SHARE and SuperDARN alike

2006-12-05

I've had a very busy couple of weeks since getting back from my holidays. The SHARE (Southern Hemisphere Auroral RADAR Experiment), our over-the-horizon ionospheric radar was to be the guinea pig for a new version of the radar control software and a new suite of radar control programs, and I was the lucky person that got to install and debug the lot of it. Two weeks later and it looks like we've ironed out all the problems we can detect and there's little else to do but leave it running for a while to see how it behaves over the longer term. In the middle of all this work it was announced that the radar was even changing its name to the Halley SuperDARN Radar, to better fit in with the global network of which it is a part, but so far we've not got the paint out, and I imagine no one here will ever think of it as anything but the Share.

So what is the Share? Physically speaking its a set of two cabooses housing transmission and control electronics connected up to an array of sixteen towers that loom along the horizon out behind the Piggott. The sixteen antennas, and a box of tricks called a phasing-matrix, allow us to electronically steer the direction that radio waves are transmitted, and so sweep out a wide swath of the sky. The radar blasts out short bursts of radio energy then waits and listens for echoes. By very carefully timing how long it takes for the echoes to return, and knowing that radio waves travel at the speed of light, we can locate patches of plasma in the ionosphere pretty much anywhere between Halley and the other side of Antartica.

When our packets of radio energy reflect of these areas of plasma their frequency will be changed depending on how fast the plasma is moving towards or away from us. By measuring the frequency of the pulses coming back we can get an idea of the winds affecting the ionosphere, effectively a way to work out the weather in the very highest layer of the atmosphere.

Of course, it's a bit more complicted than that. We don't just send out one pulse at once, but instead let loose a sequence of pulses, each very precisely timed, and listen for the returns from the first pulse in the sequence even before we start transmitting the later pulses. We have to be very careful doing this, though. The receiver electronics are very sensitive and are liable to fry themselves if we use the array to receive at the same time as it transmits a pulse. To avoid causing damage we switch out the receivers just before a pulse is sent out, and turn them on again a little while afterwards. Each pulse lasts only 300 microseconds, and the gaps between turning on the receiver take only 20 or 30 microseconds.

All these turnings on and turnings off are coordinated by the radar operating software running on a realtime operating system. And this is where I come in as the resident computery type. We were asked to install and test a pre-release version of the software, so had plenty of little problems along the way. The details are, well, detailed, but it was a new experience for me to have to use a digital oscilloscope to debug a piece of code. Initially nothing worked, little error lights flashed on and off, the pulse shape being generated was skewed and the protection circuits weren't going to turn on and off in the right places. (We had remembered to unplug the transmitting kit, just to be on the safe side.) Anyway, a week later having probed and tickled all the different bits of the radar and made lots of little changes along the way we were able to turn it on and record some real data again.

While Chris and I were working hard tucked away in our little containers, life on the rest of the station grew a little more intense. The end of November is the time of the year when we clean out the tank which melts the snow we drink and wash with. For the first time since leaving the ship we were asked, ordered even, to spend more than a fleeting minute in the shower. All the water in the tank had to be removed, and it might as well be flushed away to good cause. It was great having a good long wash, but hard fighting away the small guilty feeling that grew stronger the longer I stood in the shower.

The cleaning went well and we quickly refilled the tank by dragging great boxes of water from the Drewry, but disaster struck as the pipes from the tank, 30m under the ice, to the Laws, 8m above the surface, had frozen. Only a little disheartened the technical services crew got down to work and, after a few hard days down the tunnels, restored water supplies. In the interim we had to supply the main building with water from the summer building, which was a bit of a faff, and were prohibited from washing clothes or wasting water. I think, in the end, that I enjoyed the first short shower after the shutdown more than the luxurious stream of the long one beforehand.

It being nearly the summer, it's been incredibly warm. Once or twice we broke through into positive figures. Everywhere the accumulated snow of the winter is melting away. Under the platforms there's a constant drip, drip from the girders. The snow is getting sticky and skis need waxing, or tracks cutting, to avoid tripping over as great lumps of snow stick to their bottoms otherwise. Most importantly, so long as it's not windy its warm enough to wear shorts again, to put my chin out in the open, to feel a little more engaged with the elements, more involved in the environment. For the next bright month or two the weather shouldn't be much worse than many of you will be experiencing back in the UK, or out on the ski slopes.

Share cabooses. Control on left, Tx on right

Tx trays. One for each antenna.

Cabling on the phasing matrix

Icicles drooping outside my office

Sixteen towers

Forest of elements

Melting in the heat

Finally, as a special treat, I've worked up a Google Earth animation of my shiny new radar (save the link to your desktop, then drag into Google Earth). You should be able to see ionospheric echoes moving over the Antarctic continent earlier this morning. The brighter the blob, the stronger the echo. Each frame is from a two minute scan of the sky. You'll need a fairly recent install of Google Earth. I shall need a new head after all the scratching involved in rotating coordinate systems back and forward.