Astronomical Site Testing in the Canadian High Arctic
Winter in the Canadian High Arctic is long, dark, and extremely cold. At the northern tip of Nunavut, in the remote military outpost of Alert, the sun sets for the last time in October and does not rise again until late February [1]. During this time the temperature is typically -35 C, but it can drop to -50 C. And moreso than almost any other place on the planet, it is dry. The total annual precipitation is only 15 cm [2]. These are harsh conditions for people, but beneficial ones for infrared astronomy, provided one can find a high vantage point with clear skies. Conditions would be ideal if that location also provided good seeing. The idea that the Canadian Arctic may possess such a spot was discussed previously in Cassiopeia [3]. Here we report the first work on the ground in the North. We have scouted four candidate sites by helicopter and begun testing on two. All are on mountains between 1400 m and 1900 m high at the northern edge of Ellesmere Island. This is roughly level in latitude with Alert, only 740 km from the North Pole.
Map of Nunavut. (Place cursor over image to overplot a relief map. Java must be enabled.)
Two mountains are being tested on
the northern end of
Ellesmere Island. The highest peak in Nunavut is Barbeau, at 2616 m. It is also
on Ellesmere, but far inland, within Quttinirpaaq National Park.
The Antarctic has already attracted the focus of considerable site-testing resources, for largely the same reasons drawing us to the Arctic. One location receiving particular attention is Dome C, on a 3200-m-high glacial plateau roughly 1400 km from the South Pole. The skies at Dome C are remarkably clear [4]. But the seeing is poor by modern observatory standards. A median value of 1.8+/-0.8 arcsec was obtained in winter with a Differential Image Motion Monitor (DIMM) on top of an 8.5 m tower [5]. Earlier excitement surrounding a mean value of 0.27 arcsec was based on a combination of Multi-Aperature Scintillation Sensor (MASS) and Sound Detection and Ranging (SODAR) measurements which are insensitive to the lowest 30 m of the atmosphere [6]. It is now known that the plateau is blanketed by a thick, strongly turbulent boundary layer. Balloon measurements show that 87% of turbulence occurs within 36 m of the ice surface [5]. Even so, this could conceivably be overcome by placing a telescope on a tall pier or mound.
Northern Ellesmere has a significant advantage over the Antarctic ice plateau in this regard. There are mountains. Although none reach 3200 m, some top 1900 m, and many are over 1400 m. These could provide a firm foundation on which to perch a telescope. No large telescope has ever been built on top of ice.
The northern end of Ellesmere Island in Nunavut. Mountains are flanked by glaciers which
flow down to the Arctic Ocean.
The scale height of the atmospheric pressure profile decreases toward the poles, which makes these mountains effectively higher than if they were at a lower latitude. They should also benefit from exceptionally smooth airflow. Winds are predominantly from the west, travelling over hundreds of kilometres of uninterrupted ice. And unlike Hawaii or other mid-latitude sites, the upper atmosphere is untroubled by the jet stream. Given these unique conditions it is difficult to predict what seeing can be expected. It may be comparable to Dome C when that location's thick boundary layer is excluded. If the ground-layer contribution is more typical of the best mid-latitude sites (for example, as estimated via the empirical model of Racine [7]) the seeing could still be as good as 0.5 arcsec - the best in the world. Skies should also be clear. Analysis of satellite images of this region by Liviu Ivanescu predicts that several peaks should have clear skies 70% of the time or more in winter [8], as good as Hawaii or northern Chile.
Can a telescope be placed on any of these mountains? Last year, one of RC's summer students, Mubdi Rahman, found from topographic maps that most but not all of these peaks are permanently covered in snow. Further study of satellite images this year by JK revealed that some of these mountains possess rocky outcrops or ledges of sufficient size to allow construction of an observatory. >From these we selected four candidate mountains on northern Ellesmere for study: two on the coast, and two higher peaks further inland. All four were visited this summer, and two of these received an autonomous site-testing station.
The station camera (far right) and communications equipment. The silver control box in the centre has a
router with two spare channels for future instruments. The cable at left plugs into a standard
meteorological station used by Environment Canada.
