Satellite Oceanography for Ocean Forecasting
31
2.2.4 Altimeter data processing
So far, we have considered that altimetry provides measurements of the surface
dynamic topography. We now describe the basic altimeter data processing techniques used to extract the dynamic topography from measurements of sea surf ace
topography. Our analysis includes a discussion of measurement errors.
Oceanic signal extraction from altimetry and measurement errors
The altimetric observation of the sea surf ace topography S can be described by:
S =N+11+e
where N is the geoid, 11 the dynamic topography and E the measurement errors
(orbit error, propagation effects in the troposphere and ionosphere, tides, electromagnetic bias, inverse barometer effect, altimeter measurement noise).
Present geoids generally are not ac curate enough to estimate the absolute
dynamic topography 11 globally, except at very long wavelengths (> 2000 km) (see
also discussion in section 2.2.3). The variable pact ofthe dynamic topography 11' (11'
= 11 - <11» (or Sea Level Anomaly, hereafter SLA) is, however, easily extracted
since no prior knowledge of the geoid height is needed. The most commonly used
method is the so-called repeat track method (collinear analysis), suitable for satellites whose orbits repeat their ground tracks (to within ± 1 km) at regular intervals.
For a given track, the variable pact of the signal is thus obtained by removing the
mean profile, which contains the geoid and the quasi-permanent dynamic topography, from each profile.
Altimeter measurements of sea surf ace topography are affected by a large number of errors. These include errors on the range measurement due to propagation
effects in the troposphere and ionosphere; electromagnetic bias; errors due to inaccurate ocean and terrestrial tide models; cross-track geoid errors, and inverse
barometer effect. Most are large-scale, and do oot limit the use of altimetry for
ocean mesoscale studies. For studies of large-scale oceanic variability, most of
these corrections must be taken into account, however, since they can significant1y
contaminate the oceanic signal. Some of these errors can be corrected with dedicated instrumentation (dual-frequency altimeter for ionospheric corrections, microwave radiometer for wet tropospheric corrections). Electromagnetic bias can be
deduced by aoalyzing altimeter data.
The inverse barometer effect is actually a large, real oceanic signal which reflects
the response of the sea level to changes in atmospheric pressure. The static
response assumes a non-dynamic adjustment of the ocean due to atmospheric pressure (about 1 cm for 1 mb change in atmospheric pressure). Recent numerical and
empiric al studies suggest that the ocean generally responds as an inverted barometer except over very short time scales and in semi-enclosed seas (e.g. Gaspar and
Ponte, 1997). This correction should thus generally be applied when studying the
dynamic response of the ocean.
31
2.2.4 Altimeter data processing
So far, we have considered that altimetry provides measurements of the surface
dynamic topography. We now describe the basic altimeter data processing techniques used to extract the dynamic topography from measurements of sea surf ace
topography. Our analysis includes a discussion of measurement errors.
Oceanic signal extraction from altimetry and measurement errors
The altimetric observation of the sea surf ace topography S can be described by:
S =N+11+e
where N is the geoid, 11 the dynamic topography and E the measurement errors
(orbit error, propagation effects in the troposphere and ionosphere, tides, electromagnetic bias, inverse barometer effect, altimeter measurement noise).
Present geoids generally are not ac curate enough to estimate the absolute
dynamic topography 11 globally, except at very long wavelengths (> 2000 km) (see
also discussion in section 2.2.3). The variable pact ofthe dynamic topography 11' (11'
= 11 - <11» (or Sea Level Anomaly, hereafter SLA) is, however, easily extracted
since no prior knowledge of the geoid height is needed. The most commonly used
method is the so-called repeat track method (collinear analysis), suitable for satellites whose orbits repeat their ground tracks (to within ± 1 km) at regular intervals.
For a given track, the variable pact of the signal is thus obtained by removing the
mean profile, which contains the geoid and the quasi-permanent dynamic topography, from each profile.
Altimeter measurements of sea surf ace topography are affected by a large number of errors. These include errors on the range measurement due to propagation
effects in the troposphere and ionosphere; electromagnetic bias; errors due to inaccurate ocean and terrestrial tide models; cross-track geoid errors, and inverse
barometer effect. Most are large-scale, and do oot limit the use of altimetry for
ocean mesoscale studies. For studies of large-scale oceanic variability, most of
these corrections must be taken into account, however, since they can significant1y
contaminate the oceanic signal. Some of these errors can be corrected with dedicated instrumentation (dual-frequency altimeter for ionospheric corrections, microwave radiometer for wet tropospheric corrections). Electromagnetic bias can be
deduced by aoalyzing altimeter data.
The inverse barometer effect is actually a large, real oceanic signal which reflects
the response of the sea level to changes in atmospheric pressure. The static
response assumes a non-dynamic adjustment of the ocean due to atmospheric pressure (about 1 cm for 1 mb change in atmospheric pressure). Recent numerical and
empiric al studies suggest that the ocean generally responds as an inverted barometer except over very short time scales and in semi-enclosed seas (e.g. Gaspar and
Ponte, 1997). This correction should thus generally be applied when studying the
dynamic response of the ocean.