Features Physics World  June 2016

Widening our view of the universe

NASA is designing an ambitious near-infrared space observatory with a field of view up to 200 times greater than Hubble’s to understand why the expansion of the universe is accelerating and to explore other planetary systems, describe Bert Pasquale, Cathy Marx, David Content and Jeffrey Kruk

Ambitious An artist’s impression of NASA’s Wide Field Infrared Space Telescope (WFIRST). (NASA)

The 1998 discovery that the universe is expanding at an accelerated rate raised new and profound questions about the nature and fate of the universe. Based on measurements of the speed of distant supernovae, the discovery pointed to the existence of an invisible “dark energy” that dominates the universe’s mass-energy content and somehow drives the cosmic acceleration. Indeed, together with dark matter, physicists are blind to more than 95% of what makes up the universe. To map out the unseen we have to study the effects of dark energy on that which we can see, namely infrared light from distant galaxies, which means going into space.

Since the discovery of cosmic acceleration, various missions have been proposed to reveal the nature of dark energy, including NASA and the US Department of Energy’s Joint Dark Energy Mission study. The fruit of this effort was the 2010 categorization of the Wide Field Infrared Space Telescope (WFIRST) as a top priority by the US National Academy of Sciences. WFIRST is now a six-year mission being undertaken by NASA to observe the effects of dark energy from an L2 halo orbit 1.5m km from Earth.

A small team immediately began to work on design concepts for a metre-plus-class observatory and set out its scientific goals. Then, in 2012, it was announced that hardware for an optical telescope assembly with a Hubble-sized (2.4 m diameter) primary mirror, valued at more than $300m, was being transferred to NASA from the defence community. It was quickly ascertained that this hardware could be used for a dark-energy survey. The larger observatory, with twice the resolving power and more than triple the light collecting area, offered up to four times the scientific return of the initial design.

Since then, teams at NASA’s Goddard Space Flight Center in Maryland and the Jet Propulsion Laboratory in California have been working hard to pull together payload and instrument architectures that meet the scientific goals, while also overcoming the engineering challenges of designing around existing hardware. In December 2015 the study team presented its “Cycle 6” design to the Mission Concept Review Board. As a result, in February this year, NASA gave the mission the green light with WFIRST set to launch in the mid-2020s. Eleven science teams comprising more than 200 members are now in place to guide the mission goals.

Mission goals

WFIRST will tackle two key questions about the acceleration of the universe. First, it will address whether the speed-up is caused by a new energy component or simply by the breakdown of our understanding of gravity on cosmological scales. Second, if the acceleration is caused by a new energy component, WFIRST will see if its energy density is constant in space and time or if it has evolved over the history of the universe. To achieve this, the observatory will carry out three types of survey.

An imaging survey will measure the shapes and redshifts of a large number of galaxies and galaxy clusters, and will observe how their images are stretched as they pass tangentially around the location of dark matter. This “gravitational weak lensing” survey will cover 2000 square degrees (1/20th of the sky) to observe around 350 million lensed galaxies and 40,000 massive galaxy clusters. The results will produce a map of dark matter and cosmic structure rate of growth, thereby tracing out the distribution and time-evolution of dark energy.

A supernova survey will observe the spectral evolution of 2700 Type 1a supernovae as far out as 10.5 billion light-years via integral-field spectroscopy to correlate their distance (inferred by their brightness) and speed (inferred from their redshift). Finally, WFIRST will carry out a spectral survey of galaxies to study the imprint of primordial sound waves on the clustering of galaxies, also known as baryon acoustic oscillations. A survey of 22 million galaxies between 8 to 11 billion light-years away  will provide an absolute calibration of the expansion-rate history, as well as tests of general relativity based on localized redshift distortions in galaxy clusters.

Taken together and correlated, these three surveys vastly increase the accuracy of the 3D maps of the distribution of matter and gravity, and measure the expansion history of the universe, thereby revealing the temporal evolution of dark energy. WFIRST dark-energy programmes will measure cosmic expansion with an aggregate precision of 0.1–0.5%, which will be improved further as WFIRST’s cosmological data are combined with ground-based instruments such as the Large Synoptic Survey Telescope and other space-based missions such as the European Space Agency’s Euclid telescope.

Another mission goal for WFIRST is exoplanet science. A precise census of planets beyond our solar system is important for both understanding the formation of planetary systems and for determining the number of Earth-like planets in our galaxy. When a star in our galaxy passes between a more distant star, the gravitational microlensing effect produces a temporary increase in the brightness of the distant object. If that star also has a planet, an even shorter intensity spike may be observed, providing information about its orbit and mass. WFIRST’s microlensing survey will monitor 200 million stars in the galactic bulge over six two-month periods, with the hope of finding 2600 planets at least as large as Mars, including 300 Earth-mass planets. The microlensing survey will fill the gaps of missions such as the existing Kepler probe and the future Transiting Exoplanet Survey Satellite by revealing both smaller and farther-orbiting planets.

In addition to the wide-field surveys, WFIRST will also host a demonstration of technology for the first space-borne wavefront-controlled exoplanet coronagraph, which will image dozens of planetary systems orbiting other stars within a 40-light-year radius. The instrument will be able to see exozodiacal dust disks, far-out gas giants and smaller rocky planets nearer to their host star that are highly reflective. Direct imaging will even make it possible to detect changes in both brightness and spectrum as some exoplanets rotate on their axes. Such non-uniformities could indicate land and water masses.

