170
M.-H. Rio
but not in the direct MDT from which they have been removed by increasing filtering (resulting in slightly increasing RMS values). At scales shorter than 300 km, the
RMS values are dominated by both the geoid omission and commission errors.
The use of almost 3 supplementary years of GRACE data for the computation of
the EIGEN-GRGS.RL02.MEAN-FIELD compared to the GGM02S model results
in a reduction of the RMS difference to synthetic estimates at scales shorter than
300 km for the zonal component and 400 km for the meridian component.
To further illustrate this point, a focus is chosen in the Gulfstream area (see
Fig. 10.4). Figure 10.4a shows the direct MDT obtained at 133 km resolution
with EIGEN-GRGS.RL02. The signal is quite noisy. When a 300 km filter is
applied (Fig. 10.4b), the noise is reduced and the velocities computed by geostrophy from the direct MDT are quite consistent with the synthetic velocity estimates
(Fig. 10.4c). This is a strong improvement compared to the circulation obtained, at
300 km resolution, with the GGM02S geoid (Fig. 10.4d), and for which a lot of
noise, mainly on the meridional component of the velocity, is observed.
10.3 Toward Higher Resolution of the Geoid and the MDT
As we have seen in the previous section, huge improvements have been made during
the last 20 years for the estimation of the geoid and consequently the ocean Mean
Dynamic Topography. However, the quality of the latest “satellite-only” geoid models still limits the spatial resolution of the ocean MDT to scales larger than around
300 km. On the other hand, altimetric Sea Level Anomalies are available along-track
every 7 km and, when gridded, with an approximate resolution of 50–100 km.
The spatial resolution of altimetric Mean Sea Surfaces is much higher, down
to 20–30 km. Various methods have therefore been developed in order to increase
the resolution of the ocean Mean Dynamic Topography, for a better exploitation
of altimetric data. These methods can be divided into two main categories. In the
first method, the geoid resolution is improved, and then a higher resolution MDT
is computed using the direct method. In the second method, a large scale MDT is
first computed and further improved using external oceanographic data to resolve
the shorter scales.
In the first case, the quantity to improve is the geoid. This can be done using
in-situ gravimetric data (Hunegnaw et al., 2009; Thompson et al., 2009), most often
limited in spatial extension, resulting in local to regional improvement of the geoid.
Global improvement can be achieved using the shortest scales information of the
altimetric Mean Sea Surface (the spatial resolution of the ocean MDT being coarser
than the spatial resolution of the MSS, the shortest spatial scales of the altimetric Mean Sea Surface are only due to the shortest spatial scales of the geoid, and
can therefore be used to enhance this latter). This method is commonly used to
enhance the resolution of the satellite-only solution, resulting in the so-called combined geoid models (GGM02C, EIGEN3C, EIGEN5C, . . .), which are developed
to a higher degree and order than their satellite-only counterpart. In the case of the
M.-H. Rio
but not in the direct MDT from which they have been removed by increasing filtering (resulting in slightly increasing RMS values). At scales shorter than 300 km, the
RMS values are dominated by both the geoid omission and commission errors.
The use of almost 3 supplementary years of GRACE data for the computation of
the EIGEN-GRGS.RL02.MEAN-FIELD compared to the GGM02S model results
in a reduction of the RMS difference to synthetic estimates at scales shorter than
300 km for the zonal component and 400 km for the meridian component.
To further illustrate this point, a focus is chosen in the Gulfstream area (see
Fig. 10.4). Figure 10.4a shows the direct MDT obtained at 133 km resolution
with EIGEN-GRGS.RL02. The signal is quite noisy. When a 300 km filter is
applied (Fig. 10.4b), the noise is reduced and the velocities computed by geostrophy from the direct MDT are quite consistent with the synthetic velocity estimates
(Fig. 10.4c). This is a strong improvement compared to the circulation obtained, at
300 km resolution, with the GGM02S geoid (Fig. 10.4d), and for which a lot of
noise, mainly on the meridional component of the velocity, is observed.
10.3 Toward Higher Resolution of the Geoid and the MDT
As we have seen in the previous section, huge improvements have been made during
the last 20 years for the estimation of the geoid and consequently the ocean Mean
Dynamic Topography. However, the quality of the latest “satellite-only” geoid models still limits the spatial resolution of the ocean MDT to scales larger than around
300 km. On the other hand, altimetric Sea Level Anomalies are available along-track
every 7 km and, when gridded, with an approximate resolution of 50–100 km.
The spatial resolution of altimetric Mean Sea Surfaces is much higher, down
to 20–30 km. Various methods have therefore been developed in order to increase
the resolution of the ocean Mean Dynamic Topography, for a better exploitation
of altimetric data. These methods can be divided into two main categories. In the
first method, the geoid resolution is improved, and then a higher resolution MDT
is computed using the direct method. In the second method, a large scale MDT is
first computed and further improved using external oceanographic data to resolve
the shorter scales.
In the first case, the quantity to improve is the geoid. This can be done using
in-situ gravimetric data (Hunegnaw et al., 2009; Thompson et al., 2009), most often
limited in spatial extension, resulting in local to regional improvement of the geoid.
Global improvement can be achieved using the shortest scales information of the
altimetric Mean Sea Surface (the spatial resolution of the ocean MDT being coarser
than the spatial resolution of the MSS, the shortest spatial scales of the altimetric Mean Sea Surface are only due to the shortest spatial scales of the geoid, and
can therefore be used to enhance this latter). This method is commonly used to
enhance the resolution of the satellite-only solution, resulting in the so-called combined geoid models (GGM02C, EIGEN3C, EIGEN5C, . . .), which are developed
to a higher degree and order than their satellite-only counterpart. In the case of the