Regarding the transport equation onto the sphere

[GuillaumeBriand]

Hello everyone,

I would like to open the question regarding the correction to be made to solve the equation in spherical coordinates (latitude, longitude).
This should be improved mass conservation.
How should we proceed ?

Guillaume

[innasenina]

Hello Guillaume,

That’s great and timely! To reply to your question, please integrate these changes into the development branch and provide related documentation. We can then review the changes together and merge them with the master branch via pull request.

Regarding documentation for the change, since the transition from the regular grid on the 2D plane to the quadrilateral grid on the sphere surface represents a critical change of the model numerical scheme, basically it should be entirely rewritten, it will be necessary to update the reference manual as well. Unless you did this already, I suggest that we do this as the next step, and for now we will need to revise only the essential changes, i.e., the partial derivatives approximation (section 2.2.3 of the manual) in the spherical coordinates, and, ideally, the updated approximation of the model equations (section 2.2.4).

Note that the development branch is behind the master, so before introducing any modifications, please update it from master. Or, perhaps, we can make abother branch specifically for this issue/change. Let me know what you prefer and if you need help with anything.

Cheers,
Inna