NATURE|Vol 450|6 December 2007 EARTH MONITORING NEWS FEATURE

overlap between monitoring systems is another. in providing more information. the same time that NASA had been planning Frequently the problem is international; one José Achache, secretariat of the Group on the Hydrosphere State (Hydros) mission for country launches a spacecraft that partially Earth Observations in Geneva, Switzerland, is soil moisture — which has since been put on duplicates what another mission is already not so sure that duplication is a good way for- indefinite hold — and the Aquarius mission for doing. International steering committees are ward. “Essentially the agencies were in unofficial salinity. supposed to cut down on the overlap, but it competition,” he says. An international steering Sometimes, though, the US–European doesn’t always work that way. “Our hope is group, the Committee on Earth Observation competition can work in science’s favour. not just to fill gaps, but to avoid duplication Satellites, exists to try to cut down on dupli- With SeaWiFS possibly close to dying, NASA of effort,” says Helen Wood, a senior adviser cation for satellite-based systems, but some- is looking at how it can jump in on the Euro- to NOAA’s satellite and information services times national interests win out. The European pean MERIS instrument, aboard , to division in Silver Spring, Maryland. Space Agency, for instance, is planning a Soil get ocean-colour data, says Paula Bontempi Moisture and Ocean Salinity mission — meas- of NASA headquarters in Washington DC. Poles apart uring two of the essential climate variables — at Although the data may not be all that the In 2003, NASA launched its ICESat mis- scientists wish they were, they will be bet- sion mainly to study the ice sheets ter than nothing once SeaWiFS gives out. of Greenland and Antarctica; in And in the long run Europe plans to have 2005, the an instrument as good as or better than launched its CryoSat, which was MERIS as part of the Sentinel 3 series A. MARTIN to have done much the same of operational climate-monitoring thing (although it would also satellites. Duplication of efforts is have measured sea-ice thickness). undeniably wasteful in the Earth- CryoSat failed on launch, so now monitoring world. But relying plans are under way to send up on any single nation, even the a second version in 2009. Mean- richest and most technically while, NASA — which likes its advanced, would risk doom- mission so much, despite a prema- ing the planet to an endlessly ture laser failure, that it renamed repeated history of research one of its streets at the Goddard satellites operating long after center as ICESat Road — is look- their intended lifespans, last- ing at launching an ICESat-II. minute scrambles to keep Waleed Abdalati, who is ICESat’s things going, and possibly cata- programme scientist, says that strophic gaps. ■ so many changes are expected in Alexandra Witze is Nature’s chief the Arctic that both the US and of correspondents for America. European missions will be useful See Editorial, page 761. The crucial measurement

he “Carbon Club” began meeting on of a US$300-million-or-so satellite called the think they know roughly where it is going in,” Fridays about a decade ago, setting Orbiting Carbon Observatory (OCO). says Ross Salawitch, an atmospheric chemist at up shop in whatever spare meeting When the clock starts ticking on the Kyoto the University of Maryland in College Park, a Tplaces it could find at the Jet Propul- Protocol’s five-year commitment in January member of the OCO team and another Carbon sion Laboratory in Pasadena, California. Its 2008, developed nations that have ratified the Club veteran. “Others will tell you emphati- members, a handful of scientists with exten- treaty will be bound to a strict bookkeeping cally that land is taking up the carbon. There’s sive experience in of Earth’s system for greenhouse-gas emissions. They will nowhere close to a unanimous opinion.” atmosphere, set about brainstorming ways receive credits for mopping up their emissions As it orbits Earth, OCO will measure the to provide one of the most crucial data sets with so-called carbon ‘sinks’, such as through ‘fingerprint’ that CO2 leaves in the air between of the twenty-first century: precise measure- reforestation efforts and improved agriculture the satellite and Earth’s surface almost half a ments of carbon dioxide levels in the atmos- and grazing practices. Yet it is currently impos- million times a day. The resulting map of CO2 phere on a fine enough scale to definitively sible to pinpoint where the gases originate — concentrations will then be used, with other track the gases’ sources and sinks. “No one and no one really knows where they end up. data and modelling, to work out where CO2 is was crazy enough to say that they could do Half the CO2 pumped into the atmosphere by being emitted and absorbed. “It’s the most dif- it until we came up with a possible solution,” burning fossil fuels ends up in the or ficult atmospheric trace-gas measurement that’s says Charles Miller of the Jet Propulsion Labo- absorbed by plants on land — but how much ever been made from space,” Miller says. ratory, and one of the original Carbon Club goes each way, and precisely where, is still If OCO’s two-year mission is a success, members. That solution is due to reach orbit unclear. “Certain people will tell you emphati- it could well serve as a model for an opera- late next year or early the year after in the form cally that it’s going into the oceans, and they tional mission that might be tied directly to a

