Hubble’s blurry years

On 24 April 1990, following years of delay, astronomers’ “window on the universe,” the Hubble Space Telescope, was finally launched aboard the space shuttle Discovery. But the telescope’s arrival in orbit wasn’t quite as triumphant as astronomers had hoped. After all the anticipation about its clear view above Earth’s atmospheric haze, Hubble began returning warped, blurry pictures rather than the crisp images scientists had promised the public and themselves.

The iconic telescope’s primary mirror had been ground “nearly perfectly to the wrong specifications,” wrote historian Joseph Tatarewicz, which caused spherical aberration. The mirror’s outer perimeter was flattened by about 2.2 µm, which prevented it from focusing light rays properly. Rather than concentrating 70% of the light from point sources over an angular resolution of just 0.1 arcseconds, as it was designed to do, the mirror instead smeared that light into a halo 10 times as large.

The manufacturing error was a major embarrassment for NASA, one that threatened to, as Tatarewicz put it, “severely degrade or even scuttle” the planned 15-year mission. Scientists and engineers soon got to work on a plan to fix the flawed telescope, with repairs scheduled for 1993. As those rescue plans took shape, astronomers also began to think about what juice they could squeeze from this lemon in the meantime. After all, despite Hubble’s flaws, its position in low-Earth orbit still gave it an advantage over ground-based telescopes. “Very quickly it became apparent that the images were useful,” says Robert O’Dell, Hubble’s project scientist for 10 years prior to launch. “They weren’t perfect, but they were useful.”

Read more special content for Hubble’s 30th anniversary:

The teams that had worked on the telescope’s five instruments had each been promised 300 hours of observation for projects they designed. Hubble higher-ups now gave those teams the choice of continuing with the observations using the aberrated telescope or deferring them until the repairs were made. Around 40% of the projects went forward, recalls O’Dell.

The very center of each image from Hubble’s cameras was well rendered, so compact, bright objects—like active galaxies and groups of stars—showed up well. Astronomers could either avoid diffuse and dim features spread across the field of view or accept their degradation. The spectroscopic instruments worked well, as long as the field of view wasn’t too crowded. And any UV observations would be a boon, since Hubble was one of only a few scopes that could see in that part of the spectrum.

On top of that, scientists didn’t have to simply take what Hubble gave them. Once you know how a flawed telescope distorts an image, “you can use the information to put the photons back where they are supposed to be,” says Sandra Faber, then a member of Hubble’s science team. It’s a process called deconvolution, and the team wrote algorithms to do it. The software could, for example, reveal individual stars in an image of a stellar cluster that had looked like one big blob.

With a revised target list and algorithmic corrections in place, Hubble moved forward with science. O’Dell recalls the telescope’s serendipitous discovery of proplyds, or protoplanetary disks: just-forming solar systems. Peering into the heart-shaped Orion Nebula, his team had been looking simply for fine-scale structure. But they saw early versions of planetary systems, which showed up as bright regions of gas lit with ionizing UV radiation from newly born stars. No previous telescope had been able to resolve such features. The finding provided a hint of what the Kepler space telescope would later confirm: that planets are the galactic norm.

Also within the Milky Way, Hubble investigated the young DG Tau, a T Tauri variable star. Such stars are so youthful they haven’t yet fused their constituents, and their brightness changes as they contract and convert gravitational energy into radiation. Hubble was able to separate the jet of material ejected by DG Tau from the star itself and to determine the distance from the star at which the jet’s material became collimated—a feat of resolution that ground-based scopes could not accomplish. Studying the active outflows of young stars provides hints about how those objects evolve into older, more stable stars like the Sun.

The telescope also investigated the other endpoint of the stellar life cycle. Astronomers pointed Hubble at Supernova 1987A, the remnant of a massive star in the Large Magellanic Cloud, whose demise had been visible to the naked eye just a few years before. The telescope revealed that the remnant’s cloud of material was rocketing outward at an average of 6000 km/s, and that radiation from the explosion had created a “thin, tilted ring.”

By comparing the angular size of the ring determined by Hubble with the absolute size from earlier UV observations, astronomers calculated the distance to the supernova remnant in 1991. Then, on the basis of an informed guess about the location of the supernova within the Large Magellanic Cloud, they estimated the distance to that nearby galaxy to be about 50.1 kpc. That distance represents an important rung on the ladder that’s used to estimate greater distances, and in turn to determine how fast the universe is expanding and the characteristics of dark energy.

To understand how galaxies have evolved over cosmic time, Richard Griffiths, who worked on Hubble’s Wide Field and Planetary Camera at the Space Telescope Science Institute in Baltimore, Maryland, decided to move forward with a project called the Medium Deep Survey. During targeted observations by other instruments, the camera would stare at the slices of sky where it happened to be pointed. To spy on the universe in its younger years, Griffiths’s team observed galaxies between 3 billion and 10 billion light-years away. “We discovered the smallness of galaxies,” he says.

Those little galaxies resolved into a variety of morphologies under Hubble’s gaze, although the telescope could not see the details of faint galaxies and studied them just in statistical aggregate. Only after Hubble was fixed, Griffiths says, were astronomers able to see those dim collections of stars in more detail and observe that “there were a lot of mergers going on.” Those primordial galaxies, it seemed, were rearranging into the smooth ellipses and sleek spirals we see closer to home.

Faber and Tod Lauer, who helped create the deconvolution algorithm that made much of the early science possible, investigated one such modern galaxy. When they pointed Hubble at elliptical galaxy NGC 7457, they saw that much of its light came from a tiny central region where stars were packed 30 000 times closer than they are near Earth. Scientists had never been able to discern such super-urban density in a galaxy beyond our Local Group. Subsequent Hubble observations confirmed that NGC 7457, like most galaxies, contains a supermassive black hole at its center.

Though clearly capable of advancing observational astronomy, Hubble was still hobbled and limited in its ability to observe the faint, way-off objects that largely motivated the mission in the first place. That’s why, in December 1993, two astronaut teams spent five spacewalks reengineering the telescope to boost it to its full potential. Today, 30 years after launch, and still showing the universe’s proverbial pores in its portraits of the universe, Hubble seems to have been worth the trouble.