Could Pueo Nui be competitive?

Since Pueo came in operation in 1996, larger telescopes have been equipped with AO systems. These are the 5-m Palomar telescope and one of the 10-m Keck telescopes. Last June, the 8-m Gemini North telescope produced its first images with Hokupa'a. The second 10-m Keck telescope, and the Japanese 8-m Subaru telescope will also be soon equipped with AO. The four European 8-m VLTs will quickly follow. Could Pueo Nui be competitive in front of such instruments?

Surprisingly it could. This is because of a fundamental limitation of AO. The larger the telescope, the more modes an AO system must correct. To produce the same image quality at the same wavelength, the number of compensated modes must grow as the square of the telescope diameter. For instance to produce the same image quality as Hokupa'a on the CFHT, the Keck AO system would have to compensate 150 Zernike modes. Currently, it is one of the most powerful AO system ever built. However, it compensates only the equivalent of 35 to 40 modes, which falls short by a factor of almost 4. Because the technology does not exist, it is unlikely that a more powerful AO system will be built in a near future.

The consequence is that AO systems on large telescopes are bound to operate at long wavelengths, in the near-thermal infrared (H and K bands). At these wavelengths, they will outperform Pueo Nui because they will produce sharper images and/or use fainter guide sources. At shorter wavelengths, they will not perform as well. In the visible, they will not substantially improve CCD images. If the same systems were installed on the CFHT, they would perform only slightly better than the current Pueo system in the near-thermal infrared, but they would considerably extend the capability of Pueo toward shorter wavelengths all the way down to the visible (at least for bright sources). In this region, they would produce images with an angular resolution comparable to that achieved on much larger telescopes in the near thermal infrared. They would therefore produce an information very complementary to that obtained on large telescopes, which can only be obtained with a smaller telescope like the CFHT. This is a very important reason to upgrade Pueo. Equipped with a 36-actuator Pueo Nui system, the CFHT would be the only ground-based telescope able to compete seriously wih the Hubble Space Telescope (HST), assuming that the cryo-cooled NICMOS works as expected starting in 2001. Advantages of Peo Nui will be a smaller pixel size (35 mas versus 43 mas for HST), a larger field of view (36" x 36" versus 11" x 11" for HST/NICMOS), the flexibility to use any type of filter, and the availability of observing time

Another reason to upgrade Pueo has to do with the optical quality of the CFHT. An important application of AO is the sudy of the circumstellar environment. Using the central bright star as a guide source, one looks for faint companions (possibly brown dwarfs or large gaseous planets), or circumstellar dust (disks or dust shells) in the near environment. In this case light scattered from the central star is the main limitation, and the optical quality of the telescope becomes essential. Another example is the detection of Neptune's dark satellites and the arcs on the Adams ring (Nature, vol. 400, p. 731) which was made possible by the CFHT high optical quality. Only the lowest aberrations are compensated by AO. Small scale defects in the wave front remain uncompensated and scatter light at larger distances were one wants to observe (typically of the order of one arcsecond). Light scattered by the atmosphere is quickly smoothed in a long exposure, but light scattered by the telescope produce speckles which are often unstable and very difficult to calibrate out. Fig. 2 shows an example of PSF produced in the H band by Hokupa'a on the CFHT. To date, no other AO-equipped telescope has yet produced such a clean PSF. This is because all the larger telescopes have a lower optical quality than the CFHT. Discontinuities between the Keck segments is an important source of scattered light. The primary mirror figure is a limitation for the Palomar 5-m telescope, and may also be a limitation for Subaru. The light weight secondary mirror of Gemini is its main limitation. If it were equipped with a high order AO system, the CFHT would remain the best telescope to detect faint stellar companions.
 
 

Fig. 2. A typical PSF recorded by Hokupa`a on the CFHT in the H band. It is  displayed here in a log-log scale (black dots). Diffraction by the telescope aperture (full line), atmospheric turbulence (dotted line), and telescope optics (dotted line) all contribute to scattered light.

A third reason to upgrade Pueo is the availability of telescope time on the CFHT. With the improved resolution provided by AO, more and more objects which showed no significant time evolution on a large scale are found to evolve with time and need to be monitored. Because of the competition for telescope time, this will be better done on smaller telescopes especially if they can achieve the same angular resolution as larger ones by observing at shorter wavelengths, that is if they are equipped with a sufficiently powerful AO system. An area where AO has been highly productive is the study of our own solar system. Fig. 3 shows an infrared image of Neptune obtained with Hokupa`a on the CFHT. Image quality is comparable if not better than that of Neptune's HST images obtained in the same wavelength range, allowing the details of Neptune's atmospheric activity to be monitored from the ground for the first time. Because ground-based monitoring is a very important complement to space missions, NASA has provided funding for a 36-actuator Hokupa`a-type system on the IRTF which has only a 3-m primary. Both Canadian and French planetary observers would highly benefit from an upgraded Pueo system on the CFHT.