The video is a time-laps of the sky during one day, although it was almost certainly taken at night (i.e. winter) with the camera rigged to provide high enough sensitivity such that it looks like it's day out. So you're actually looking at the moon. Notice that the moon drifts along the horizon rather than away/towards it as we're used to.
|Measuring the Earth's tilt at the equinox; SPT in background|
I took a break from work today to measure the tilt of the Earth for fun. This tool is a sextant, designed to let the user measure the inclination of an object with respect to the horizon. I measured that the sun is 23.4 degrees above the horizon, thus the Earth's rotational axis is at a 23.4 degree tilt with respect to it's orbital plane.
This technique was used by both Admussen and Scott to navigate to the pole 100 years ago. Remember that a magnetic compass is useless near the magnetic pole of the Earth, so you're stuck using a sextant and celestial bodies. If you're not standing at the pole, the sun's elevation will bob up and down over the course of a day, so repeated measurements can help guide you to the Southern-most point on the globe. In fact, when Admussen reached the pole, he stayed for three days to conduct repeated measurements of this sort to definitively prove he had made it. An experienced user can determine their latitude to within a mile with this tool.
There's another treat about halfway into the above video, when you can see the Aurora Australis (Southern Lights). These are the same phenomenon as the Northern Lights, arising when charged particles from the Sun are directed along magnetic field lines to the polar regions, where they collide with gas in the upper atmosphere, causing a glow. I won't get to see this in person (thank god!)
One final cute thing in that video: looking at the beginning, notice that the stars do not change there orientation with respect to horizon. That's not the case for mid-latitudes where the constellations "rotate", as seen in this youtube video:
This actually presents a mild-challenge to our experiments since the mechanics of our telescopes only let us scan the sky parallel to the horizon (in azimuth, as we say). This can give rise to stripy looking maps that concentrate certain noise in just on direction. Mid-latitude telescopes essentially probe more spatial modes on the sky thanks to that rotation.