Mercator vs. Web Mercator for Large Scale Topographic Maps

Question

I have 8 terabytes of imagery that is in either lon/lat or UTM projection that I will be using for the basis of large scale topographic maps.

I have been studying the differences between Mercator and Web Mercator, but I am still not sure what projection is more suitable for large scale (scales between 1:24k - 1:63k) topographic maps. I do understand that Mercator is conformal unlike Web Mercator which seems like an advantage.

I would like to project these datasets in UTM as I understand that is the recommended projection by United Nations Cartography Committee for topographic maps because it minimizes angular distortion at large scales (and at detail greater than 1:24k) - and because about half of my data is already in UTM.

At the same time I would like the ability to seamlessly zoom from a large geographic area (~1:250k up to 1:24k. That makes the UTM zones problematic as I will not have the capability of reprojecting adjacent UTM zones into a common UTM zone on-the-fly.

Answer 1

Web Mercator is rarely the right answer, unless you want pictures that line up with other stuff in Web Mercator.

An extract from the NGA's Implementation Practice Web Mercator Map Projection, which is worth a read in full:

5.2 The Web Mercator map projection has several defining mathematical formulas and parameters that make data referenced to Web Mercator incompatible with WGS 84 ellipsoid referenced GEOINT. These incompatibilities include:

5.2.1 Spherical equations. The use of simple spherical mathematics to convert ellipsoid based latitude and longitude data to spherical based plane coordinates.

5.2.2 Non-conformal. The use of a point scale factor that varies as a function of the azimuth and creates angular distortions. This means that Web Mercator is not a conformal projection and that true rhumb lines are not straight lines on a Web Mercator projection.

5.3 The mathematical differences described in 5.2.1 and 5.2.2 can cause errors in spherical based Web Mercator plane coordinates of over 40 km compared to WGS 84 ellipsoid based Mercator plane coordinates.

5.4 Visually, Web Mercator plane coordinates appear to portray a correctly rendered ‘Mercator-style’ map, and numerous layers of Web Mercator data will align on a map or computer screen. The shifts become apparent only when spherical based Web Mercator plane coordinates are compared to WGS 84 ellipsoid based Mercator plane coordinates, e.g. control points, GPS data, and WGS 84 ellipsoid registered maps, data, and GEOINT.

Answer 2

It depends on the extent of your area of interest.

Transverse mercator is conformal along the central meridian, while Mercator (in its original form) is conformal along the center latitude. That does not have to be the aequator.

So if your area is mostly north-south orientated, tmerc is better, if it is more east-west, merc is better.

Transverse is restricted to the visible half of the world, so it is not suitable for the whole of Asia (or Russia).