Discovery of gravitationally lensed galaxies with Herschel Space Observatory

Herschel Space Observatory, a European Space Agency mission with important NASA contributions, launched on May 14 2009. Its 3.5m telescope is the largest astronomical telescope to be launched into space. It houses three instruments that operate over the wavelength range of 55 to 700μm, bridging the far-infrared and sub-millimeter wavelength regimes. While the Infrared Astronomical Satellite (IRAS) in the early 1980s showed the importance of far-IR wavelengths to understanding dusty star-formations missing in optical wavelengths, we knew very little about the galaxies visible at these long wavelengths. With Herschel, we are now finally unveiling the far-IR and sub-millimeter universe.

One of the first cosmologically important Herschel discoveries comes from the sky maps made with the Spectral and Photometric Imaging Receiver (SPIRE), one of the three instruments on Herschel. SPIRE can image simultaneously at 250, 350, and 500μm. When overlaid on surveys at optical wavelengths such as the Sloan Digital Sky Survey (SDSS), the first SPIRE map of the Herschel-ATLAS survey showed a handful of bright sub-millimeter galaxies that overlapped with optical galaxies (see Figure 1). About 50% of the galaxies that are bright at sub-millimeter wavelengths are relatively close spiral galaxies with dusty star-forming regions. The rest seem to be associated with massive early-type/elliptical galaxies. With very low star-formation rates, such galaxies were not expected to be bright at far-IR/sub-millimeter wavelengths.

Figure 1. The main image is a 4×4 degree map of the first Herschel-ATLAS data. Five white squares indicate the lensed sub-millimeter galaxies, while zoomed areas show the bright red color. (Courtesy: ESA/SPIRE/Herschel-ATLAS/SJ Maddox.)

One hypothesis is that the bright sub-millimeter galaxies imaged by SPIRE are not physically associated with the galaxies seen optically. Instead, they are along the same line of sight at distances much farther away. Such fortuitous events—where two galaxies overlap in the same line of sight—can be expected. The dark-matter-dominated galaxy in the foreground gravitationally lenses the background galaxy, a phenomenon predicted by Einstein as part of his theory of general relativity (see Figure 2). A large gravitational lensing leads to multiple images of the background galaxy. When the alignment of the two galaxies is perfect, the background galaxy appears as an Einstein ring around the foreground galaxy. There are now more than 200 examples of these types of strong lenses in optical and radio surveys.

Figure 2. An illustration of gravitational lensing, in which a galaxy magnifies a second, more distant galaxy. The Herschel telescope and Earth are shown to the right. A foreground galaxy is shown in blue, located approximately three billion light-years away. A more distant galaxy, about 11 billion light-years away, is shown in red. The gravity of the foreground galaxy bends the light from the distant one, as shown with the red lines. The pink lines show what we actually see—a distorted and magnified view of the distant galaxy. An example of what a final image might look like is at the far left. (Courtesy: NASA/JPL-Caltech.)

Once the positions of candidate lensed galaxies are known from Herschel data, such galaxies can be examined using ground-based facilities to confirm the lensing nature and to study gas, dust, and stars in the distant galaxy (as well as mapping out the dark matter content of the foreground galaxies). The dust emission is usually mapped at sub-arcsecond resolution with the Sub-millimeter Array (SMA), an eight element, six-meter telescope array atop Mauna Kea in Hawaii.

In the first study of lensed sub-millimeter sources with Herschel,¹ SMA images revealed multiple components similar to known gravitational lensing events (see Figure 3). The lensed galaxies, while bright in sub-millimeter, are extremely faint in the optical, mostly due to the extinction associated with copious amount of dust in actively star-forming galaxies. The lack of detectable visible light leads to additional challenges.

Figure 3. This composite shows a gravitationally lensed and magnified galaxy discovered by the Herschel telescope. The galaxy SDP 81 is the yellow dot on left (Herschel) and the pink smudges on right(taken by the Sub-millimeter Array in Hawaii). The galaxy is 11 billion light-years away and buried in dust, but it is magnified and distorted by a foreground galaxy (blue blob at right), causing it to appear in multiple places, seen as the pink smudges. The foreground galaxy was then spotted in optical light using the Keck Observatory. (Courtesy: ESA/NASA/JPL-Caltech/Keck/SMA.)

The redshift of a galaxy is associated with the change in wavelength of the emitted light. Traditionally, redshift measurements involve spectral observations at optical wavelengths that look for emission and absorption lines. Instead of optical redshifts, lensed galaxies are measured directly at millimeter wavelengths using spectrometers outfitted to the 10.4m Caltech Sub-millimeter Observatory (Z-spec) and the 100m Green Bank Telescope (Zpectrometer). Such instruments look for the multiple transitions of the rotational lines of the carbon-monoxide molecule. Z-spec covers 195 to 300GHz with a parallel-plate waveguide grating and an array of 160 silicon nitride micromesh bolometers operating at 100mK.² Zpectrometer covers the frequency range of 26 to 40GHz with a set of analog lag correlation spectrometers.³ Redshift measurements are also now coming from interferometers such as the Plateau de Bure Interferometer in the French Alps and the Combined Array for Research in Millimeter Astronomy in eastern California. Unlike spectrometers, these provide information on the spatial distribution of the gas by tracing molecular gas.

While rare, lensing provides both magnification and a corresponding increase in spatial resolution to study intrinsically faint distant galaxies. Thus, lensed galaxies are likely to be key targets for high-resolution observations with the Atacama Large Millimeter Array (ALMA), an international astronomy facility on the Chajnantor plain of Chilean Andes. Once fully constructed, ALMA will involve 66 telescopes operating between 0.3 and 3.6mm. Early science operations of ALMA are expected to begin in 2011 with 16 telescopes. The discovery rate in our initial study shows that existing extragalactic programs with Herschel can easily find a sample of between 300 and 500 lensed galaxies before the end of 2012.

Herschel-SPIRE might also be able to map beyond the current 750 square degrees of sky area. A survey that moves the telescope rapidly to scan once with SPIRE could cover a very wide area in a reasonable observing time, leading to the anticipated discovery of about 2000 bright-lensed galaxies.⁴ Such a large sample of lensed sub-millimeter sources will be very beneficial for cosmology, with follow-up observations spanning a decade or more over the entire electromagnetic spectrum.

The author gratefully acknowledges the Herschel SPIRE Instrument Team, Herschel-ATLAS survey consortium, and the directors, staff and scientists involved with ground-based follow-up observations of Herschel sources. Herschel is a European Space Agency cornerstone mission, with science instruments provided by consortia of European institutes and the Herschel Project Office at NASA's Jet Propulsion Laboratory.