Satellite Oceanography for Ocean Forecasting
33
Merging of multiple-satellite altimeter data
The merging of multi-satellite altimetric data sets is necessary for a better mapping of sea level and oceanic circulation variations. To merge muti-satellite altimetric missions, it is first necessary to have homogeneous and inter-calibrated data
sets. Homogeneous means that same geopotential model and reference systems for
the orbit and same (as far as possible) instrumental and geophysical corrections
(e.g. same tidal models, same meteorological models, etc) should be used. Intercalibrated means that relative biases and drifts be corrected for at the mm level and
also that the orbit be reduced (to have consistent data sets). A methodology we proposed and successfully tested is to use the most precise mission (TIP, Jason-1) as a
reference for the other satellites (Le Traon et al., 1995; Le Traon and Ogor, 1998).
Once altimetric data have been homogenized and inter-calibrated, the next step is
to extract the Sea Level Anomaly for the different missions and to merge the data
using a mapping or assimilation technique. To eXtract the SLA, it is preferable to
use a common reference surface (e.g. a very precise mean sea surface or mean profiles consistent between the different missions) to obtain the SLA relative to the
same ocean mean. Then data can be merged using a mapping technique as previously described.
2.2.5 Results from TIP and ERS-1I2
An overview ofresults obtained with TIP and ERS-1I2 is now given to illustrate
the contribution of altimetry to ocean forecasting. It contains TIP results on mean
sea level variations, large scale sea level and oceanic circulation variations, Kelvin!
Rossby waves, El Nino and the comparison with global eddy resolving models. TI
P and ERS-1I2 results deal with the combination of TIP and ERS-1I2 for regional
and mesoscale analyses. The reader is referred to the two TIP (1994, 1995) and the
ERS-1I2 (1998) Jouma1 of Geophysical Research special issues for a more complete overview.
To forecast possible global sea level rise, you need to observe the mean sea level
trend over a very 10ng period with accuracy better than 1 mm/year. To achieve this
accuracy from 1000 km altitude is a real challenge. TIP is the first altimetric mission which enab1es us to observe the global mean sea 1eve1 variations with such an
accuracy. The signa1 over the first five years is between 1 and 2 mm/year (10 to 20
cm per century), consistent with resu1ts derived from tide gauges which were (and
still are) extremely useful for validating altimeter measurements. The main advantage of altimetry is global coverage, which makes measurements much more representative of the global mean sea level. Satellite altimetry also offers a unique
ability to observe mean sea level variations at smaller regional scales. We know
that global warming will not induce a uniform rise in the mean sea level. TIP estimations show that the sea level rise can be much higher in some places (a few cm/
year) although there are some regions where the signal is clearly related to an interannual ocean signal (not a secular drift). This also shows why it is so difficult to
monitor the global mean sea 1eve1 from a limited number oftide gauges. TIP data
have also provided very useful estimates of mean sea level variations in regional
33
Merging of multiple-satellite altimeter data
The merging of multi-satellite altimetric data sets is necessary for a better mapping of sea level and oceanic circulation variations. To merge muti-satellite altimetric missions, it is first necessary to have homogeneous and inter-calibrated data
sets. Homogeneous means that same geopotential model and reference systems for
the orbit and same (as far as possible) instrumental and geophysical corrections
(e.g. same tidal models, same meteorological models, etc) should be used. Intercalibrated means that relative biases and drifts be corrected for at the mm level and
also that the orbit be reduced (to have consistent data sets). A methodology we proposed and successfully tested is to use the most precise mission (TIP, Jason-1) as a
reference for the other satellites (Le Traon et al., 1995; Le Traon and Ogor, 1998).
Once altimetric data have been homogenized and inter-calibrated, the next step is
to extract the Sea Level Anomaly for the different missions and to merge the data
using a mapping or assimilation technique. To eXtract the SLA, it is preferable to
use a common reference surface (e.g. a very precise mean sea surface or mean profiles consistent between the different missions) to obtain the SLA relative to the
same ocean mean. Then data can be merged using a mapping technique as previously described.
2.2.5 Results from TIP and ERS-1I2
An overview ofresults obtained with TIP and ERS-1I2 is now given to illustrate
the contribution of altimetry to ocean forecasting. It contains TIP results on mean
sea level variations, large scale sea level and oceanic circulation variations, Kelvin!
Rossby waves, El Nino and the comparison with global eddy resolving models. TI
P and ERS-1I2 results deal with the combination of TIP and ERS-1I2 for regional
and mesoscale analyses. The reader is referred to the two TIP (1994, 1995) and the
ERS-1I2 (1998) Jouma1 of Geophysical Research special issues for a more complete overview.
To forecast possible global sea level rise, you need to observe the mean sea level
trend over a very 10ng period with accuracy better than 1 mm/year. To achieve this
accuracy from 1000 km altitude is a real challenge. TIP is the first altimetric mission which enab1es us to observe the global mean sea 1eve1 variations with such an
accuracy. The signa1 over the first five years is between 1 and 2 mm/year (10 to 20
cm per century), consistent with resu1ts derived from tide gauges which were (and
still are) extremely useful for validating altimeter measurements. The main advantage of altimetry is global coverage, which makes measurements much more representative of the global mean sea level. Satellite altimetry also offers a unique
ability to observe mean sea level variations at smaller regional scales. We know
that global warming will not induce a uniform rise in the mean sea level. TIP estimations show that the sea level rise can be much higher in some places (a few cm/
year) although there are some regions where the signal is clearly related to an interannual ocean signal (not a secular drift). This also shows why it is so difficult to
monitor the global mean sea 1eve1 from a limited number oftide gauges. TIP data
have also provided very useful estimates of mean sea level variations in regional