Text and pictures © 1993-2023 Guillaume Dargaud
Last updated on 2021/11/05
"The land looks like a fairytale." — Roald Amundsen about Antarctica.
Antarctica is one of the places on Earth where the sky is a constant source of amazement. Cold temperature, low humidity, lack of variation on the terrain, proximity of the magnetic and geographic poles all contribute to displays rarely seen at other latitudes. Add to this the fact that there is little else to do than watch the sky (and work)...
Above: Midwinter sun at DdU
Although DdU is a tiny bit south of the Antarctic circle, the sun is always visible at noon, even at the darkest of the winter. The reason is because of the refraction of light on the low atmosphere layers: you can actually see it while it's somewhat below the horizon. The picture above was taken on June 21st 1993 with 9 exposures set 11 minutes apart if I remember right. One hour of sunlight is all we get in winter. Enough to go for a walk, but little else. I created my animated HR bar starting from this picture.
So why is the sun visible at all since DdU is just below the polar circle ? That's because the low angle bends the light rays in a way that although the sun really is below the horizon, it is still visible. This phenomenon makes the setting sun appear somewhat flat instead of round (this is true at any latitude).
Above: Midnight sun at DdU
This picture is the exact opposite of the previous one, taken on December 21st 1993, the longest day of the year. So why does the sun disappear at midnight since the station is almost exactly on the polar circle ? It's still above the horizon, but below the mass of the continent dominating the south view (compare with the visible sea). 8 exposures set about 25 minutes apart.
After years of attemps, I managed to take an extraordinary picture by following the sun during 24 hours on the longest day of the year.
Right: Mirage of an iceberg
Yes, this is a genuine picture of a real mirage. In Antarctica they happen quite often, each time there is absolutely no wind (well, okay, that's rare enough around windy DdU). The atmospheric turbulence needs to be really low, the air really pure (not a problem there) and dry. We saw dozens of them. The stripes in the foreground are alternating slabs of ice, snow-covered ice and sea water; the picture was taken in autumn at a time when only parts of the sea are frozen.
Left: Another mirage effect, here limited to the deformation of the guy lines holding the tower upright: they should obviously be straight. The white pools in the back are actually reflections of the sky on a very low inversion layer. And the horizon is much higher than it should, up to the point that we get an impression of vertigo, of being inside a large hole.
People sometimes tell me that it's impossible to take a picture of a mirage; not true: if you can see something, you can take a picture of it (except for visual hallucinations). Well, here the quality's not so good because I used a doubler on a 400mm lens, and this is an enlargement of the center of the resulting picture since the mirage was quite far away and thus very small.
Left: Peary's arcs, aka sundog at Dome C.
Right: Halo and parhely geometry.
Parhelies are an extension of the more common halo circles. Halos are a group of optical phenomenon due to the reflection or refraction of solar light on ice crystals suspended in the atmosphere (cirrus clouds, thin snow, icy fog, blown snow). The basic halos take the form of circles around the sun, at 22° (more common and brighter) and 46°. Red inside, purple outside with other rainbow colors in between. They can also happen with the moon. Halo circles are often visible in winter when there are some thin high clouds. The crystals forming the halos are usually high up in the atmosphere, but if the temperature is low enough they can be next to the ground, like in Dome C where I could actually see the individual ice crystals blinking on and off in the halo as they fell slowly to the ground on a windless day. More information about halos at Space.com's Fact vs. Fiction: Reading Weather in the Sun, Moon and Stars.
Left: 22 and 46° (barely visible) parhely circles, as well as sundogs and horizontal parenthely circle.
Right: On this straightened 180° fisheye image, the Lowitz arcs are clearly visible.
Left: Sundogs and Lowitz arcs behind the 'Xmas tree' which I used to hide the sun and lower the backlighting brightness.
The parhely circle is a horizontal circle at the same height than the sun, barely visible here. There may be brighter points on it, at the intersections with the halos called parhely, at 120° called parenthely and at the opposite of the sun called anthely (rare). When those points are bright, they are often called "fake suns" or "sun dogs" and they are visible here.
Arcs going from parhely points to the small halo are called Lowitz arcs or Peary's arcs. I had never thought them possible in summer until I saw two in Rome in June on two consecutive evenings.
Right: A rare anthely circle behind the Dome C summer camp. Notice the shadow of a radio tower pointing towards the center of the circle.
The visibility of a halo depends on the shape of the crystals, but also of their quality, for instance if they are too small (<10 microns) or contain air bubbles, the light gets diffused instead of being diffracted. The best halos are produced by crystals above 0.1mm, although the largest ones also rarely have good optical quality, explaining why a good halo is not so common although high altitude crystals are very common in cirrus clouds.
