The “Columbus Geoid” of the Ohio State University, based on terrestrial gravimetric measurements, brought the last pre-space draft, rough and incomplete, of the gravity-shape of Earth. Since then, and especially after the mid 1960's, the race for the global calculation of the geoid was well and truly on. The first geoid model of the Standard Earth program of the Smithonian Astronomical Observatory, resulting from the satellite monitoring data (optical photograph against a background of stars or by Doppler-Fizeau effect on transmitter satellites), was released in 1966.
CNES entered the game in 1970 with its GRGS team (Groupe de Recherche de Géodésie Spatiale - Research Group on Space Geodesy). In a partnership with a German team, currently called GFZ (GeoForschungs-Potsdam), the Space Geodesy (GS) team kept developing increasingly more accurate patterns.
So they understand now the gravity variations on Earth which correspond to displacements of masses as little as one centimetre of water in a 400km x 400km basin, which would correspond to the detection from space of mass variations of 160,000 tons.
Enough to accurately determine the annual ice melting in Greenland, a thousand times more important (see below).
Mapping of the annual polar ice melting (in cm of the equivalent water height), modelled from 7 years of GRACE data.
But before getting here, the monitoring measurements (optical at first, then Doppler and laser) of some 40 space missions were used to improve our knowledge of the gravity field. CNES had in fact a pioneering role in the launch of satellites dedicated to the gravity measurement, such as Starlette in 1975 and then Stella in 1993. Those little dense spheres (48 kg, 48 cm diameter), simply featured laser retroreflectors to be observed by laser monitoring stations with a millimetric precision, brought many useful information from the GRIM series to the end of the 1990's.
Until 2000, our knowledge of the geoid increased, on average, by a factor ten every ten years, benefiting from monitoring system enhancements and resulting from ever greater accuracy in geodesic orbit calculation, from a hundred of metres to a centimetre.
This improvement even experienced a considerable acceleration since 2000 (see below), thanks to the successive low orbit missions (< 500km) dedicated to the study of gravity such as CHAMP (DLR, 2000-2010), GRACE (N/DLR, 2002-) and GOCE (ESA, 2009-).
The three missions benefited from the technology of the on board GPS receptors which allow a continuous 3D monitoring of the orbit as well as the electrostatic accelerometers' technology of the Onera which are in charge of isolating - and measuring - the surface's accelerations (mainly friction and radiation pressure) from the gravity's. CNES provided the STAR accelerometer on board CHAMP and the GS team was responsible for the validation phase of the instrument.
The GRACE mission was, on its part, composed of two twin satellites in K/Ka-bands of which the interferometric measurement of the distance between them was in a few micrometers. Such accuracy was impossible to reach with a terrestrial monitoring because of the troposphere in which the waves delay could only be modelled in the millimetres.
This mission brought a new interest for the gravimetry study from space and demonstrated that the very accurate relative orbital observation between two satellites gave a lot of information about mass transfers in general, an especially water transfers. This technique was efficient for the global monitoring of the displacements of continent, polar and oceanic's freshwater (see below). It provided an integrated measurement unlike the space altimetry, which made the combination of the two techniques complementary to distinguish, for example, the steric effects (1.1mm/year) from the mass effects (1.8mm/year) in the ocean or even the variations of the underground water compared with surface water.
It has been transposed to the Moon to refine the Selene gravity field (GRAIL mission, NASA, 2011) and it was even more promising in the future using a interferometric laser system a hundred times more accurate.
Seasonal evolution (in 2007) of mapped mass variations of the geoid in height (on ± 15mm, images at the top) and converted in equivalent liquid water height (on ± 90cm, images at the bottom), on the basis of the GRACE's data.
The GOCE mission launched into an even lower orbit, at a height of 255 km and proposed by a European consortium which includes CNES was based upon the principle of an in situ measurement of the gravity by differentiation of the gravitational attractions of masses between 6 accelerometers fixed in the satellite 50cm apart from one another. The increased measurement accuracy up to 10-12m/s2/?Hz allowed shortest wavelengths (between 300 and 150 km) to be achieved.
The EIGEN-6 model thus gave a more accurate and homogeneous reference of altitudes. On the continents, the corrections made on an anterior reference geoid such as EGM2008 could have reached the metre in some areas like Himalaya, Amazonia, Equatorial Africa, Antarctica (see below).
Variations of the geoid height between the EIGEN-6 and EGM2008 models
On the oceans, the new geoid model allowed the oceanic general circulation deduced from the space altimetry to become more accurate at scales of 100 km.
Average dynamic topography of the North-West Atlantic (filtered at 100km), differences between an average oceanic surface deduced from space altimetry and the EIGEN-6 geoid. The arrows indicates the amplitude and direction of the corresponding geostrophic speeds (source: CLS).