improve reservoir schematization in wflow

[shartgring]

I have made Wflow branch (res-lake-sq) where SQ approach is implemented. This branch will still work with previous Wflow models. It checks if a SQ-table exists (e.g. res_sq_1.csv) and if this is the case, it will apply the SQ-approach. And otherwise, it will use the standard approach with the fractions.

This means that this branch can be used in this project. Implementation of the res-lake-sq branch in the main code of Wflow will take time as there are plans to merge reservoirs and lakes in wflow, and this branch will be used to reuse the code. So in the meantime, it can be an option to test the reservoirs in the model by using the res-lake-sq branch.

The shared Rur-model will also be based on use with this branch

[laurenebouaziz]

thanks @shartgring ! this is indeed a very good quick fix! so if I understand correctly, we don't even need to add a layer ReservoirOutflowFunc in the staticmaps, adding a csv file where required is sufficient. and max volume of the reservoir is used from the staticmaps?

[shartgring]

Yes exactly! It may even be possible to extend the code to get the max volume from the .csv file, but for now it uses the ResMaxVolume netcdf map. It uses resloc to find the .csv file, so also good to know that reservoirs should have unique ID's (which of course is good practice)

Also, code passed all tests after running runtests.jl except the following (good to keep in mind):

• Stale dependencies (1 fail)
• Compat bounds (1 fail)
• BMI -> BMI functions -> model information functions (2 fails)