Every element of the Earth system is being studied by hundreds of researchers working in scientific laboratories the world over. Just slip into one of the many international scientific conferences and you'll become aware of the enthusiasm of these women and men as they share the results of their work in sessions, publications and meetings.

You might not easily understand the subject of some sessions (for example KaRIN : an Instrument for Measuring High- Resolution Sea-Surface Topography and Fresh Water Extent, Stage, and Slope), or you might be put off by some pages covered with complex equations, but you can be sure that they'd very patiently and modestly explain what they are about: namely, that models need to be supplied with accurate, regular data in order to validate their pertinence. Such data are of fundamental importance for forecasting, because you can't observe the future!

A long gestation
The goal of a space mission dedicated to climate change is to measure, from space, the key characteristics of a sphere in order to improve our understanding of it and our ability to forecast how it will change. Satellites are the favourite tool for this because they can be used for global, repeated and automatic observation of the whole planet. From the moment a satellite is first required until its launch, the time needed to complete a mission can be as long as ten or even fi fteen years. The expected lifetime of an Earth observation satellite is around 5 years, but this is often exceeded: the ESA satellite ERS-2 was launched on 21 April 1995, and is still operational in 2008. Everything starts with the very large number of mission projects proposed by the scientific community. Space agencies are in charge of selecting the missions that will be given priority, within a given calendar and budget.

smos_clean
The SMOS (Soil Moisture and Ocean salinity) ESA satellite in integration. In the foreground, the radiometer. The objective of SMOS mission is to provide maps of soil humidity and ocean salinity, two key variables of climate monitoring.
© Yoann Obrenovitch
A mission is made up of the space segment, which is the satellite and its instrument (if there are several instruments, this is called the payload), and the ground segment. The cost of a mission can range from 100-300 million euros, which can be compared with the cost of a kilometre of motorway, which is about 10 million euros. The agencies organise the preparation of the mission and entrust its total or partial execution to an industrial prime contractor*. The project gets under way with a conceptvalidation phase during which the agency and the P.I. (Principal Investigator, the chief scientist on the mission) ask one or more competing industry players to look into the technical feasibility of the whole project, establish a schedule for completion, and quote a price.

There then follows a consolidation phase during which the industrial primecontractor makes the first models, and provides tangible proof that the requirements of the mission can be met. This is then followed by a construction phase carried out by the prime-contractor with its own teams and external subcontractors. A major European programme involves thousands of people working for dozens of subcontractors and hundreds of manufacturers. During this phase, the P.I., in coordination with the agency, leads one or more validation campaigns on test areas, for example with a plane or sounding balloon equipped with an instrument similar to the one that will be on board the satellite when it is launched.

Then comes the validation phase of data provided by the satellite, which pass through the ground segment before arriving at the scientifi c laboratory. This is an essential stage, since it is necessary to make sure that the measurement given by the instrument is correct by corroborating it in particular on the test areas. Finally comes the phase when it is used by scientists to tackle climate problems. This requires regular, accurate data, with the guarantee that these observations will continue over a long period.