Currently, the datasetdocgen.py script produces maps with a grid and longitude/latitude labels, but no scale bar. In the discussion on https://github.com/deepesdl/deepesdl-doc/issues/3, @AliceBalfanz suggested adding a scale bar. I think this is a good idea but it's not entirely trivial to implement: currently the map uses a Mercator projection, so the horizontal scale varies significantly with latitude and we therefore can't produce a scale bar which will apply to a whole plot. If we're going to have a scale bar we should instead use projections which minimize distance error -- maybe a locally centred Lambert azimuthal equal-area for local/regional data sets, and Mollweide for any bounding box big enough to be shown on a map of the whole Earth. Cartopy supports both of these. Dropping Mercator is probably a good idea in any case because of the notorious vertical distortion at high latitudes.
Currently, the
datasetdocgen.py
script produces maps with a grid and longitude/latitude labels, but no scale bar. In the discussion on https://github.com/deepesdl/deepesdl-doc/issues/3, @AliceBalfanz suggested adding a scale bar. I think this is a good idea but it's not entirely trivial to implement: currently the map uses a Mercator projection, so the horizontal scale varies significantly with latitude and we therefore can't produce a scale bar which will apply to a whole plot. If we're going to have a scale bar we should instead use projections which minimize distance error -- maybe a locally centred Lambert azimuthal equal-area for local/regional data sets, and Mollweide for any bounding box big enough to be shown on a map of the whole Earth. Cartopy supports both of these. Dropping Mercator is probably a good idea in any case because of the notorious vertical distortion at high latitudes.