Above: 360° panorama of Concordia (and self-portrait) with a halo circle around the sun.
Above: A parhely circle around the moon, often nicknamed 'moondogs'.
Additional bibliography: Atmospheric Halos by Walter Tape and Rainbows, Halos, and Glories by Robert Greenler. There's also a halo raytracing simulator you can download. Fun to play with, but doesn't beat seeing the real thing.
Left: Brocken spectrum seen from a mountaintop in Ecuador.
Yet another colorful and rare phenomenon, the Brocken spectrum (also know as a glory), here a picture taken in central Italy. It is sometimes visible from mountaintops and looks like a circle of rainbow colors around one's shadow projected on the haze below. The sky above needs to be clear and usually some wind is needed to form the haze just on the downwind side of the mountain. Up until 2003 no really clear explanation had yet been given, but a deeper mathematical analysis revealed that the main cause is a tunneling effect of incoming light on the surface of the water droplets, an effect which is not necessary in order to explain rainbows for instance. In other words, a part of the light from the sun which passes very close to a droplet actually goes in the droplet thanks to a tunnelling effect, and is sent back in the direction it came from via usual optical phenomena, including wavelength separation of colors, hence the aspect of multiple concentric rainbows. Or in yet other words, it is a direct and beautiful macroscopic view of a microscopic quantum effect.
Right: Brocken spectrums are more commonly seen from airplanes when the shadow of the aircraft is projected on the clouds below, here when flying above the Transantarctic range.
Right: The green flash of the sun
It took me 12 years after first seeing it, but I finally managed to take a picture of the infamous and elusive green ray (aka green flash) of the sun, which I've seen a few times in winter. It's visible on the sea as well as on the ice but doesn't turn out too well in picture. I tried several times and failed because I wasn't fast enough (it can last only a split second), because of turbulence (exposure time is then too slow even though it's still very bright), lack of patience (the sun takes a long time to set in Antarctica), technical difficulties (frozen lens) and what else...
The green ray happens fairly often in Antarctica because the horizon is uncluttered (either frozen sea or continental plateau, both totally flat); the sun's trajectory is almost horizontal so the sunset lasts a long time (several days if you happen to be right at the pole); when there's no wind (granted that's rare) the air is truly calm over extended regions as there is no horizontal thermal gradient to create turbulence. Once I could make the green color appear and disappear by just lowering or raising my head; it lasted 2 or 3 minutes. I attempted to take it in picture with a 800mm a few times when the conditions seemed right; but when I was ready and waiting the phenomenon didn't happen. Finally, on march 18th 2005 at Dome C I managed to take a picture with a 1250mm Celestron telescope and a Nikon D70 attached to it. The turbulence was awful, but then the next day conditions were even better and I was ready. That's the picture you see on the right. I have a whole series spanning a few minutes.
Left: A polar stratospheric cloud, a rare and impressive sight.
Also known as PSC clouds, noctiluminescent clouds, noctulent clouds, noctilucent clouds, NLC, NNL clouds, mother-of-pearl clouds or nacreous clouds, the polar stratospheric clouds are very high clouds brighter than the full moon and clearly visible at night or when the sun is low. Those clouds form and disappear rapidly when water condensation occurs at high altitude, often caused by the presence of some chemicals like nitric acid (HNO3) or sulfuric acid. They form in stable dark stratosphere when the air temperature falls below -80°C, conditions which happen usually during winters at the poles but sometimes also after volcanic eruptions at milder latitudes.
Those clouds are too high (20~80 km high, much above cirrus clouds and barely below the auroras) for weather balloons to probe and too low for satellite analysis. When seen from space they are called PMC (Polar Mesospheric Clouds). The recent increase in the number of observations of PSC clouds is strongly suspected to be caused by increased global air pollution.
One theory as to their formation is that they may result from meteor showers ionizing the upper atmosphere.
Right: Iridescent clouds can often be seen when the angle with the sun is low on the side of thin clouds. It's better to hide the sun for better viewing.
Left: Iridescent clouds behind people working on the mast to install new satellite calibration systems.
Right: The Sun Pillar, a common weather phenomenon
The sun pillar, visible as a column of light when the sun is under the horizon, is caused by multiple reflexions between flat ice crystals, all falling with the same orientation. More information about sun mysteries at Space.com's Sun pillars and other curious effects
Left: The infamous katabatic wind.
