A map projection is a mathematical rule for flattening the curved Earth onto a plane, and a coordinate reference system (CRS) combines a projection with a datum — a model of the Earth’s shape and its anchoring. Every projection distorts something: area, shape, distance or direction. Understanding projections is not academic; mismatched coordinate systems remain the single most common cause of GIS data landing in the wrong place.
Geographic vs projected coordinates
Geographic coordinates — latitude and longitude in degrees — locate points on the ellipsoid itself. They are universal but awkward for measurement, because a degree of longitude shrinks toward the poles. Projected coordinates map those positions onto a flat plane in metres or feet, enabling honest distance and area computation within the projection’s valid zone.
WGS84 latitude/longitude is the lingua franca of GPS and web data exchange, while national grids and UTM zones dominate engineering and mapping work. Web maps almost universally render in Web Mercator, a display convenience that should never be used for area measurement.
Datums and why they matter
A datum ties the coordinate framework to the physical Earth. Different datums — WGS84, ETRS89, NAD83, or older local datums behind many national grids — can place the same monument tens or hundreds of metres apart in coordinates. Vertical datums do the same for heights, separating ellipsoidal height from orthometric height above the geoid.
Datum transformation is therefore a precision operation with multiple published methods of differing accuracy. Production workflows must record exactly which transformation was applied, or centimetre-grade survey data quietly degrades to metre-grade.
Choosing and managing a CRS
The practical rules are simple: know the EPSG code of every dataset, work in a projected CRS appropriate to your area for measurement, and reproject deliberately rather than letting software guess. For countrywide or global work, expect to manage several zones or use equal-area projections for statistics.
Most “my data is 200 metres off” support calls trace to an undefined or wrongly defined CRS — a metadata failure, not a software one. Rigorous CRS bookkeeping is a hallmark of professional spatial data production.