Add preprocessed grids for certain fetch calls

[mdtanker]

Description of the desired feature:

Currently, all the fetch calls automatically return the full resolution of the grids. For example, DeepBedMap's initial spacing is 250m. Any spacing supplied by the user results in a pygmt.grdfilter and grdsample, which takes a long time for these large datasets. We should include low-resolution versions of each of these which are automatically returned unless the user specifies a higher res.

For now, these datasets include ice_vel, bedmachine, bedmap2, deepbedmap.

See the low and high res preprocessing versions of fetch.ice_vel() for how to implement this.

I think 5km would be a good spacing for the low res versions. Any added datasets with original grid spacing smaller than 5km should include a low-res version.

Are you willing to help implement and maintain this feature?

Yes