In this article, I reflect on my 2008 project at the Renaissance Computing Institute at North Carolina State University, where I focused on visualizing a data simulation of a Perfect Storm. This scientific visualization aimed to depict data relevant to weather forecasting and was conducted in collaboration with the United States Department of Energy (US DOE). The animation created from this work received an award from the US DOE and is now publicly accessible. A related Youtube video is linked at the end of this article.

The Hurricane Visualization Project

In climatology, meteorologists utilize computational models to analyze weather dynamics, including hurricane formation. For this project, we simulated a perfect storm using data from Hurricane Katrina (2005) on a supercomputer, employing the Weather Research Forecasting model. The resulting data was visualized using VisIt, an open-source tool designed for analyzing large-scale simulations, developed by the US DOE's Advanced Simulation and Computing Initiative. My role involved enhancing the time series animation that illustrated the hurricane's development over time.

Creating the Time Series Animation

To construct the Perfect Storm animation, we merged multiple data sets in VisIt to achieve a realistic hurricane visual. Each data set generated its own three-dimensional isosurface with a distinct colormap. By rendering each time step and assembling them, we created a timelapse of the hurricane's progression. Meteorologists requested the use of wind vectors to display variations in wind direction, which were shown alongside the evolving 3D isosurfaces. Initially, our visualizations were presented in black and white to better assess how different elements combined to create the hurricane effect.

Designing the Visualization's Color Palette

To begin colorizing the Perfect Storm visualization, I established a foundational color context for the animation. Given that hurricanes originate from the ocean, a blue backdrop was a fitting choice, allowing the 3D isosurfaces to emerge effectively. As illustrated in the black and white image, a dark reference map helps highlight the Gulf of Mexico as the storm intensifies. I selected a deep blue for this background.

Next, I needed to devise a color scheme for the 3D isosurfaces and wind vectors. The hurricane's eye is formed by merging two data sets, while the wind vectors remain distinct. The visualization required two blending colors for the isosurfaces and a contrasting color for the wind vectors.

Utilizing the ColorBrewer tool, I opted for a four-color diverging scheme of purple and orange, where purple represents the isosurfaces and orange denotes the wind vectors. I also checked ColorBrewer's color deficiency feature to ensure accessibility. Thankfully, the chosen scheme passed this test. The configuration is presented below, with map border lines omitted and the color deficiency icon indicated.

Applying Color Theory to Isosurfaces

In VisIT, two hurricane data sets evolve over time to define the hurricane's center. Each point that contributes to the isosurfaces can be visualized as a dot. In color theory, the technique of pointillism—pioneered by Impressionist artists in the late 19th century—can be applied here.

By employing pointillism, magenta and blue-cyan dots placed closely together can create the visual effect of a purple isosurface. Although this blending function isn't available in ColorBrewer, the concept can be illustrated through it. Below, you can see how sequential magenta and blue-cyan colors can combine to form a purple gradient.

To grasp the sequential, divergent, and categorical color selection concepts in ColorBrewer, refer to Michael Yi's insightful discussion on "How to Choose Colors for Your Data Visualization" (October 24, 2019).

For the visualization, I selected a magenta sequential color scheme for one data set and a cyan-blue scheme for the other. When these data sets merge, the animation reveals the purple eye of the hurricane.

Colorizing the Perfect Storm Visualization

Below is a still image from the Perfect Storm animation created in VisIT. The combination of magenta and cyan-blue data sets produces a simulated purple hurricane eye, with orange wind vectors contrasting against the purple hurricane and blue ocean backdrop. This contrast allows viewers to track the wind direction changes as the hurricane's eye develops.

The full time series movie, titled “WRF (Weather Research Forecasting) Simulation of Hurricane Katrina,” can be found on Youtube.

Final Thoughts

There are diverse methods for developing color schemes for data visualizations. In this piece, I highlighted the ColorBrewer tool, which I used for the 2008 “WRF Simulation of Hurricane Katrina.” Below, I present a flow chart outlining my entire colorization process.

Additionally, alternative methods for colorizing this hurricane data set are discussed in Chapter 6 of my 2016 book, "Applying Color Theory to Digital Media and Visualization," published by CRC Press. You may discover various approaches depending on the color tools you utilize in your visualization practice.

About the Author:

Theresa-Marie Rhyne is a Visualization Consultant with extensive expertise in producing and colorizing digital media and visualizations. She has worked with the Stanford University Visualization Group on a Color Suggestion Prototype System, the University of California at Davis's Center for Visualization, and the Scientific Computing and Imaging Institute at the University of Utah, focusing on the application of color theory to Ensemble Data Visualization. Previously, she established two visualization centers: the United States Environmental Protection Agency’s Scientific Visualization Center and the Center for Visualization and Analytics at North Carolina State University.