The station has a digital camera housed in
a heated enclosure. This offers a wide viewing angle, sufficient
to see both local ground conditions and
a large patch of sky. An automatic iris, exposure meter, and infrared cut filter will
respond to darkening light conditions in order to measure night-time cloud cover.
Basic meteorological measurements - windspeed and direction, temperature, pressure, and humidity -
are taken with a standard Environment Canada instrument suite.
And there is room for more instrumentation to be added later.
The data are all stored on a flash drive, via a Linux-based PC board. These are off-the-shelf components, and integration was in collaboration with a commercial surveillance firm in Vancouver. We have tried to learn from the failed early Antarctic efforts, and so everything is solid state - no fans or disk drives. Everything is rated to -40 C. We assume that at least once the station will freeze down to this or colder, and must come back up on its own. Limited communication with the outside world is possible through a satellite transceiver, but the only available network is Iridium, which has too low bandwidth to get all of the data out this way. We must return to retreive the full dataset.
The station framework is a 2-m-tall aluminum tripod. Solar panels are of no use during the Arctic winter, so the station is topped by a rotary wind turbine. It charges a bank of batteries in the same warm enclosure housing the electronics. Because our mountain sites are only accessible by helicopter, the station is light, under 100 kg, and very compact. The tower dissassembles and is small enough to fit within the passenger compartment of a Bell Ranger 206L - a medium-sized machine. For big loads, of say 2000 kg or more, something in the heavy-lift category would be required, such as a twin-rotor Chinook.
Logistical support for this project was provided by Natural Resources Canada through the Polar Continental Shelf Project (PCSP) which operates out of a base in Resolute Bay. Their offer of support came in March of this year, just three months before we would need to be in the field. This required quick development of the stations. In fact, the first test of the complete stations did not take place until the field party (ES, DB, and JK) converged in Resolute on 13 July. We test assembled the stations near the airstrip, just outside the PCSP hangar. It is from here that PCSP oversees aircraft logistics for dozens of scientific research groups all over the Arctic.
Undergraduate summer student Jonathan Klein poses in front of the two stations just after the completion
of testing in Resolute.
The first images sent home by the two cameras are inset. Note the bush planes and helicopter in the background.
Throughout the summer all manner of researchers pass through the PCSP base on their way to and from the field: climatologists, paleontologists, paleomagnetists, ichthyologists, and others whose specializations are even harder to spell. Many astronomers would find the Resolute base familiar. It is similar to the Hale Pohaku facility which supports the Mauna Kea Observatory. Except in Resolute, evening conversations in the mess hall are more likely to stray into the sexual lives of arctic fish. And perhaps more disturbing, the sun never sets.
Inside the Twin Otter on the way to camp. Dell Bayne looks out over our equipment, including two drums of
jet fuel for the helicopter.
The PCSP base is roughly 700 km south of our candidate sites, far too distant to reach with a
loaded helicopter. So instead we flew in the field party, gear, food,
and fuel on the "Workhorse of the North" - a deHavilland Twin Otter bushplane. This involved a somewhat
uncomfortable 4 hour flight in cramped conditions,
but the landing on the gravel bank near where we would later establish camp was surprisingly smooth, even
with a full 1000 kg load.
Our camp was a clump of bright yellow dome tents, a cookstove, and a portable outhouse. Very luxurious. We also had a radio for checking in with Resolute twice a day, and a shotgun in case a polar bear strayed into camp. Local hunters suggested we were too far north to expect any bears to visit, but even so it is best to be prepared. The radio is key. If a camp misses two radio calls in a row, the base facility will send a plane to look for them. It was also our means of communicating with the other camps, and knowing when the helicopter would arrive. We anxiously awaited the helicopter, which rendezvoused with us on 18 July. This marked the beginning of the field work.
The first task was to establish which of the candidate peaks offered a safe landing site for the helicopter. By its nature, such a spot would probably also be sufficient for setting up a station. It must not be covered by thick snow or ice, and have enough bare rock to provide a suitable footing for the tripod. On the first morning, two of us (ES and JK) flew over all four candidate peaks and quickly confirmed that none had such conditions at their true summits. Nor did the two higher peaks possess good spots lower down their flanks. But the two coastal mountains have ridges that run down to the sea, forming headlands at either edge of a large ice-locked bay. We selected high rocky outcrops at the leading edge of each of these, one at 1100 m elevation (designated Site 11A) and another at 800 m (Site 12A). These sites are within view of each other across the bay.
