Scientists Fight Radio Interference to Forecast Drought
Soil-moisture satellite encounters rogue transmitters; new satellite being designed to dodge them
18 September 2012—For scientists who want to create an early-warning system for drought, satellite data is gold. So it was with much anticipation that the European Space Agency (ESA), in November 2009, launched its first-ever Soil Moisture and Ocean Salinity (SMOS) satellite for measuring global soil-moisture levels, a crucial gauge for drought monitoring and prediction.
But as soon as operators in France flipped on SMOS, trouble hit. The radiometer picked up bright, high-power, man-made signals that weren’t supposed to be there. “We see signals that are much stronger than what we would expect from natural emissions of the Earth’s surface,” says Susanne Mecklenburg, the SMOS mission manager. “They drown the data we’re trying to measure.”
The SMOS radiometer looks at areas approximately 50 by 50 kilometers, detecting minute microwave emissions from land and water to construct global maps of soil moisture and ocean salinity. The signal frequencies fall between 1400 and 1427 megahertz. This so-called L-band is protected, set aside by the International Telecommunication Union solely for radio astronomy and Earth observation. But in practice, the SMOS team found, there’s illegal interference from radar systems and TV and radio transmitters. “They can corrupt entire 50- by 50-kilometer snapshots, making them unusable for scientific applications,” Mecklenburg says.
The result is a global soil-moisture map that might have blank, dataless patches. NASA’s Aquarius satellite, launched last year to measure ocean salinity, faces similar woes. Plus, in 2014 NASA plans to launch its own soil-moisture satellite, which could also be plagued by illegal L-band interference.
Both the American and European space agencies are making headway in the fight against radio interference, but the battle might not be easily won.
ESA’s combat tactic has been straightforward: Go to national authorities and ask them to get the radiation sources turned off. SMOS images show that China, India, and some areas in the Middle East, eastern and southern Europe, and the northern United States are radiation hot spots, Mecklenburg says. Spain, for instance, was riddled with interference sources at the start of the mission. The country is one of SMOS’s funders, and at the ESA’s behest, Spanish authorities were able to clean up much of the interference. “We’re now able to produce very good maps over Spain, and the data has been used for drought monitoring,” she says.
ESA’s direct approach won’t work everywhere. Some countries simply might not enforce frequency regulations. “There are military radars we’ll never get a handle on,” Mecklenburg says. “We will never get all [interfering] sources to switch off. Some switch off, but others might come on.”
And then there are the unintentional culprits. Airport and weather radar, for instance, emit between 1200 and 1300 MHz, but some of their energy can leak into the 1400 edge of the L-band. “This is a tricky area to address,” says Simon Yueh, a Jet Propulsion Laboratory engineer on the Aquarius mission. “They’re probably not illegal, because the main frequency is outside of the L-band. There needs to be consensus on how to manage those, and stricter regulations.”
(A similar issue helped kill off LightSquared’s plan for a terrestrial L-band cellular network, because GPS receivers unintentionally picked up the cellular signals, even though the cellular signals were not in the GPS band.)
The best way to battle radio interference may be with technology. NASA’s Soil Moisture Active Passive (SMAP) satellite, scheduled for launch at the end of 2014, will carry smart receivers that detect and discard unwanted radio signals. The microwave emissions from land and water fluctuate. A conventional radiometer deals with this by measuring signal power across a wide bandwidth and integrating it over a long time interval to get an average.
By contrast, the SMAP radiometer will use time intervals one-fiftieth as long and divide the 27-MHz-wide L-band into 1.5- MHz portions. “A radar transmits a pulse for a short period of time, [and] if I integrate over a long period of time, it’s not easy to tell the radar was there,” says Joel Johnson, an Ohio State University electrical engineering professor who helped design the new radiometer. “Instead, if you chop up one large time interval into many little intervals, you can see the radar pulse in one of those time intervals, throw it away, and keep the rest.” Chopping up the band has the same effect, in that the system can throw out sub-bands containing man-made interference.
The more sophisticated receiver comes at a cost, says Jeffrey Piepmeier, the lead hardware engineer working on the radiometer. Apart from the added expense of developing, building, and testing it, the instrument is also bigger and bulkier, which adds to the cost of heaving it into orbit. And the technique has a weak spot: It may not work against all possible interference sources. It wouldn’t work, for instance, with emissions that look similar to the small natural signals the radiometer is trying to measure, such as those from transmitters that use spread-spectrum communications schemes, Johnson points out.
The ESA, meanwhile, can’t make any improvements to the SMOS radiometer hardware. But researchers are developing intelligent data-processing algorithms to detect and eliminate interference, Mecklenburg says. And the agency is intent on laying groundwork to enforce regulation. They have so far gotten about 200 interfering sources turned off around the world—“a huge success,” she says. “We’ve gotten a long way in two and a half years, but the problem will persist.”