Buzz
When designers aim to depict changes in data, such as in heat maps or topographical surveys, they often choose the rainbow scale. It's become almost a standard for visualizing scientific and engineering data. However, placing the entire spectrum of the rainbow into a single image isn't ideal. According to Scientific American, it creates issues for people with colorblindness, rendering the visualization unreadable. Even if someone can distinguish all the colors, understanding the transition from red to violet isn't always intuitive.
Researchers at the U.S. Department of Energy’s Pacific Northwest National Laboratory in Washington have introduced an alternative to the rainbow scale. Their new color model, named cividis, was designed to make visualizations clearer for both people with colorblindness and the general public. This model, based on how the human brain perceives color differences, uses only shades of blue and yellow, the hues that someone with colorblindness would see when viewing a rainbow-colored scale.
The newly developed blue-yellow color scale is named CVD-Jet | Nuñez et al., PLOS ONE (2018)They took traditional rainbow color maps and processed them using software that transformed them into a version more closely resembling the blue-yellow scale, representing what a person with the most common type of colorblindness would see. Then, the software fine-tuned the color and brightness to ensure it aligned better with how people typically interpret data. One key issue with the rainbow scale is that people often perceive the brightest color as a peak, which can lead to misinterpretations. While yellow is in the middle of the Roy G. Biv spectrum, it often stands out as the most extreme color on the map, even though red is actually the highest. In this new color scale, brightness increases with higher values, making it easier to understand.
In general, most people don’t instinctively know the correct order of rainbow colors. While red and violet are at opposite ends of the Roy G. Biv scale, their placement isn’t always obvious. By limiting the range to just two colors, readers can more easily pinpoint where a specific value lies on the scale.
This two-color scale also presents data changes more gradually, while a rainbow scale can make differences appear overly stark. This smooth progression can highlight more subtle variations.
It's not merely about aesthetics. A complex, attention-grabbing rainbow visualization can mislead scientists into misinterpreting their own data. In contrast, a simpler scale allows them to spot patterns more easily. For example, in a 2011 study cited by Scientific American [PDF], researchers from Harvard discovered that doctors were quicker and more accurate at identifying heart disease indicators when examining 2D images of arteries on a color scale with just black and red, compared to 3D rainbow visualizations.
Cividis has already been incorporated into the color-scale libraries of various image-processing software, and its developers are working to encourage more scientists and designers to adopt it moving forward.