The helicopter flies across the bay. (Click on the image for a
video clip. This is a 4 Mb MPEG file.)
Both sites are large flat areas strewn with boulders - perfect for piling on the feet of the stations.
They are also perfect for building inukshuks, the waymarkers of the North. And in a way, these
seemed conspicuously absent. We did not really expect to find any, and a
condition of our permit to test here required that we would not disturb any, if we did. Even
so, both locations seem to demand one, as these headlands naturally point out towards the North Pole.
Station 2 up and running. On the right is Eric Steinbring, putting the final touches on
this new "inukshuk."
Joined by our weather station expert (DB), the full field party then shuttled to the two selected sites
and set up the stations over the next day.
Both were up and running by noon on 19 July, just a day after the helicopter arrived.
Working to such a tight schedule was worth the extra effort, because using the helicopter is expensive, charged
per hour, and subject to a minimum fee of $6000 per day.
It may come as a surprise to some how well established logistics are in the far North.
Weather may
delay flights in and out of either the Resolute base or the field location by a day or so.
Fog is a particular problem in the glacial fiords, and delayed flying into our camp by one day, for example.
However,
two of our party suffered worse delays at the hands of southern carriers in their attempts to return
home to Saskatoon and Toronto.
The view from Site 12A captured with the station camera. The field of view includes one
leg of the tower. The base of the satellite antenna is also visible in
the upper centre of the image. We do not expect any icing of the station due to the low humidity, but this
needs to be shown.
Now we wait.
The robotic stations will take data throughout the winter
without any intervention from us.
We know that a computer glitch has cut off data transmission
up to the satellite, but this will not affect the datataking. Next summer
we plan to return and retreive the data as well as fix the modems.
We will also prepare for new instrumentation, especially a means of measuring the seeing.
But the first winter's data should already be sufficient to show if the satellite analysis
of cloud cover is correct and skies are clear.
The view from Site 11A. This bay remains locked in ice year round. The North Pole is off to the left in this picture, across 740 km of ice.
What science can be done if it turns out that conditions are as good as we expect? The unique location - just 7 degrees from the pole - is ideal for circumpolar objects. The winter sky should be very dark - perhaps as much as 100 times darker in the near infrared than typical mid-latitude sites [4] - and dry, which suggests a whole new window might open up for near-infrared and submillimetre science. If true, the exoplanet and observational cosmology communities would be very interested, as this window would be open 24 hours a day for months on end, perfect for long-term monitoring of short-period and transitory events.
Just what sort of telescope this demands is yet to be determined, but anything of 0.5 m size and up could be quite interesting. A dark sky can make a small telescope very efficient; even moreso if the seeing is particularly good. And the first two sites we have selected are suitable for supporting much larger instruments. A design will develop as the data come in, which should allow plenty of time to consult with local communities. We recognize that this would not only be a new Canadian observatory, it would be Nunavut's first. We welcome all Canadian astronomers to join in our project as we start looking toward this.
Acknowledgements
Many individuals have helped the project get this far. We gratefully acknowledge the support of Greg Fahlman as well as guidance from Paul Hickson, Chris Pritchet, Howard Yee, Rene Racine, Tony Travouillon, and Brian Leckie. We thank Liviu Ivanescu and Mubdi Rahman for their photographic analyses. We appreciate the helpfulness of John Tarduno with the University of Rochester paleomagnetism group, whose sharing of aircraft time kept our research costs down. ES is indebted to the HIA purchasing department, whose quick action rescued some of our equipment from customs purgatory only just in time to make it to Resolute. This research was supported by funds from the Natural Sciences and Engineering Research Council, the National Research Council, and Environment Canada. Support from the Polar Continental Shelf Project is through Natural Resources Canada, and we particularily thank their staff, without whose arctic expertise this work would not have been possible.