search

UMEP-dev / SUEWS

Surface Urban Energy and Water Balance Scheme
https://suews.readthedocs.io/
Mozilla Public License 2.0
12 stars 8 forks source link

Which supy version for citation? #246

Closed feigenwinter closed 3 months ago

feigenwinter commented 7 months ago

Hi all, we had two master thesis started in JAN 2023 that used the supy version 2022.9.22 which was the standard version installed by "pip install supy" at this time. Later we recognized that this version was declared as "yanked" and the current stable version is 2022.4.7. (But the thesis were finished at this time). When changing from 2022.9.22 to 2022.4.7 we had to change input parameters to receive similar (but not exactly the same) results as with 2022.9.22. This refers mainly to the input parameter df_state_init.g6 (the soil moisture deficit function) when using df_state_init.gsmodel = 4. (changing from 0.36 (Havu et al., 2022) to 0.05 or so...) This affects QE and FcPhoto.

In #241 sun05 wrote: "However, it is noted a relatively old supy is being used - please do the following: update to a recent dev version: pip install supy==2023.7.3.dev0" Since citing a "yanked" version in a publication is not appropriate, I wonder if citing a "dev"-version is really better? All this said, we wonder if it's really worth to install and work with a "dev" version, if we can not use the same input parameter values?

sunt05 commented 7 months ago

Hi @feigenwinter, apologies for the inconvenience caused by these changes. We had to yank v2022.9.22 due to an error in calculating the frontal area index, which becomes significant when tree coverage is large.

Since the formal release of 2022.4.7, we've been working on numerous under-the-hood improvements and experimental features that are yet to be finalised; hence, please stick with 2022.4.7 for stability.

sunt05 commented 3 months ago

Hopefully, my answer above addresses this issue, so I'll close it for now. Otherwise, feel free to re-open and provide more details to help resolve it.