Viametris backpacks and road scanners capture GNSS data, allowing users to easily geo-reference any scan. Having georeferenced data opens up a world of possibilities for what can be done with it, but how to know which projection is the right one for the data and applications?

The map of the world that is shown in school to learn Geography, as an accurate representation of the Earth, is only one way of representing the planet (usually in the Mercator projection).

The Earth is not a flat surface. Its shape can be defined as an ellipsoid that approximates a sphere. This is why the most accurate way of representing it is by using a globe. However, making and carrying a globe is not practical and poses limitations. For this reason, people invented maps and geographic projections.

What are geographic projections?

Projections are a way of representing the spherical Earth as a flat surface. This might sound like a simple task but it poses many challenges and there are many approaches to this problem. Let’s imagine a fully inflated beach ball representing the Earth. If the ball is cut in an effort to have its whole surface lying flat on a table, one ends up with some parts of it full of wrinkles and other parts stretched. This is the reason why it is impossible to represent the Earth’s surface on a flat map.

This is the problem projections are trying to solve. Let’s imagine a translucent globe with a light source inside of it. The whole surface of the earth would be projected outside of the globe. By placing a structure (cone, cylinder, plane) around or tangent to the globe, it is possible to capture this projected image into a flat surface, as can be seen in the image below.

Types of geographic projection

There are different ways of classifying map projections. In this article we will talk about two of them:

Classification based on the tangent surface

As mentioned in the previous section, a projection can be obtained by placing a surface tangent to the earth’s ellipsoid. Based on the way this is done, there are three main categories:

Planar

These projections are also known as Azimuthal or Zenithal projections and they project map data into a flat surface that is tangent to the globe at only one point. In most cases, it is used to map polar regions, so this contact point corresponds to the North or South pole.

Conic

Conic projections map data into a cone placed over a globe. The cone is tangent to the globe in a single latitude line, called the standard parallel. Distortions increase north and south of this line, so they are usually applied to only portions of a hemisphere.

Cylindrical

Cylindrical projections place a cylinder around the globe to project the data onto it. The cylinder can touch the globe along a parallel (latitude) or a meridian (longitude).

Classification based on preservation of a metric property

All projections introduce some type of distortions because they are trying to represent the globe in two dimensions. There are four metric properties that can be affected by projections: area, shape, direction, and distance. These properties are handled differently by each projection and at least one of them is preserved but not all can be preserved at the same time. Based on this, four main categories can be defined:

Equal-area

Equal-area projections present any area on the map in true proportions to its real area on Earth. Generally, this leads to distortions in shapes.

Conformal

Conformal projections preserve angles and shapes. With these projections, all angles measured from a point are correct. This comes at the cost of distortions in area measurements.

Equidistant

In equidistant projections, the length of a particular line in the map must correspond exactly to the length of the lines on the curved surface (the ellipsoid).

Azimuthal

Azimuthal projections ensure that all directions from a single point to any other point on the map are true. This point is where the plane is tangent to the ellipsoid.

How to choose a projection

We’ve talked about many different types of projections, but how to know exactly which one to use? To answer this, two aspects must be considered:

1. What is the purpose of the map?
2. What is the area of interest (AOI)?

The answer to the first question will determine the metric attributes to prioritize. For example, to create a dot density map that illustrates quantities in relation to an area (e.g., population density), an equal-area projection is adequate to avoid misrepresenting the data.

Then, the area to map must be determined. An easy rule can be to choose based on the latitude of the AOI: tropical areas (near the equator) can be mapped using cylindric projections, middle latitudes can use conic projections and polar regions can use planar projections. Nowadays, advances in computation make it easy to model different types of projection to any specific area. The important aspect to consider is to choose one that will present the less amount of distortions for the AOI.

In summary

Projections are a way to represent the Earth’s spheroid as a two-dimensional map. All map projections sacrifice some metric attributes in order to do this. However, as long as the purpose of the map is clear, it is possible to minimize these distortions and present the data accurately. Viametris systems and software allow to easily work with any projection, to be chosen wisely then!