Add functionality for spherical system

[alarshi]

This PR addresses one of the bullet points in issue #5626. In particular, it allows the user to use spherical coordinate system while using initial_topography function model.

Before your first pull request:

• I have read the guidelines in our CONTRIBUTING.md document.

For all pull requests:

• I have followed the instructions for indenting my code.

[bangerth]

Do you want to say

Suggested change

	ExcMessage("While using a box geometry, use cartesian coordinate system."));
	ExcMessage("While using a box geometry, you can only use the cartesian coordinate system."));

?

[bangerth]

Suggested change

	Assert (coordinate_system == Utilities::Coordinates::CoordinateSystem::spherical,
	ExcMessage("Make sure that you are using a spherical coordinate system While using a spherical geometry."));
	Assert (coordinate_system == Utilities::Coordinates::CoordinateSystem::spherical,
	ExcMessage("Make sure that you are using a spherical coordinate system while using a spherical geometry."));

[bangerth]

Move the assertion to the top of the code block. Everything in this block is wrong if the assertion is not satisfied, we might as well abort there.

[gassmoeller]

@alarshi: This is actually not quite the implementation I had in mind in #5626. I was thinking we should allow users to choose which coordinate system to use (as in boundary_temperature/function.cc), not force them to use the coordinate system corresponding to their chosen geometry model. It is possible that that is not the right approach for initial_topography, but can you think about it and let me know what you think is most useful?

[alarshi]

@gassmoeller : I see. I can think of cases where a user might want to use the initial topography from another model run (fastscape?) as an input to ASPECT models. I agree that having a choice between cartesian/spherical would be useful.
I will add that, thanks for the suggestion!

[alarshi]

@gassmoeller : Hmm, I see why the function in initial_topography was implemented based on the geometry model we are using. Unlike, the boundary_temperature and heating_model implementations in which we compute the boundary values at a point, initial_topography computes boundary values for a surface_point. This implies that in order to use the Function Parser, we need convert the surface_point to a point and this will depend on the geometry model.
Let me know if you have other thoughts or I am misunderstanding what you said.