785 NATURE|Vol 450|6 December 2007 EARTH MONITORING NEWS FEATURE

post-Kyoto regulatory system. But design, is cloud cover. The instrument the team is keen not to offer opera- has a very small viewing window tional data too early. “A prototype because a thin column of air is less

always produces challenges,” says likely to be beset by clouds than is a OCO/JPL/NASA David Crisp, OCO principal investi- broader swath. This should mean that gator and a senior research scientist the instrument gets enough data to do at the Jet Propulsion Laboratory. its job — but places that are frequently Today, a network of ground-based cloudy, such as the Amazon rainfor- stations strung across the globe meas- est, could still prove troublesome, says ures CO2 and other greenhouse gases Pieter Tans, a senior scientist with the at Earth’s surface with high precision, National Oceanic and Atmospheric but patchy coverage (see page 789). Administration in Boulder, Colo- Large expanses of Earth, including rado, who provides the ground-based Africa, India, Siberia and much of measurements that will be used as a South America, have very few, if any, validation standard for OCO. Aero- monitoring stations — even North Do look down: the Orbiting Carbon Observatory could provide precise sols, too, are a potential problem, says America, which hosts the high- data about the origin of carbon emissions. Moore. But he’s still excited about the est concentration of measurement mission: “It is potentially going to be stations, has significant gaps in its coverage has gigantic implications for the carbon sinks a huge breakthrough on this source/sink — the Yukon, for example, and large chunks that we would infer,” Salawitch says. “Our whole problem, and it will be a terrific pathfinder.” of Quebec and the US southwest. What’s more, science is driven by small spatial gradients in the network was specifically designed to avoid the gas.” Dynamic duo picking up the fluxes that OCO is interested But the measurements face various problems. A Japanese satellite named Greenhouse gases in. “The network was actually sited as far away For instance, they rely on sunlight, so they can’t Observing Satellite, or GOSAT, is scheduled for from known sources and sinks of CO2 as pos- be made at night, or during polar winters. This launch in August 2008, and will complement sible so that we could get good, clean, average is a “fundamental shortcoming” says Berrien OCO. GOSAT will measure methane, water measurements,” says Crisp. Moore, a mathematician at the University of and ozone as well as CO2. Whereas OCO uses New Hampshire in Durham, who studies the a spectrometer based on diffraction gratings, Bounce back carbon cycle. The OCO team has argued that which achieves a high signal-to-noise ratio and OCO came into being in 2001, when NASA this is not terribly problematic (C. E. Miller et thus a precise determination of levels, GOSAT set up a competition for low-cost Earth-sci- al. J. Geophys. Res. 112, D10314; 2007). A bigger will obtain its measurements with a spectrom- ence missions. Thirty-three proposals went in: concern for them, reflected in the instrument’s eter that operates at both short and long infra- OCO came out. Its instrument works by meas- red wavelengths. Long wavelengths allow it uring visible and near-infrared sunlight that is to measure emissions even when there is no reflected back from Earth’s surface — sunlight sunlight, avoiding the issues associated with that has travelled through the atmosphere night-time or polar winters. And whereas OCO twice, once going down and once returning will make spatially contiguous measurements up. As sunlight shines down and is reflected along a narrow field of view (10 kilometres) back, various molecules absorb some of it at over a 16-day cycle, GOSAT will measure iso- distinctive wavelengths. By comparing the lated footprints of the gases over a broad (up to different bands associated with CO2 and with 900-kilometre) swath that repeats every 3 days. other gases (which serve as calibrations), the “It is a tremendous advantage to the global car- instrument comes up with an estimate of the bon-cycle community that both approaches number of CO2 molecules in a column of air are being used during the flagship missions,” just 10 kilometres in cross-section. Feed these Salawitch says. data, which have much higher resolution than So far, OCO has glided through its first those obtained in previous efforts, into models testing stage inside a thermal vacuum cham- of atmospheric circulation and you can work ber with no insurmountable problems. The out how and from where the gas is spreading. team is getting ready to test the performance The OCO team is attempting to measure dif- of the instrument early in 2008. But every day ferences in trace gases with a 1 part per million between now and launch poses challenges, says precision against a background of 380 parts per Crisp. Perhaps to remind himself as much as million of CO2 equivalent (the approximate anyone else, Crisp says, “This is fundamentally concentration of CO2 in the atmosphere today) a science experiment. We’re asking whether while the spacecraft travels at 7 kilometres per this technique will work as well as the models second. Although measurements of trace gases are telling us it will.” With Kyoto taking effect on Earth and even on Mars are made down to and a reliable bookkeeping system for carbon parts per billion, the interest there is in absolute “This is fundamentally a science sources and sinks sorely needed, it’s not just the ■ levels, not in small changes. “If we measure two experiment.” — David Crisp scientists who will be awaiting the results. parts per million more CO2 over the eastern Amanda Haag is a science writer in Colorado. part of the country versus the western part, that See Commentary, page 789.

787