It took a long time to get as far as this success, beginning over a year previously. Under considerable time pressure, a group of astrophysicists led by Jan Stenflo designed, built and tested their own instrument for this purpose. Then SoFiE, as the experiment (SonnenFinsternis-Experiment – solar eclipse experiment) was now called, was packed in crates and sent off, with its constructors, to the desert. (1) In Tripoli, the SoFiE consignment was only handed over in return for an extorted “ransom”, but this was in fact the team’s only unpleasant experience with the otherwise friendly Libyans. The onward journey to the tented encampment in the heart of the Sahara was delayed because a sandstorm had destroyed part of the infrastructure there, but in the end a military transport aircraft and several helicopters took the ETH physicists, together with scientists from all over the world, to “Eclipse City”.

Scientists from all over the world are flown into the desert in the belly of a Libyan military transport aircraft. large

The team spent the week in this camp, in one of the remotest spots in the world, anchoring the telescope as securely as possible in the ground, installing the power supply, electronics and computers, carrying out calibration and test measurements, practising the timing of the steps to be taken when it came to the real thing, and also taking an active part in the international symposium that Stenflo had organised with Osama Shalabiea from the University of Sebha, which was taking place in the same location. There was barely enough time to enjoy the perfectly clear view of the starry sky before the short period they had for sleep. The infrastructure provided by the University of Sebha was perfect in almost every way, yet the team would have been cold in the SoFiE tent in the mornings without their thick windjammers, whereas during the day temperatures could soar to 40 degrees. Minor sandstorms from time to time covered all their equipment with a layer of dust. What was the point of it all?

From a segment to a crescent to darkness: the middle picture shows the flash phase, lasting just ten seconds. By recording it in high time resolution, the ETH astrophysicists now hope to learn more about the chromosphere. large

Scalpel cut through the surface of the sun

The sun does not have a solid surface like the earth. The incandescent gas of its atmosphere simply becomes denser towards the centre, and therefore more opaque. However, the transition from “almost transparent” to “opaque” takes place in a relatively thin layer about 400 km thick, the so-called photosphere. This is where most of the sun’s light that we see comes from. For solar physics, the layer immediately above this, the richly structured chromosphere, is particularly interesting. Its weak shine is normally greatly exceeded by the photosphere.

Branded by the desert: the SoFiE team in front of their work tent. From left to right: Alex Feller, Renzo Ramelli, the SoFiE instrument, Daniel Gisler and Jan Stenflo. In the background to the right you can also see one of the straw huts in which the physicists lived. large

Immediately before a total solar eclipse, however, there is a rare opportunity: the moon is already completely covering the photosphere, but a thin crescent of the chromosphere can still be seen. It is only about ten seconds – the so-called flash phase – before the edge of the moon covers this last sliver too. If you “film” the crescent during this time using high time resolution measurement, then by using high spatial resolution you can learn about the layers within the chromosphere.

It is not practicable simply to simulate the coverage by the moon with a screen in the telescope. For one thing, it is virtually impossible to keep a set-up like that free of the diffused light from the far brighter photosphere. Furthermore, the image of the sun is blurred and distorted by turbulence and pressure gradients on its way through the earth’s atmosphere, which makes it impossible to position the screen accurately – the moon, on the other hand, is seen against the still undistorted light waves in the vacuum.

New data for solar physics

The aim of SoFiE is to extract as much information as possible from these ten seconds. To this end, the light is split into a spectrum (“rainbow”). In addition, a Savart plate divides the spectrum into two components, polarised in parallel and vertically to the rim of the sun. In this way it is possible to measure both the intensity and the polarisation of the light in relation to wave length (“colour”). The composition of the sun’s atmosphere is revealed in the countless absorption and emission lines which can be made out in the spectrum. Their polarisation signatures provide valuable information about the thermodynamic and magnetic conditions there. This is why spectropolarimetry has long been a useful tool in solar physics; until now, however, no-one had attempted it during a flash.

Mission successful

Even though the ETH astrophysicists had had no opportunity really to simulate the flash in advance, and so to test whether SoFiE would work properly, when it came to the real thing, everything went like clockwork: the flash spectrum was indeed measured, and the results exceeded all expectations. The next few months will show what new discoveries for physics can be made from the data. In fact, SoFiE’s world tour is not yet over. The next useful total solar eclipse will be in 2008 in Siberia and Mongolia, and it is expected that SoFiE will be there too.