By/par Ray Carlberg
(Cassiopeia – Summer/été 2016)
Some of the background of the loss of the site permit in Hawaii was discussed in the Vernal Equinox edition of Cassiopeia. TMT has announced that it plans to resume construction in April 2018 or sooner, but of course it needs a site on which to build. Since the project was scientifically, technically and financially built around Hawaii we would all like to build there, and, it is not that easy to simply transport an entire multi-national project someplace else. Doubly so when significant decisions need to be made over the short time left this calendar year.
Hawaii continues to provide an impressive flow of news stories about TMT. There is a laudable local effort to create a more positive environment, but that is a long term effort that seems very unlikely to overcome a deeply entrenched opposition in a short time. The legal process required to grant a permit is itself clouded. It took about 5 months to appoint a hearing officer to oversee the process. The opponents pointed out at least one significant appearance of bias. Their case is strong enough that all parties, pro and con, have petitioned for a new hearing officer. In TMT’s case we do not want to go through a long permit process knowing that there is already a substantial basis for a court to rule the entire effort to be invalid. Somewhat astonishingly, Hawaii has recently reaffirmed the hearing officer’s appointment.
So, TMT is looking for an alternate site. the list of really good sites is not all that long. The physics of an astronomical site is interesting in itself and reasonably simple. The air needs to have as little turbulence as possible and be as dry as possible. Turbulence in the air has three components. Mountains are always windy and the aerodynamics of wind over the surface gives rise to a ground shear layer that becomes turbulent. Getting above the local ground layer is the reason that CFHT has a 44m high dome for a 3.57m telescope. TMT has a 60m high dome so gets a little further into good air. Ground heating of the air on the way to the mountain generates thermal convection between the ground and the cloud layer at around 3000m. The development of turbulent convection is minimized for steep mountains near coastlines, high enough to get above cumulus clouds. Good seeing needs to start with higher altitude airflow that is as laminar as possible, which occurs in the trade winds. These originate in the Hadley cell airflow in which the thermal energy of air near the equator causes it to rise high in the atmosphere, descending at +/- 20 degrees as extremely dry air with remarkably steady wind. It also generally means relatively cloud free skies (and often deserts on the ground). The outcome is that at a good site there is mainly a ground layer, which good design can minimize, and, a shear layer near the tropopause around 10km altitude, which airplane passengers frequently experience. The beauty of having a single dominant turbulent layer distorting light is that it means that relatively simple adaptive optics systems can focus (conjugate) on and provide remarkable correction, although better correction and wider fields of view does require 3D correction.
The other major factor in a good site is simply height. A higher mountain increases UV transparency and reduces the water in the atmosphere which has such a dense spectrum of lines that both reduce transparency and emit radiation. Chile has the advantage of a very cold current from Antarctica off the coast rather than the tropical water of Hawaii which leads to more water column at the same height. For some observations, water is simply a nuisance, but for those interested in exoplanets, water is an important molecule to detect. And of course physics of molecular lines means that lots of interesting molecules have their lines in the infrared and mid-infrared. A non-scientific issue is that costs tend to go up with height as well, but ALMA and other telescopes (including the ACT built by Empire Dynamic Structures) have greatly developed high altitude construction techniques. Another non-scientific factor is that people have been climbing mountains for a long time and nobody in TMT wants to go to another mountain with cultural artefacts and native population connections.
All that doesn’t leave a long list of viable mountains. The best are Mauna Kea in the north and in the south the mountains of the northern Atacama Desert of Chile. Polar sites are also very good, but have additional technical challenges and high costs. To help make a comparative scientific assessment manageable, TMT developed the “Nelson-Cohen” site merit function, with an estimate of the speed of acquiring a given scientific goal in the optical in natural seeing, near-infrared with AO and mid-infrared with AO. These are representative science uses, and specific ones will give somewhat different outcomes. The sites currently under consideration are Mauna Kea (13N at 4050m), Roque de los Muchachos Observatory (ORM) in the Canaries (2250m), San Pedro Martir (SPM) in Baja Mexico (2830m), Vicuna Mackenna (3100m) near Paranal and Honar (5400m) south of the ALMA site. For optical band natural seeing observations Mackenna is best with Honar, Mauna Kea and ORM being about 15-18% less desirable although transparency is a larger issue at ORM. SPM is about 28% less effective than Mackenna. In the NIR Honar is best with MK 13N and Mackenna down 14-18% and SPM, ORM down 34-42%. In the mid-IR Honar truly stands out, with MK13N and Mackenna down 40-43% and SPM and ORM down 53-56%. In the south, the choice between Honar and Mackenna depends on how important mid-IR is for your science. By any assessment AO assisted observations and NIR and mid-IR are becoming more important uses as both science interests move to redder bands and as IR technology improves. If only a small fraction of the time was required for mid-IR, a poor site does give that from time to time, which a queue system (not in the TMT baseline) could give. However, for time critical observations (such as planet transits and all sorts of LSST followup) a small queue fraction is not likely to be satisfactory.
The E-ELT is being built on a site essentially identical to Mackenna although Mackenna does have room for more telescopes in the future. Clearly TMT on Hawaii would give us a competitive telescope with complementary sky access. Other sites in the North would give us a smaller telescope on a worse site, which would be hard to justify. A high site in Chile would give us an advantage over E-ELT for a range of interesting science. Canadians are of course familiar with the advantages of dealing with the difficulties and long term payoff of pioneering a high, remote site, which Hawaii once was.
The TMT fund is not money that astronomers can move around at will which is true for all members of TMT, which is where the complications begin. The Canadian SAC members have so far mainly pointed out the increased scientific return of the Chilean sites, particular relevant to scientific observations (exoplanets) which are likely to become a major use of ELTs. It would be sad to leave the North but we may have no choice. We in Canada have built our case on the century of tradition of “second to none” which has important implications for the TMT site choice for Canadian astronomers.