What Is Projection Distortion?

Projection distortion is the change in geometric relationships caused by representing the Earth's curved surface on a flat plane. Depending on the projection and location, mapped area, local shape, distance, direction or scale will differ from the corresponding relationship on the Earth.

Distortion is unavoidable, but it is not unknowable: each projection distributes it according to a mathematical design. The useful questions are therefore which property changes, where the change occurs, how large it is and whether it threatens the purpose of the map or analysis.

Distortion is not one number

Calling one projection “more distorted” than another is incomplete unless the property and extent are specified. An equal-area world map can keep every area ratio correct while visibly stretching coastlines. A conformal world map can retain local angles while enlarging polar regions enormously.

Scale can also differ by direction at one location: the scale along a meridian may not equal the scale along a parallel, and a third direction can behave differently again. A single average score may be useful for comparing general-purpose world projections, but it cannot decide whether a projection is appropriate for a cadastral parcel, an airline route or a map of forest area.

Four effects to examine

Area

Area distortion changes the mapped size of a region relative to its ground area. It can alter the visual weight of polygons and corrupt planar area calculations. Equal-area projections control this effect globally by keeping the areal scale constant.

Angle and local shape

Angular distortion changes the angle at which lines meet. Conformal projections preserve local angles, so an infinitesimally small circle remains circular. Larger shapes can still look different because map scale varies across them.

Distance

Map distance is ground distance multiplied by local scale. On most projections, that scale varies with location and direction. An equidistant projection preserves a defined set of distances—often from one centre or along selected lines—not every possible pairwise distance.

Direction

Some azimuthal projections preserve directions from a centre. Mercator makes rhumb lines, which follow a constant compass bearing, appear straight. Great-circle routes follow a different rule and usually curve on Mercator. “Direction-preserving” must therefore identify both the reference point and the meaning of direction.

Tissot's indicatrix makes distortion visible

Tissot's indicatrix imagines an array of infinitesimally small circles placed on the globe and then projected onto the map. Their transformations show two important effects at once.

If the symbols remain circular but change size, the projection is locally conformal and area varies.
If their areas remain constant but the circles become ellipses, the projection is equal-area and local angles change.
If both size and shape change, neither property is preserved at that location.

The indicatrices are diagnostic symbols, not geographic features. Their exact pattern depends on the projection's parameters and the map's aspect. They are most useful when plotted densely enough to show how distortion changes across the actual study area.

Numerical diagnostics can go further: local scale factors, angular deformation, areal scale and maximum/minimum scale directions can be mapped or sampled at control points.

Extent changes the judgement

Every projection has favourable and unfavourable regions. Web Mercator is indefensible for comparing continent area, yet over a small city the scale change across one viewport may be negligible for ordinary display. UTM controls distortion by giving each six-degree zone a separate central meridian. A national grid can be designed around a country's shape and official accuracy needs.

The area of use is therefore part of the CRS definition. Although software may produce coordinates outside that area, successful computation is not proof that the result meets the projection's design assumptions.

Map scale matters too. Small-scale in cartographic terminology means a large geographic extent shown with little detail; this is where global distortion becomes visually prominent. Large-scale local mapping shows less territory and can use a projection tuned to that region.

Distortion is different from a data or CRS error

Expected projection distortion changes geometry continuously according to known rules. The coordinates can be completely correct and the projection can be behaving exactly as designed while a country looks stretched, a high-latitude region appears enlarged or scale varies across the map. A data error is different: the source coordinate, geometry, boundary or measurement is wrong. A CRS metadata error can place an otherwise valid layer in the wrong location or create a systematic offset.

This distinction matters in practice. Greenland looking enormous in Web Mercator is not evidence that the Greenland polygon is bad data; it is the expected result of that projection's scale behaviour. If one layer is displaced from another, by contrast, investigate the source CRS, datum, axis order and coordinate transformation before blaming ordinary projection distortion. If the layers align but measured areas vary implausibly by latitude, the analytical projection may be the problem. An outdated boundary or poorly surveyed point remains inaccurate in any projection.

Calling every unfamiliar shape a “projection error” confuses the geometry of the map with the quality of the data. It also encourages a common mistake: repeatedly assigning new CRS labels until a layer appears close to the basemap. Assignment changes the interpretation of existing coordinates; it does not correct projection distortion or transform the data.

A practical distortion review

State the operation or claim. Area comparison, visual orientation and precise distance demand different checks.
Map the full extent. Include disconnected features and the outermost latitudes and longitudes.
Identify the protected property. Confirm what the candidate projection preserves and under what conditions.
Sample distortion where it matters. Check scale factors or Tissot indicatrices at the centre, edges and sensitive locations.
Compare an alternative. Recalculate or rerender with another defensible CRS and examine whether the conclusion changes.
Record the method. Keep the CRS and coordinate operation with the result.

A projection is suitable when its distortion is controlled for the actual task—not when its name is familiar, its coordinates use metres or its output looks plausible.

References

Map Projections: A Working Manual. US Geological Survey treatment of scale, distortion and projection properties.
PROJ projection documentation. Technical catalogue of projection methods.