Another exciting result obtained by high-dispersion spectroscopy

The second topic relating to high-dispersion spectroscopy is the extra-solar planet search. HDS is an optical spectrograph with high wavelength resolution that enables us to measure the intensity of light with each 1/100,000 wavelength interval. By using this high spectral resolution, we have the chance to find extra-solar planets like earth.

How can we find extra-solar planets?

Doppler shift

For an observer far away from the sun, spectral lines of the sun shift to blue when the sun is approaching the observer. On the contrary, spectral lines shift to red when the sun is receding.

For observers far away from the solar system, it is difficult to detect the faint light from the earth because of strong radiation from the sun. However, observers could discover the existence of planets by measuring the sun's periodic motion. Now, assuming our solar system consists of only the sun and the earth, we would say that the earth orbits the sun. However, this statement is not correct in the strictest sense. Actually, both the sun and the earth orbit around the center of gravity of the two bodies. However, the center of gravity is close to the center of the sun because the sun is so huge compared to the earth. Since the sun is also orbiting around this center of gravity, its motion can be detected as periodic oscillation in the observed solar spectrum. When the sun approaches the observer, spectral lines of the sun shift to blue, and when the sun recedes, the lines shift to red. This is known as the Doppler shift. Therefore, by observing the periodic variation in the Doppler velocity of the spectral lines we can find the existence of planets orbiting bright stars. Very small Doppler velocity variation can be observed with HDS enhanced with optional tools.

We have observed many candidate stars that might have orbiting planets and searched for the periodic spectral variations caused by orbiting planets. On the basis of data on periodic variation, we can estimate the mass of a planet and the separation between a star and a planet.

The extra-solar planet search is currently one of the most exciting projects. Over 100 extra-solar planets have been found so far, almost all of which are gigantic planets like Jupiter. We expect to detect earth-like planets in the near future by improving the accuracy of the Doppler velocity measurements

Principle of the grating spectroscopy

“High dispersion spectrograph,” or HDS, is an instrument used to obtain high wavelength resolution spectra using a reflective Echelle grating, on which evenly spaced grooves disperse light depending on the wavelength.

Take a look at the diagram. Incident parallel light falling onto two adjacent grooves A and B on the surface of the grating results in an optical path difference of Δℓ after diffraction. If the path difference (Δℓ) is an integral multiple of the wavelength, the diffracted light waves are mutually strengthened in such directions that the equation Δℓ=mλ (m: integer, λ: wavelength) is satisfied, which is called m-th order diffraction.

When the diffraction angle is specified multiple orders (m=0, 1st, 2nd, ...), diffracted light can be detected simultaneously. If the diffraction angle deviates a little, the wavelength of the diffracted light changes a little, so the incident light can be dispersed depending on the wavelength. You have probably learned that light has a dual nature, particle and/or wave. Diffraction is the typical nature of the wave.

Principle of diffraction by the grating

The above figure shows that there is an optical path difference of Δℓ between light beams diffracted at two adjacent grooves A and B. If the optical path difference (Δℓ) is an integral multiple of the wavelength (m times), the light is strengthened. When m=1 (i.e. λ0=Δℓ), the diffraction is called 1st order diffraction. Higher order (2nd , 3rd, ..., mth order diffraction) corresponds to the wavelengths of λ0/2, λ0/3, .....λ0/m. Light beams with wavelengths of λ0+Δλ, (λ0+Δλ)/2, (λ0+Δλ)/3, etc. are strengthened in the slightly deviated direction.

The merits of higher order diffraction

HDS has a high wavelength resolution capable of measuring light intensity at every 1/100,000 wavelength interval. Since the wavelength difference (Δλ/m) for a given diffraction angle shift (Δθ ) is smaller for higher order diffraction, high-dispersion is easily realized for high order diffraction.

However, for the high order diffraction spectrum, many orders (...(m-2)^th , (m-1)^th, m^th ,(m+1)^th, (m+2)^th...) are superposed, so we have to separate individual orders. The cross-disperser grating separates this superposed spectrum into individual orders in a direction perpendicular to the direction dispersed with the Echelle grating. Since the directions of the dispersion caused by Echelle and cross-disperser are perpendicular from each other, final spectrum (Echelle spectrogram) extends two dimensionally, which can be obtained by the 4,000x4,000 pixel CCD put on the focal plane of camera optics.

HDS optical layout

Left: Sample of Echelle spectrogram extended two dimensionally
Right : A portion of raw HDS spectrum of comet LINEAR C/1999 S4). The longest wavelength is at the upper-right corner, and that of the shortest wavelength is at the lower-left corner. Bright spots in this image correspond to NH₂ emission lines.

How did you become involved in the development of HDS?

I started studying infrared astronomy at graduate school. This field was at a primitive stage at that time, and we were obliged to make our own instruments. There weren't many astronomers in Japan in those days that had experience in developing spectrometers, so my experience was valuable for the HDS project. I participated in the HDS project in 1995 because I had experience in making infrared spectrometers, though only small ones.

The name of Nikon was, of course, familiar to me. Nikon's technology was fairly well known to Japanese astronomers because instruments developed by Nikon had contributed to Japanese astronomy even before Subaru's construction. However, I only really came to understand Nikon's abilities after we began collaborating. I think our demands exceeded our budget from Nikon's view-point and I was afraid we would get Nikon employees into trouble (laughs). Looking back on our collaboration, we owe the success of the HDS project to the excellence of the Nikon staff assigned to the project. I would like to express my gratitude to Nikon's staff. Actually, I'm quite sure that this interview is due to the success of the HDS project achieved because of the remarkable efforts made by Nikon's staff. There are still many unknowns in astronomy but I hope our ongoing collaboration with Nikon will contribute to future progress in astronomy.

• Some images are courtesy of the National Astronomical Observatory of Japan

• National Astronomical Observatory of Japan website
• Subaru Telescope website

Posted July 2006