When I started with UTM tiles for Trainz back then we already had two other similar concepts of texture carrier objects. One was called "Basemaps", the other, if I remember correctly, "UTM maps". The former focussed on the British Ordnance Survey map grid, the latter on UTM, of course. Map projections for large scale almost always use a metric grid and typically, you have 1000m grid lines on these maps.
Now, the natural thing was to use these 1000m grid lines as easily identifiable dividers for the tile objects. Those original tile concepts were meant for a manual approach, though. You had to cut out that part of the map you wished to use, create a texture from it, apply that to the tile object and finally place the tile object into your route, all done by hand. You may now argue that with the automatic approach for UTM tiles in TransDEM, the 1000 m grid lines on maps are no longer that important, or for ortho-imagery, where there are no grid lines at all. While that is correct, there is another aspect to it: human-readable tile names. Although not actually required for processing the tiles in TransDEM or Trainz, the UTM based naming convention allows us to locate any tile for any route anywhere in the world by just reading the name.
Compare the UTM tile names with the new baseboard-based file name structure in upcoming T:ANE SP2. The baseboard files will be named by baseboard index, relative to the route origin and valid only for one particular route. As route origins can change when merging routes, the baseboard file names are rather volatile, too. Nevertheless, they make sense for automatic processing and Driver and Surveyor. UTM tiles, on the other hand, may be created for a specific route, but basically, they are independent of the route.
So, in essence, UTM tiles in TransDEM reflect the geo data source grid, the UTM map projection, with human readable absolute names. And that only makes sense with a grid width of 1000m.
|