Ouranosinc / xclim / 17269752551

92%

push

github

Snowfall and rain approximation from Dai 2008 (#2208)

<!--Please ensure the PR fulfills the following requirements! -->
<!-- If this is your first PR, make sure to add your details to the
AUTHORS.rst! -->
### Pull Request Checklist:
- [x] This PR addresses an already opened issue (for bug fixes /
features)
    - This PR fixes #1752
- [x] Tests for the changes have been added (for bug fixes / features)
- [x] (If applicable) Documentation has been added / updated (for bug
fixes / features)
- [x] CHANGELOG.rst has been updated (with summary of main changes)
- [x] Link to issue (:issue:`number`) and pull request (:pull:`number`)
has been added

### What kind of change does this PR introduce?

* New methods "dai_annual" and "dai_seasonal" for snowfall and rain
approximation.
* New argument `landmask` to control which formulation (land or ocean)
we use.

### Does this PR introduce a breaking change?
No. Only new arguments added, no default changed.


### Other information:
@Zeitsperre : The `landmask` argument is polymorphic (my favorite, your
nemesis (?)), it can either be a DataArray (which is assumed to be of
boolean dtype, but no need to check) or a straight boolean. This is not
a `Quantified` because boolean have no units. Right now, I put the
annotation that resolved into the "optional variable" `InputType`, which
means "bool" is not part of it. Should we create a new `InputType` for
boolean mask ? It would be similar to `Quantified` in that it can be a
scalar or a DataArray, but without the units check.


@tlogan2000 : I implemented the rain approximation as well. However, in
opposition to the 3 other methods already implemented, the Dai rain
approximation is not equal to `pr - prsn` because it considers some
small fractions of the total precip to be of another phase (denoted
"sleet" in the article). Thus, to imitate DonnéesClimatiques, ones need
to compute `prra` as `pr - prsn` instead of using `rain_approximation`.
Is this ok ?


Tests need to be ... (continued)

42 of 57 new or added lines in 2 files covered. (73.68%)

7839 of 8512 relevant lines covered (92.09%)

8.07 hits per line