The capacity of the United States to monitor Earth’s vital signs is being stymied by tight budgets and poor coordination. Alexandra Witze reports.It seems like such a little thing, the ability to lie back and look up at the full Moon. A moment of wonder or romance on a summer evening, perhaps, but not something vital to the way you do your job. Unless, that is, your job is measuring the amount of photosynthesis going on in Earth’s oceans. SeaWiFS is a NASA satellite that spends most of its days staring down at the ocean, measuring the subtle colour changes that come about as phytoplankton levels wax and wane. But once a month, SeaWiFS takes its electronic eyes off the water, rolls itself backwards and takes a picture of the full Moon. Without the check-up on its colour percep-tion that this regular ‘lunar calibration’ provides, SeaWiFS could find its judgement drifting slowly off-kilter.Unfortunately, SeaWiFS can’t keep this up for ever. It was designed for five years and has lasted ten. And at the moment, there is no replacement quite as good. Two of NASA’s other Earth-observing satellites carry sensors that can measure ocean colour, as do Europe’s Envisat mission and some other satellites. But none of them is as good at monitoring ocean colour as SeaWiFS, says the project’s chief scientist Gene Carl Feldman of NASA’s Goddard Space Flight Center in Greenbelt, Maryland. And the US instrument designed as a direct follow-on to SeaWiFS is just not as good, many say. It has fewer wavelength bands, it might have problems correcting for atmospheric turbulence, its observations could be corrupted by stray light leaking in around the sensors — and it will never take time out to check its colour vision by staring at the Moon. In other words, the next generation of ocean-col-our sensors, built by the most advanced research nation in the world, will in some respects be a step back. And this is not an isolated problem. Climate scientists have a list of a couple of dozen ‘essen-tial climate variables’ (see ‘The dimensions of the problem’) that they would wish to see monitored in perpetuity. In fact, they’d like a lot more than that — the list started out with more than 150 vari-ables and was winnowed down in large part on the basis of what data were readily available. “There was a fair amount of pragmatism,” says Kevin Trenberth, a climatologist at the National Center for Atmospheric Research in Boulder, Colorado. Nevertheless, in some cases the rel-evant measurements are not yet being made (see ‘The crucial measurement’, page 785). And although some data sets are being interrupted or degraded, others are duplicated — provided by multiple satellites and multiple nations. Not enough eyes on the prizeSEAWIFS PROJECT, NASA/GODDARD SPACE FLIGHT CENTER, ORBIMAGEAtlantic productivity revealed by SeaWiFS.782NATURE|Vol 450|6 December 2007NEWS FEATURE EARTH MONITORINGThe problems are global, as each nation strug-gles to fund and maintain data streams from satellites that serve its own interests (which may be shaped by the particular research interests of its scientists). But the issues are most apparent in the US government’s civilian Earth-observ-ing satellites, of which there are 30. Turf battles among multiple US agencies, as well as tight budgets, threaten the future of the country’s Earth monitoring. The National Academies, not known for alarmist views, was prompted earlier this year to note that “the United States’ extraordinary foundation of global observa-tions is at great risk”.The administration sees talk of a crisis as unwarranted. “I think the panic is greatly exaggerated,” says Vice-Admiral Conrad Lautenbacher, head of the US National Oceanic and Atmospheric Administration (NOAA), which oversees the country’s weather satellites as well as its fisheries and other ocean resources. Lautenbacher is, among other things, a great believer in putting national capabilities in a broader global context. He has been a driving force in America’s contribution to the creation of a worldwide coordinating network for Earth observations, the Global Earth Observation System of Systems (GEOSS), and hopes that such networking can provide the seamless inte-gration of Earth-monitoring systems needed to protect society against natural hazards. GEOSS was launched two years ago with a ten-year mandate to get the world’s view of its common house in order. Lautenbacher and other sup-porters say that although it is moving forward slowly (see ‘All in this together’), it has garnered enough political support to address some of the major issues with data gaps and other observa-tional problems.GEOSS does little, however, to solve one of the sector’s fundamental problems: bridging the long-standing gulf between the scientific com-munity, which generally wants to fly cutting-edge instruments with which to discover things, and the operational community that has the job of providing long-term but unglamorous data sets. In the United States, NASA builds and launches research satellites and NOAA han-dles the operational systems such as weather satellites. But NOAA does only oceans and atmosphere — keeping track of things on land is the responsibility of the US Geological Survey (USGS), which runs the Landsat Earth-observ-ing satellites with NASA. Depending on what piece of information about Earth is needed, it could have been gathered for any number of purposes by any one of the three agencies. And that’s not even counting private remote-sensing spacecraft, nor the military satellites gathering both classified and unclassified data.United frontOne way to simplify things would be to have a unified system for operational measurement of the variables of interest. This is the purpose of Europe’s Global Monitoring for Environ-ment and Security (GMES) programme (see page 778). Yet attempts to unify disparate sys-tems can bring problems of their own. The National Polar-Orbiting Operational Environ-mental Satellite System (NPOESS) combines activities previously carried out by the NASA and NOAA low-Earth-orbiting satellites with the defence department’s weather-satellite pro-gramme. Begun in 1994 and run by Northrop Grumman, by 2005 the NPOESS had accumu-lated so many cost overruns that it triggered a mandatory federal review. Its estimated cost of nearly US$7 billion had soared to at least $11 billion, and the initial launch date has slipped from 2009 to