Some 20,000km above the Earth, 30 GPS satellites are beaming down signals, helping thousands of sales reps figure out how far it is to the next motorway services.

But these signals could have other uses. UK researchers are developing equipment which could use GPS signals to predict storms at sea, determine the advance and retreat of ice-shelves, and refine models of climate change.

According to Martin Unwin of Surrey Satellite Technology, the key to the breakthrough is the way that GPS signals are reflected from the Earth. Using the UK-DMC imaging satellite, Unwin and a team from the Surrey Space Centre recently proved that reflected signals could be picked up from sea, land and ice, and that information about the characteristics of the reflecting surface could be derived from the signal.

Several important discoveries have already been made by studying these reflected signals. These have generally used radar altimetry, where the satellite sends the radio signal and receives its reflection, to measure the height of the surface.

For example, Unwin said, the TOPEX/Poseidon satellite gave NASA researchers accurate measurements of the height of the sea surface, which allowed them to track the El Nino current. 'But the problem with altimetry satellites is that they're big and power hungry, because they have transmitters on them, and they point straight down so they cover quite a small area — a few tens of kilometres,' said Unwin.

'They trace out a track over time, but only repeat on the same area every 20 days or so. There are big gaps in the measurements, so if you're trying to measure sea eddies and other features, you miss information.'

There are several advantages of using the signal from GPS (or, more properly, GNSS — global navigation satellite system — of which GPS is one example) rather than radar altimetry satellites. ESA hopes to launch its own GNSS, Galileo, in 2010.

First, the satellites are already in orbit and sending out their signals, so there's no need for extra transmitters. The reflectometry satellites only need to be equipped with receivers, therefore they use less power and can be quite small, 'so you could potentially put several satellites up,' said Unwin.

Moreover, there are many GNSS satellites, allowing a multiplicity of reflected signals to be received and potentially much more complete coverage.

But of course, it isn't that simple. The reflected signals are weak, with interference from the ionosphere, making them much more difficult to interpret than the very clean, strong signals that radar altimetry produces. Unlike TOPEX/Poseidon, the reflectometry system can't directly measure the height of the sea. 'So instead of measuring the absolute altitude, we're measuring how much scatter there is,' said Unwin.

As an example, he suggested, imagine looking at the reflection of the sun off the sea from an aircraft. 'If there's no wind and the ocean is mirror-like, you'd get a very strong, direct image. But when the ocean gets rough, the reflection is splurged out into a large ellipse.' Exactly the same happens with the reflected GNSS signals, he said. and analysing how much the signal is spread, and the delays and Doppler shifts in the reflected signals, will give Unwin's team information on what the surface of the sea is like — and that will give them valuable information about wind speed and direction.

'What we think we can measure more accurately than anyone else is something called the mean square slope, which is the average angle of the ocean surface. The angle increases with the speed of the wind; it's a complex relationship. But if you have a high roughness, you can warn of storms.'

Information on sea conditions is also valuable for climate scientists, as is other information gleaned from GNSS reflectometry. Reflections from land are affected by the level of moisture in the soil, said Unwin. The technique is also very good at detecting the boundaries between water and ice, and even between different sorts of ice on the surface; the system could find uses in tracking the advance and retreat of polar ice sheets.

In the current project, Unwin and his team are developing a new receiver which can process the signals on board the satellite. 'At the moment we take 20 seconds of data, store it on board and download it to our ground station where we process it all night.

'If we process on-board, we'll be able to collect signals continuously.' The system will be able to decide where to look for reflected signals, and divide the signal into separate domains for direct reflections, delayed signal and Doppler-shifted signals. 'We'll need to get all the processing right first go,' said Unwin.

Unwin aims to complete a prototype detector within two years, and hopes to mount it on one of SST's satellites. 'The aim would be a constellation of small satellites, once we find the application that will support this,' he said. 'We need customers, whether they're commercial, scientific or both. But I'm sure this will be a powerful tool.'