A small step toward seeing habitable planets

March 10, 2014

An image of the exoplanet Beta Pictoris b taken with the Magellan Adaptive Optics VisAO camera. This image was made using a CCD camera, which is essentially the same technology as a digital camera. The planet is nearly 100,000 times fainter than its star, and orbits its star at roughly the same distance as Saturn from our Sun. (Credit: Jared Males/UA)

University of Arizona researchers have snapped images of a planet outside our solar system with an Earth-based telescope using a CCD imaging sensor, which is also found in digital cameras, instead of an thermal infrared detector.

“This is an important next step in the search for exoplanets because imaging in visible light instead of thermal infrared is what we likely have to do if we want to detect planets that might be suitable for harboring life,” said Jared Males, a NASA Sagan Fellow in the UA’s Department of Astronomy and Steward Observatory and lead author on a report to be published in The Astrophysical Journal (also available on open-access arXiv).

The image was taken at a wavelength (985 nanometers) that is close to visible red light, showing that the use of a digital camera-type imaging sensor — called a charge-coupled device or CCD — opens up the possibility of imaging planets in visible light, which has not been possible previously with Earth-based telescopes.

“This is exciting to astronomers because it means we now are a small step closer to being able to image planets outside our solar system in visible light,” said Laird Close, a professor in the Department of Astronomy, who co-authored the paper.

Cooler planets require visible-light imaging

He explained that all the other Earth-based images taken of exoplanets close to their stars are far-infrared (thermal) images, which detect the planets’ heat. This limits the technology to Gas Giants — massive, hot planets young enough to still shed heat. In contrast, older, possibly habitable planets that have cooled since their formation don’t show up in infrared images as readily, and to image them, astronomers will have to rely on cameras capable of detecting visible light.

“Our ultimate goal is to be able to image what we call pale blue dots,” Close said. “After all, the Earth is blue. And that’s where you want to look for other planets: in reflected blue light.”

The photographed planet, called Beta Pictoris b, orbits its star at only nine times the Earth-Sun distance, making its orbit smaller than Saturn’s. In the team’s CCD images, Beta Pictoris b appears about 100,000 times fainter than its host star, making it the faintest object imaged so far at such high contrast and at such relative proximity to its star. The new images of this planet helped confirm that its atmosphere is at a temperature of roughly 2600 degrees Fahrenheit (1700 Kelvin). The team estimates that Beta Pictoris b weighs in at about 12 times the mass of Jupiter.

“Because the Beta Pictoris system is 63.4 light years from Earth, the scenario is equivalent to imaging a dime next right next to a lighthouse beam from more than four miles away,” Males said. “Our image has the highest contrast ever achieved on an exoplanet that is so close to its star.”

Adaptive optics

In addition to the host star’s overwhelming brightness, the astronomers had to overcome the turbulence in Earth’s atmosphere, which causes stars to twinkle and telescope images to blur. The success reported here is mostly due to an adaptive optics system developed by Close and his team that eliminates much of the atmosphere’s effect. The Magellan Adaptive Optics technology is very good at removing this turbulence, or blurring, by means of a deformable mirror changing shape 1,000 times each second in real time.

The team also imaged the planet with both of MagAO’s cameras, giving the scientists two completely independent simultaneous images of the same object in infrared as well as bluer light to compare and contrast.

“An important part of the signal processing is proving that the tiny dot of light is really the planet and not a speckle of noise,” said Katie Morzinski, who is also a Sagan Fellow and member of the MagAO team. “I obtained the second image in the infrared spectrum — at which the hot planet shines brightly — to serve as an unequivocal control that we are indeed looking at the planet. Taking the two images simultaneously helps to prove the planet image on the CCD is real and not just noise.”

Males added: “In our case, we were able to record the planet’s own glow because it is still young and hot enough so that its signal stood out against the noise introduced by atmospheric blurring.”

“But when you go yet another 100,000 times fainter to spot much cooler and truly earthlike planets,” Males said, “we reach a situation in which the residual blurring from the atmosphere is too large and we may have to resort to a specialized space telescope instead.”

Development of the MagAO system was made possible through the strong support of the National Science Foundation MRI, TSIP and ATI grant programs. The Magellan telescopes are operated by a partnership of the Carnegie institute, the University of Arizona, Harvard University, Massachusetts Institute of Technology and the University of Michigan. The work of NASA Sagan Fellows Jared Males and Katie Morzinski was performed in part under contract with the California Institute of Technology funded by NASA through the Sagan Fellowship Program executed by the NASA Exoplanet Science Institute.

Abstract of arXiv paper

We present the first ground-based CCD (λ<1μm) image of an extrasolar planet. Using MagAO’s VisAO camera we detected the extrasolar giant planet (EGP) β Pictoris b in Y-short (YS, 0.985 μm), at a separation of 0.470±0.010′′ and a contrast of (1.63±0.49)×10−5. This detection has a signal-to-noise ratio of 4.1, with an empirically estimated upper-limit on false alarm probability of 1.0%. We also present new photometry from the NICI instrument on the Gemini-South telescope, in CH4S,1% (1.58 μm), KS (2.18μm), and Kcont (2.27 μm). A thorough analysis of our photometry combined with previous measurements yields an estimated near-IR spectral type of L2.5±1.5, consistent with previous estimates. We estimate log(Lbol/LSun) = −3.86±0.04, which is consistent with prior estimates for β Pic b and with field early-L brown dwarfs. This yields a hot-start mass estimate of 11.9±0.7 MJup for an age of 21±4 Myr, with an upper limit below the deuterium burning mass. Our Lbol based hot-start estimate for temperature is Teff=1643±32 K (not including model dependent uncertainty). Due to the large corresponding model-derived radius of R=1.43±0.02 RJup, this Teff is ∼250 K cooler than would be expected for a field L2.5 brown dwarf. Other young, low-gravity (large radius), ultracool dwarfs and directly-imaged EGPs also have lower effective temperatures than are implied by their spectral types. However, such objects tend to be anomalously red in the near-IR compared to field brown dwarfs. In contrast, β Pic b has near-IR colors more typical of an early-L dwarf despite its lower inferred temperature.