Somewhere Over the Rainbow: How to Make Effective Use of Colors in Scientific Visualizations
One of the many challenges associated with (atmospheric) sciences is the analysis and utilization of large, usually very complex datasets. One common way to process such data, to gather and communicate the information, is to graphically visualize them. Such visualizations cover a large range from simple one-dimensional plots (e.g., bar plots or time series plots) to complex multidimensional charts (e.g., spatial three or four dimensional plots).
Color is often applied to improve the readability and to enhance the information content or level of detail. Nowadays, most common software packages supply a variety of different plot types frequently using a red-green-blue (RGB) rainbow palette as default color map. Even though the adequate usage of color maps for scientific visualizations has improved over the last years, they are still often used unconditionally without questioning their validity. Applying ineffective color maps can make plots hard to interpret or even lead to (severe) misinterpretations.
In the scope of this seminar an alternative color space called Hue-Chroma-Luminance (HCL) is introduced. In contrast to the (in)famous RGB color space the HCL color space is based on the human perception rather than a technical demand. Beside some more technical details about the HCL color space some easy-to-use tools (R package `colorspace’; www.hclwizard.org) are introduced. These tools allow to integrate custom-tailored HCL based color maps in your daily workflow. Chosing effective color maps requires only very little additional effort but can clearly help to improve the assesability and readability of your results and findings.