Processing images with wavelengths outside the visible spectrum can produce spectacular results – like this iconic image of the so-called Cosmic Cliffs by the James Webb Space Telescope. The gas and dust in the field of view surround NGC 3324 near the Carina Nebula. The filters used in this image range from 0.9 microns to 4.7 microns; all are infrared filters, which are outside the range of human vision and block out about 0.7 microns. Image credit: NASA, ESA, CSA, STScI
In terms of appearance, the universe is a tricky place. Visible light occupies only a small part of the electromagnetic (EM) spectrum. To explore the cosmos in its entirety, scientists must look beyond visible light with special instruments such as radio telescopes and X-ray telescopes. And the James Webb Space Telescope (JWST) detects infrared (IR) radiation, penetrating veils of dust that block visible light.
The researchers and developers who process this data face a special challenge: What color is light that humans cannot see? Answering this question and bringing these cosmic scenes to life requires an impressive mix of art and science.
The process of decoding data from non-optical telescopes is often called false colorization, but the word “false” doesn’t do it justice. The technique has been used for decades to create color photos from raw data, including some of Hubble’s most famous images, which combine the telescope’s optical capabilities with its ultraviolet (UV) and IR data. Like any other camera, this processing creates images from otherwise meaningless strings of zeros and ones.
“Representational colors is a more accurate term,” says Joe DePasquale of the Space Telescope Science Institute (STScI) in Baltimore, Maryland, which manages JWST and its images. As one of only two space visualization developers at STScI, DePasquale is eager to dispel any public skepticism generated by the term “false colors.”
“It’s a fine blend of art and science that leans more towards science,” he explains. “We work with scientists to make aesthetic judgements and highlight important scientific features without altering the data. It’s about showing what the telescope observes as best as possible.”
Nothing wrong with wrong color
The process of representational colorization involves assigning colors based on relationships that mimic the human perception of visible light. Since JWST detects light in the near and mid-infrared, this range of wavelengths must be shifted—or mapped, as image processors call it—into the visible color space.
“The color mapping is pretty well established,” says Alyssa Pagan, the other space visualization developer at STScI. Typically, this remapping follows the same order as the visible part of the spectrum, “with the shortest wavelengths getting bluer and the longer wavelengths getting redder,” she explains. “But there’s a lot of flexibility. We can make the middle wavelengths a little more cyan or green to get a fuller color representation.”
JWST’s different filters – 29 in the mid-infrared and 10 in the near-infrared – reveal different aspects of an object’s structure. A nebula might have filaments of dust overlapping with clouds of hot gas, each of these filaments accentuated by different wavelengths. Adjustments are made to highlight these features.
“You do your best to either subtract or enhance,” says Pagan. “Even though scientists use six filters to get an image, you don’t necessarily want to use all of them. That can neutralize overlapping features, like mixing too much color and turning brown. You have to be careful, but you want to get the widest color gamut possible.”
Covering the spectrum
In addition to the JWST, astronomers use a wide range of observatories to observe the rest of the electromagnetic spectrum. From low-energy radio waves to microwaves, infrared, UV, X-rays and finally gamma rays, these telescopes reveal the unimaginable, including the cosmic microwave background radiation left over from the Big Bang, supermassive black holes, the inner structures of nebulae and exploding supernovas. (One key observatory in this fleet, the Chandra X-ray Observatory, was recently cleared for decommissioning by a NASA committee because no replacement was in sight – to the dismay and protest of scientists.)
I
Although the same basic rules apply to interpreting the data from these different telescopes, combining the data across the entire electromagnetic spectrum is challenging. However, the results – so-called multi-wavelength images – can be spectacular and insightful.
Pagan compares the process to trying to combine a regular photograph of a human arm with X-rays of the bones in it. “They don’t match very well, but you can assign one color to your X-ray and another to your skin to see what’s really happening. It’s like creating a graph with five different colors to show different numerical information.”
Visualizations for the public
In spring 2022, ahead of the release of the first science images from JWST, STScI gave itself a short window of opportunity to select and map some of the most spectacular targets the telescope had imaged to date. “The team met daily for six weeks,” says DePasquale. “We had scientists, writers, graphic designers, and then Alisa and me.”
The resulting images immediately achieved cult status. Pagan worked on the Cosmic Cliffs – the nickname the team gave to a surging wall of gas and dust that surrounds the star cluster NGC 3324 near the Carina Nebula. DePasquale’s work included the Tarantula Nebula (30 Doradus), a huge star-forming site in the Large Magellanic Cloud.
DePasquale says he doesn’t take the task of colorization – or its impact – for granted. “It’s a big job to interpret what we get from a telescope and create the images that shape the public’s perception of the universe.”
If you would like to try your hand at representative coloration, STScI provides JWST data and software at the mission websiteFor more information on processing your own JWST images, see Warren Keller’s story in our September issue.
