Feature: TC surge from bathtub model

[tovogt]

Due to problems with the elevation package under Windows (issue #9), the old surge module has been removed from the main and develop branches and archived under archive/feature/tcsurge. A completely new TC surge module is in the pipeline in the feature/tc_surge_geoclaw branch. However, due to it's heavy computational needs, there is still some interest in the old simplified approach, most recently by @QinhanZhu and @arunranain who managed to remove the dependency on the elevation package from the old module.

I added some references from the literature, implemented unit tests and modified the tutorial notebook. As discussed in #9, users are now required to provide their own DEM data set as a raster file. Two freely available data sets are mentioned in the tutorial notebook. I think, the module is now ready to be merged back into the develop branch.

Naming: I called the old model a "bathtub" model to make clear that it's a very simplified approach. The term "bathtub model" is often used in the storm surge literature to refer to a model that ignores all hydrodynamics and instead basically subtracts a digital elevation model from an assumed surge height.

@chahank: Please feel free to assign more/other reviewers.

[chahank]

Were these changes discussed with @QinhanZhu and @arunranain ?

[Evelyn-M]

Thanks a lot for the work, I just have a few, mostly conceptual / readability comments.
Nothing to remark on the jupyter notebook, looks good!

[tovogt]

Thanks for the valuable feedback @Evelyn-M! I will comment directly on each of your points.

Were these changes discussed with @QinhanZhu and @arunranain ?

They gave me access to their modified version and communicated that they don't plan to continue working on a transformation into an update for the CLIMADA core.

[chahank]

Thanks for the valuable feedback @Evelyn-M! I will comment directly on each of your points.

Were these changes discussed with @QinhanZhu and @arunranain ?

They gave me access to their modified version and communicated that they don't plan to continue working on a transformation into an update for the CLIMADA core.

Is the new implementation compatible with the way they used the TC_surge module for the Vietnam study? Because they are also modifying the TC_surge module at the moment to make it possible to use any DEM model. The crux is the clipping part. I just want to be sure that not two competing version of the same module are developed here.

[tovogt]

Is the new implementation compatible with the way they used the TC_surge module for the Vietnam study? Because they are also modifying the TC_surge module at the moment to make it possible to use any DEM model. The crux is the clipping part. I just want to be sure that not two competing version of the same module are developed here.

As I said, they gave me access to their code and no, they are not modifying it at the moment - they actually stopped working on it. If I'm not mistaken, my PR is 100% compatible with what they did and it's supposed to reproduce the exact behavior of the legacy tc_surge module with a somewhat polished API and improved implementation. The API changes I introduced only affect parts of the API that are not used for that Vietnam study (as I said, I have access to their code - I checked this).

[chahank]

OK great. Just wanted to be sure. But could you please double check with @QinhanZhu as I talked to him recently and (probably I misunderstood) he seemed to imply to wanting to change something about the clipping of DEM in TC_surge.

[tovogt]

Thanks @emanuel-schmid, I addressed all of your points in my latest commit.