Beyond Hubble

One way to quantify a telescope’s scientific power is its “A*Omega” value, which is the product of its aperture and its field of view. WFIRST’s primary mirror shares the same 2.4 m-diamteter aperture as Hubble, but Hubble was designed to accommodate five science instruments plus three fine guiders, each with limited fields of view. WFIRST has just one large Wide-Field Instrument (WFI) with three main modes, and the Coronagraph Instrument (CGI) with two main modes. But they pack a punch: of the 20 scientific questions listed in the National Academy of Sciences’ decadal survey of astronomy and astrophysics, WFIRST addresses 15 of them.

The WFI contains the Wide-Field Channel (WFC), which can view 0.28 square degrees of sky (about the same area as the Moon) with a resolution of 300 megapixels. In fact, the WFC will cover almost 90 times more sky than Hubble’s Advanced Camera for Surveys, and more than 200 times more sky than Hubble’s wide-field infrared mode. Furthermore, whereas Hubble’s low-Earth orbit allows only a 60% average orbital visibility period of just under one hour per 96-minute orbit, WFIRST’s L2 orbit allows continuous operation that will outpace Hubble’s infrared survey capabilities by a factor of 360.

Triple vision The WFIRST payload consists of a Wide-Field Instrument (right half of image), which contains the Wide-Field Channel and Integral Field Channel, and the Coronagraph Instrument (left half). A three-mirror anastigmat comprising primary, secondary and tertiary mirrors allows a large area of sky to be mapped at high resolution. (NASA/B Pasquale (GSFC))

So how does WFIRST maintain or improve the quality of its images and increase the field of view despite having the same mirror size as Hubble? The answer is an advanced optical design that reuses the telescope’s primary and secondary mirrors, then adds a third tertiary mirror to create what is called a three-mirror anastigmat (TMA). Unlike Hubble’s two-mirror Ritchey–Chrétien design, a TMA can be optimized for a wide and flat diffraction-limited field. Therefore, a large array of sensors can be used at the focal plane to image a very large piece of the sky in high resolution.

This is perfect for WFIRST’s imaging, spectral and microlensing surveys. The WFI boasts a 300 megapixel array of 18 custom mercury-cadmium-telluride CMOS detectors, each with 4088 × 4088 pixels measuring 10 µm across to see wavelengths from 0.76–2.0 µm. To accomplish this, NASA’s optical designers used a carefully selected set of design parameters and constraints in optical-design software. The specific optical design is the result of a delicate balance of performance, costs and programmatic issues.
One challenge in designing a system with such a high value of A*Omega is the tight volume and interface constraints. Despite having more than three times the volume of Hubble’s Wide-Field Camera, two additional flat mirrors are required to fold the optical path into the instrument volume. A carbon-fibre bowl-shaped motorized carrousel called the element wheel interchanges six bandpass filters and one spectroscopic dispersion unit called the grism. The grism combines a diffraction grating, powered prism and binary optical elements to maintain image quality over a wide field.

Another challenge was to extract the spectrum of a supernova cleanly from that of the host galaxy for the supernova survey. To do this the WFI also contains a breadbox-sized instrument called the Integral Field Channel, which contains a specialized image slicer that uses 120 micromirrors that can dissect a single distant galaxy into 0.15 or 0.3 arcsecond samples before passing it through a spectrograph. A small section of WFIRST’s telescope’s field is also picked off for the CGI. In order to image orbiting exoplanets within 0.1 arcseconds of the host star, its light must be blocked with a contrast factor of at least 1 part per billion. That is 1000 times greater than current instruments, including those on NASA’s James Webb Space Telescope (JWST). Cutting-edge technologies were implemented to achieve precision wavefront control unlike anything flown in space, including sensing and controls for two deformable mirror arrays, a fast steering mirror and a pupil occulting mask.

Many institutions and facilities are involved in making the WFIRST mission successful. Currently, the project team is developing a clear path for WFIRST integration and testing activities, including verification plans for each subsystem, individual instruments and the entire observatory. To accomplish this the team is refining optical error budgets and integrated modelling methods to account for each step of the way, from fabrication and assembly to on-orbit commissioning. Development and risk-reduction activities are still under way as WFIRST heads towards a detailed design in 2017.

To the future

In science, each mission or new tool informs and enables the next. WFIRST observations of new planets by the coronograph instrument will lay the technical and scientific foundations for a future imaging and spectroscopy mission that could image the atmosphere of exoplanets to look for “Earth 2.0”. Survey results, meanwhile, will make dark-energy equation-of-state parameters 10 times more precise and perhaps tell us how dark energy may eventually cause the fabric of space–time itself to break down.

Regardless of the results, the data produced by WFIRST will open up new questions for future missions to explore, and call for a new set of instruments. Flagship missions such as Hubble, JWST and WFIRST span more than two decades from initial studies to final launch, and NASA researchers are already looking at what the next such mission might look like. Hubble’s big surprise was how much we did not even know to look for. With WFIRST and other tools of the coming decade, we fully expect to be surprised at the questions they will lead us to.