Fastest sustained wind on the planet, the katabatic wind can be a true danger. Only tornadoes are faster but they don't last very long. This wind can lift you right off the ground and also last for days. It's not a wind caused by storms or clouds, but by gravity slowly accelerating cold air along the long smooth slopes of the continent until it reaches record speeds near the shores, sometimes augmented by funneling effects of valleys, like at the old Port-Martin base where a world record wind of 325km/h wind has been recorded. If there is some recently fallen snow on the continent, it gets carried by the wind into a total whiteout of wind and snow fury that get inside any entrance: your collar or a slit on the side of a building window. After it has been blowing for several days and all the snow's been carried away, you can get this crazy wind with a clear blue sky like this picture shows. Since the wind always comes from the continent, it doesn't have time to form huge waves... usually. Look here to see what happens when it can form waves.
The wind carries dry snow that can build up electric charges on people outside if your footwear doesn't evacuate the charge. We used to get 'zapped' all the time when touching doorknobs or any metal.
Sometimes things get even weirder. The Loewe phenomenon is a katabatic flow jump, or in simpler words a kilometer high wall of snow and wind fury that can form when the katabatic wind jumps over a sharp drop. 'Sharp drop' being all relative, a slight angle change in the slope is enough to trigger it under the right conditions (laminar flux going turbulent, high wind speed...). I have a satellite pictures of it.
There are other effects of the cold that are not visible. First is the silence of the high Antarctic plateau: turbulence is very low and the wind very laminar. The little turbulence present is limited to heights close to the ground (below 30 meters) where the wind forms some small sastrugi. The silence is incredible... until you start walking. The packed and very cold snow contains lots of air near the surface and amplifies footsteps like a drum.
The atmosphere can play strange effects with the sound due to its thermal structure: the ground is very cold and so is the atmosphere in direct contact with it. The temperature increases as you go higher above the ground, before coming down again above about 600 meters. The speed of sound is higher is warm air than cold air even if that seems counterintuitive (cold air is denser so one would expect the sound to be faster like in denser materials such as water; that is not the case). And when a wave (be it optical or acoustic) goes from a low to a high density substrate, it deviates in the direction it's coming from (that's the classic prism effect). Going from high speed to low speed, if the angle is low enough (low incidence), the wave cannot come out; it's what happen in a swimming pool when you are under the surface and look at the surface: it reflects light like a mirror.
A very small difference in density can be enough if the angle is small. This has other practical application, for instance in the ocean the pressure of the water increases with the depth, and so does the sound speed. But at the same time the temperatures and the salinity vary with the depth, resulting in a maximum sound speed at a depth of about 300 to 1000 meters. This layer is called the thermocline. There are two different kind of customers for this piece of info: submarines and whales. Submariners hide their ships under the thermocline and this way are invisible to surface sonars: the sound coming from the sonars is deviated and even reflected by the thermocline and cannot go under it. On the other hand, sounds emitted while inside the thermocline carry horizontally over very long distances and some species of whales use this to communicate remotely, sometimes over 1000 kilometers. The power decrease with the distance is 1/r instead of the usual 1/r2 in a normal 3D medium.
Back to Antarctica where the low layer of cold air acts this way with sound. Simply put, the sound cannot go up and is constrained to travel close to the ground, resulting in very long distances. Once in 2005 I could distinctly hear Claire talking normally about one and a half kilometer away, and backing beepers from construction vehicles can be often heard several kms away. This same effect can happen virtually everywhere you have flat ground and stable cold air, like in the early morning on a lake or plain.
Note: yes, I know, the following pictures have nothing to do with Antarctica, but they are example of other interesting 'sky' pictures.
Left: Star rotation shown by a long pause.
This is a picture of the night sky taken with a long 40 minutes exposure on a tripod with a wide angle lens (20mm). OK, anyone with a basic understanding of astronomy can tell that this picture wasn't taken in Antarctica. Why ? Because the stars do an apparent rotation around the celestial pole, and the elevation of this pole is nothing else than your latitude; so if you are at the pole the stars rotate around a point above your head, and if you are at the equator they rotate around a point that is away on the horizon. In this case, the picture was taken from about 38° latitude north. in the Utah desert, with the camera pointed directly at the polar star (which Jenny found for me). Notice the different colors of the stars ? Star colors are very hard to see with the naked eye, but a camera has no problem revealing them.
Right: A shooting star and a passing satellite caught by a long exposure photography.
Enlargement of the right side of the previous picture to show some interesting details: you can see a shooting star on the right and a passing satellite leaving a diagonal trail of dots (barely visible). The shooting star apparently disappeared before hitting the ground, as is very common. The dots on the satellite trail comes from the fact that the satellite rotates on itself while reflecting the light of the sun. This non-Antarctic picture ended up on this page 'cause I had nowhere else to put it and it's an interesting complement to other sky pictures.
Right: Classic, but still very nice, bright rainbow on very dark background above the Utah desert, 2003.
Right: Full double rainbow above the Utah desert, 2003.