Confusion in regards to projection

[ARiisgaard]

Hey Carsten

I have been visualising my data using the EPSG:4326 projection. I originally did so because both the data and the code, which I build my solution on was using it. I chose not to have a scale on my map, since this projection is having degree have measurement, which leads to a scale with the same.

However as I start to write about it I'm getting a bit more confused.

Borden D. Dent mentions in "Cartography - Thematic Map Design" that:

Equivalent projections are very important for general quantitative thematic map work, particularly at a global scale. It is usually desirable to retain area properties, particularly when enumeration units are compared or if area is part of the data being mapped (see Chapter 4); population density is a prime example

However I can not find any information about if the EPSG:4326 projection is an equivalent projection? I since I can't find anything say that it is, I get the feeling that it probably is not.

I am therefore struggling a bit with an argument for the chosen projection - or to be honest, Im not really sure if the projection shouldn't be a different one

[crstn]

4326 is the EPSG code for lat/lon coordinates, using the WGS84 ellipsoid. It is the horizontal component of the system used by GPS. Because these coordinates describe positions on an ellipsoid, putting them onto a plane (lake a paper map or a computer screen) will always create distortions, which are getting larger towards the poles (where the graticules converge, but appear parallel when projected onto paper). This (non-scientific) video explains that quite well: https://www.youtube.com/watch?v=kIID5FDi2JQ

And equivalent project takes those coordinates and projects them onto the plane while maintaining area, i.e. the area in the projected map is the same as on the original ellipsoid (which is still just a simplified model of the globe, but it is generally quite close). Distances and/or directions will be wrong on equivalent maps, though, since it is impossible to project an ellipsoid onto a plane without introducing any kind of distortion.

Does that make sense? Other datasets with similar information, such as the Global Human Settlements Layer, use the Mollweide projection. In any case I think it is too late now to change the projection of the whole project.

[ARiisgaard]

It does make sense and I know that it is too late for changing the projection now. It was mostly to have the knowledge before the exam if something should have been done differently.

So if you were to create a tool such as mine, you would be using the Mollweide projection?

[crstn]

It depends ;) At global level, it would definitely be a better choice to visualize the data because of the huge distortions that the Mercator projection introduces (internally, OL needs to somehow put the 3D coordinates on a plane, and does that using Web Mercator. At local level, however, the distortions are pretty consistent, so everything looks OK. If we were to do any analyses with the areas (including densities), however, it would be extremely important to use an equivalent projection.

[ARiisgaard]

Thanks for all of these last second answers

So would these analysis be limited to transformation of the data (creation of a new layer. e.g. male polulation density = density*percentage of males) or would it also be relevant for describing the currently displayed data (average density in the current map extent or a histogram based on the current extent)