High-contrast imaging

Using imaging to detect an Earth-like planet circling a star like our Sun is not possible with today’s technology. The light reflected by the Earth is ten billion times dimmer than that emitted by the Sun, and no instrument can detect such a weak source so close to such a bright one. The same is true for giant planets like Jupiter, which remain undetectable with current instruments even though they reflect about 100 times more light than Earth-like planets.

For nascent planetary systems, however, the situation improves. When giant planets are forming, they are at a temperature of around 3000 °C. They cool rapidly in the subsequent millions of years by emitting light, mainly in the infrared domain. During this cooling phase, the ratio between the brightness of the planet and the star is approximately 100,000 to 1 million. Although detecting the planet with today’s imaging technology is a challenge, it is possible none the less.

Difficult beginnings

Work to directly detect exoplanets started at the Université de Montréal some 15 years ago. For his doctoral dissertation supervised by René Doyon and Daniel Nadeau, Christian Marois built an instrument called TRIDENT in order to detect exoplanets. The instrument, used on a number of occasions in tandem with the CFH telescope, did not lead to the discovery of any exoplanets, but the very detailed understanding of the problems that limited TRIDENT’s sensitivity allowed the UdeM team to suggest different techniques for substantially enhancing the sensitivity of high-contrast imaging instruments.

Following the TRIDENT experiment, Christian Marois and the Institute team developed 2 techniques: angular differential imaging, or ADI, and locally optimized combination of images, or LOCI. Combining ADI and LOCI made it possible to improve sensitivity by a factor of 30 or more as compared with the previous techniques. ADI and LOCI are now recognized as the most effective techniques for observing and analyzing data in the field of high-contrast imaging, and the advances made by the Institute’s team have played a considerable role in recent discoveries of exoplanets by other teams (examples: 12345).

HR8799 and 1RXS J1609-2105: the first discoveries

Figure 1. Star 1RXS J1609-2105 (centre) and its giant planet (upper left) discovered by David Lafrenière in 2008. Additional observations published in 2010 made it possible to identify various molecules in its atmosphere, including water vapour and carbon monoxide.

The development of new observation and data processing techniques allowed the Institute’s team to carry out one of the first large-scale surveys to determine the frequency of massive exoplanets around young stars in the solar neighbourhood. The GDPS survey conducted by David Lafrenière established that such planets are rare and that fewer than 10% of stars have exoplanets 0.5 to 10 times Jupiter’s mass in the outskirt regions of planetary systems.

The team owed its first successes in directly detecting planets to David Lafrenière, who surveyed planetary companions around stars in a group of young stars called the Upper Scorpius association. Based on his observations made at the Gemini North observatory in Hawaii, he obtained the first image of a planet orbiting a Sun-like star. (Figure 1).

Figure 2. HR8799 and its 4 planets. Planets b, c and d were discovered in 2008 and planet e, 2 years later.

Another survey around stars hotter than the Sun was done by Christian Marois in collaboration with David Lafrenière, René Doyon and an international team, and led to the discovery of a planetary system that to date remains unique in the annals of astronomy; the team identified 3 giant planets (7 to 10 times the mass of Jupiter) orbiting HR8799, a star more massive than the Sun. The development of LOCI also allowed the Institute to rediscover 2 of the planets of HR8799 from data obtained by the Hubble telescope in 1998, and which were undetectable at the time because of the lack of sufficiently powerful analytical tools. The HR8799 system has been studied many times since it was discovered, with the initial publication describing the system cited over 500 times in astronomical publications over the past 5 years.

The future: GPI and JWST

The Institute’s team is involved in 2 projects that will make it possible to directly detect new massive planets around nearby stars. GPI (Gemini Planet Imager) is an instrument that has just been installed at the Gemini South observatory, in Chile. It is a much more sensitive device than those currently operated on large telescopes. It can be used to find planets that are much less massive than those detectable during previous survey campaigns, such as the GDPS campaign. GPI has several components including an advanced adaptive optics system, a sophisticated internal calibration module and a spectrograph working between 1 and 2.4 µm. The Institute’s team built the GPI spectrograph and helped develop the analysis algorithms for GPI data, largely inspired by those used for LOCI and ADI.

The Institute team also developed an instrument called a Near-InfraRed Imager and Slitless Spectrograph (NIRISS) for the JWST telescope. NIRISS has a unique feature: a non-redundant mask (NRM) that blocks out all but 15% of the light from a star on carefully chosen sections of the telescope mirror. The NRM solves a number of observation problems limiting the sensitivity of planet detection and makes it possible to find planets much closer to their stars than the other JWST observation methods. This mode operates at a wavelength of 4 µm, which is difficult to achieve using a ground-based telescope. NRM data will make it possible to measure the infrared flux of a number of the planets discovered by GPI. Étienne Artigau is the Institute member responsible for this mode.