Our 2019 article on GIS colour palettes highlights the potential problems arising from the poor use of colour when creating maps. These include choosing colours that lead to the misinterpretation of data, or that create difficulties for people with colour vision deficiencies (CVD). As these issues have become more widely recognised, approaches and tools have been developed to avoid them.
Such tools help to produce attractive maps whilst accurately conveying information, minimising risk of misinterpretation and are accessible.
However, before considering any of these tools, it is important to ask the following three questions:
What will work best here?
It is important to decide if colour is needed in the first place.
Colour is just one way to distinguish between elements; other visual cues such as size, symbol and labels may work just as well, if not better.
To determine whether colour is the best choice, it is important to consider if there is a lot of information to present, as the use of too many colours can introduce confusion, making it difficult to distinguish between map values.
Grey is often considered the most important colour in data visualisation, as it can be used for background elements. This helps by reducing the number of colours required, reserving other colours for highlighting.
What information is being displayed?
The type of information to be presented in a map will guide the best approach to selecting an appropriate colour palette.
For sequential data (i.e. ordered data such as temperature, elevation or density) the distance along a colour palette should be approximately equal to how the change in colour is perceived. This can be achieved using ready-made colour palettes that have this characteristic, known as perceptual uniformity, e.g. Colorbrewer.
Other examples of these types of colour palettes have been added to GIS software to provide easy access for map makers.
The visualisation of categorical data has similar requirements, but many of the colour palettes used for sequential data are not suitable for this type, as the viewer is likely to interpret order in the data classes.
Instead, the colour palette should make it easy to discriminate between discrete classes. The Colorgorical tool provides a means to select sets of colours that are easily discriminable by selecting colours based on their perceptual distance.
These tools provide straightforward ways to generate colour palettes that are appropriate for various visualisation tasks.
Can everyone understand/read this?
It should be noted that the use of these tools alone does not guarantee that the resulting visualisation will be free of CVD issues, especially if additional colours feature alongside the chosen colour palette.
Fortunately, there are also tools available for testing for CVD problems. Viz Palette is a particularly useful tool that provides a way to combine colour selection with tests relating to colour vision deficiency.
Viz Palette takes the user’s choice of colour palette and simulates various forms of colour vision deficiency, thereby highlighting problematic colour combinations that can subsequently be avoided. This provides an important final check on the choice of colours used in a map or other data visualisation.
At ABPmer we use these questions, tools and GIS features to ensure that all our outputs are accessible.
Prepared by Aidan Walsh, GIS Analyst
Like this? Sign up for our newsletter