the code, but in the JRSE paper it is given as 2. Which is recommended?

csd-library/models/BrightSun2020CSDs.m

Line 185 in f2d81b6

BrightSunCriteria.ghi_c3_lower_line_length=-0.5;

[JamieMBright]

@cwhanse much with the other issue, I cannot fully investigate this in the lockdown. One of the reviewers of my paper suggested that 2 was very high so I played around a lot with different values. It would be unfortunate if I made the suggested change but left a typo in the paper, or vice versa...

[cwhanse]

@JamieMBright All's fine. I am porting from MATLAB to python to add this technique to pvlib, we'll make our best guess for now and can certainly adjust when you are able to resolve the questions.

[JamieMBright]

@cwhanse that's great! let me know if I can help in any way.

[cwhanse]

@JamieMBright any chance you've been able to look into this issue further?

[JamieMBright]

@cwhanse I am still not allowed back to Singapore at the moment. My main opinion is that the coded version is correct and the the paper is incorrect.
This is mainly because I only published the paper once I was happy with the code, and I did revisit some coefficients between revisions, I think I just did not update the paper correctly. This makes significanltly more sense than the alternative.

[cwhanse]

Thanks, that's what we'll assume and move ahead. Best wishes to you.

[JamieMBright]

Thanks @cwhanse. Once it is on PVlib, undoubtely with the contributors, all errors will out. Perhpas just a tentative comment next to the criteria saying something like "#value of -2 in the publication".
When I get back to normal working, I can check it out at all BSRN stations again, and when it is on PVlib, I will set some students to investigating.
I appreciate the effort you hvae put in to doing this.
Best,
Jamie