Coordinate systems & projections · 4 min read

Lambert conformal conic, transverse Mercator and oblique Mercator zones

Three conformal projections cover the United States. A zone's shape decides which it uses, and that decides whether scale varies with latitude or easting.

American State Plane zones use one of three conformal projections, and the choice is dictated by the shape of the region being covered. Knowing which one your zone uses tells you immediately what your scale factor depends on — and therefore how it will change as you move around a project.

That is the practical payoff. In a Lambert zone, driving east or west changes nothing about your scale factor. In a transverse Mercator zone, driving north or south changes nothing. Getting that backwards is a common and avoidable mistake.

Lambert conformal conic

Picture a cone set over the ellipsoid, cutting through it along two parallels of latitude — the standard parallels. Where the cone cuts the ellipsoid, the two surfaces coincide and the scale factor is exactly 1. Between the standard parallels the cone lies inside the ellipsoid, so grid distances are shorter than ellipsoid distances and the scale factor is less than 1. Outside them the cone lies outside the ellipsoid and the scale factor exceeds 1.

Unroll the cone and you have a flat grid on which parallels of latitude are concentric circular arcs and meridians are straight lines radiating from the cone apex. Scale is constant along any parallel, which is what makes the projection indefinitely extensible east and west.

Standard parallel
A parallel of latitude along which the projection surface intersects the ellipsoid, so the grid scale factor there is exactly 1.0. A secant Lambert zone has two, positioned to balance the scale error across the zone.

The standard parallels are placed so the scale error is roughly balanced: slightly too small in the middle band, slightly too large at the north and south edges. Zone widths in latitude are chosen so that the worst case stays inside the design budget.

Transverse Mercator

Take a cylinder, lay it on its side so its axis lies in the equatorial plane, and shrink it slightly so it cuts the ellipsoid along two lines running north and south either side of a central meridian. That is a secant transverse Mercator. Along the two lines of intersection the scale factor is 1. On the central meridian, which lies inside the ellipsoid, the scale factor is less than 1 — typically 0.9999 in State Plane, or 1 part in 10 000. Beyond the lines of intersection it exceeds 1.

Scale here is a function of distance from the central meridian, which is to say a function of easting. It does not depend on latitude, so the projection extends indefinitely north and south — ideal for a tall, narrow state.

Oblique Mercator

Some regions are long along neither a meridian nor a parallel but along a diagonal. Covering them with either standard projection wastes most of the distortion budget on empty space. The oblique Mercator solves this by tilting the cylinder so its line of tangency or secancy follows the skewed axis of the region.

The classic American application is the Alaska panhandle, where the coastline runs strongly northwest to southeast. Oblique Mercator zones are less common than the other two and their parameters are less familiar, which makes it especially important to record the full definition when you use one.

Side by side

The three projections compared
PropertyLambert conformal conicTransverse MercatorOblique Mercator
Developable surfaceConeCylinder, axis in the equatorial planeCylinder, tilted to a skew axis
Lines of true scaleTwo standard parallelsTwo lines parallel to the central meridianTwo lines parallel to the skew axis
Scale factor varies withLatitudeEasting (distance from the central meridian)Distance from the skew axis
Zone extends freelyEast and westNorth and southAlong the skew axis
Typical zone-defining parameterTwo standard parallelsCentral meridian scale factor, often 0.9999Azimuth of the skew axis and a scale factor
Meridians on the gridStraight lines converging to the cone apexCurved except the central meridianCurved

Why UTM is transverse Mercator and State Plane often is not

UTM covers the whole planet in 60 identical strips running pole to pole, so a transverse Mercator with a fixed central-meridian scale factor of 0.9996 is the obvious choice: one formula, sixty parameter sets, global coverage. It accepts a much larger distortion budget in exchange for that uniformity.

State Plane optimises the other way. It accepts an irregular patchwork of zones, projections and parameters in exchange for keeping distortion roughly ten times smaller. This is the fundamental trade between the two systems and it explains most of their differences in practice.

What to take away

  • In a Lambert zone, your scale factor is a function of latitude. Move east or west and it does not change.
  • In a transverse Mercator zone, your scale factor is a function of easting. Move north or south and it does not change.
  • All three are conformal, so angles are preserved locally in every case and only distances need correction.
  • The scale factor equals 1 exactly on the lines where the projection surface cuts the ellipsoid, and departs from 1 in both directions away from them.
  • An oblique Mercator zone needs its full parameter set recorded, because nobody will guess it.
  • Whatever the projection, grid distance is still not ground distance until the elevation factor has also been applied.

Questions

How do I tell which projection my zone uses?

The published zone definition states it. As a rule of thumb, zones stacked north to south across a wide state are Lambert conformal conic, zones side by side across a tall state are transverse Mercator, and the Alaska panhandle uses oblique Mercator.

Does the scale factor in a Lambert zone change as I move east?

No. In a Lambert conformal conic zone the scale factor depends only on latitude, so travelling due east or west leaves it unchanged. In a transverse Mercator zone the opposite is true — scale depends only on distance from the central meridian.

Why is the central meridian scale factor less than 1?

Because the projection surface is secant rather than tangent: it cuts through the ellipsoid rather than touching it. That places the central meridian inside the ellipsoid, where grid distances come out short, and it halves the worst-case error compared with a tangent design.

Are all three projections conformal?

Yes. Lambert conformal conic, transverse Mercator and oblique Mercator are all conformal, meaning scale at a point is the same in every direction and angles are preserved locally. That is precisely why surveying uses them and not equal-area projections.

Sources

  • NOAA / National Geodetic Survey — Projection definitions and parameters for State Plane Coordinate System zones, including oblique Mercator zones in Alaska.

Work it out