RUSSIAN JOURNAL OF EARTH SCIENCES VOL. 10, ES1002, doi:10.2205/2007ES000262, 2008
Preliminary Results
[38] With respect to planned work the following results were achieved during the first year of the project:
Mediterranean Sea
|
Figure 1
|
[39] The work in the Mediterranean Sea contributed to the development and improvement of the coastal
altimeter data processor, working by screening multiple along-track altimeter data together, rather than
individual ground points. This methodology permits both better determination of abnormalities caused by
the altimeter and the radiometer (impacting on wet tropospheric and ionospheric path delays for instance),
and retrieval of invalid corrective terms (by high order polynomial interpolation). With respect to Ssalto/DUACS
products, this innovative methodology allows use of more altimeter data in the coastal ocean and ensures
improved quality of the derived products. The coastal altimeter processor also improves local corrections for
ocean tides & short period atmospheric forcing (Mog2d-Medsea) and computes a new local Mean Sea
Surface (MSS) (Figure 1).
|
Figure 2
|
[40] Several results have been obtained using the new data screening, de-flagging and re-interpolation,
demonstrating how they lead to a reconstructed sea level profile, which is then de-aliased using a regional
model. A higher resolution MSS (including across-track effects) has also been computed from an inverse
method - least-square fitting on the altimetry. This MSS is along track with a 5-point grid across-track.
Differences with respect to the CLS01 MSS are significant in proximity to the coast. The wet tropospheric
correction - and the land flag based on radiometer values - are still the main cause of both data drop-out
and reduced data quality. Over the Mediterranean Sea the work is well advanced: we use multi-mission altimetry,
data editing and state of the art corrections (Figure 2).
Black Sea
|
Figure 3
|
[41] Monitoring the mesoscale water dynamics in the Black Sea is a very important application of
altimetry. It is well known that the circulation displays a chain of anticyclonic eddies, transported
cyclonically by the Rim current around the basin. Smaller eddy features, with intense anticyclonic eddies
and upwelling are present in the northeast, with some anticyclonic eddies along the southern coast and high
chlorophyll concentrations at the mouth of the Danube River. Some comparisons have been made between
SST and/or chlorophyll maps of the Black Sea against the TOPEX/ERS 2 composites that SIO get from
CCAR mapped to 3 day. From the comparison it is clear that the small anticyclonic, long-lived eddies
(60-80 km) along the Anatolian coast in the southern part are not resolved in the CCAR data (Figure 3).
Other examples highlight how the standard altimetric products miss some important features at the mesoscale,
despite capturing the overall circulation at basin scale and the largest eddies in the centre of the basin.
This underline the need for an improved product in this basin. The performance of altimeter-derived
measurements (and model measurements contained in the RADS regional archives) were also investigated in
relation to the sea level and meteo station data (wind) data. The ground tracks closest in space to the in situ
stations were extracted and the data closest in time to the altimeter passes were selected. With reference to
wind amplitudes, correlations in time are not very good - 0.3-0.4 - as shown by scatter-plots. With reference
to sea level, correlations with respect to Jason-1 and Envisat are not very high. High mesoscale activity might
be the reason for the observed discrepancy.
Caspian Sea
|
Figure 4
|
[42] In Central Asia, the application of satellite altimetry has a particularly important role for water
availability monitoring. For the Caspian Sea satellite altimetry can help estimate the water budget. The
Caspian shows a strong annual signal and interannual trends, including an increase from 1992-1995 followed
by a fall to 2001-2002 (Figure 4).
|
Figure 5
|
|
Figure 6
|
[43] A technique for estimating the water budget has been illustrated, using an example of runoff from
the Caspian Sea into the Kara Bogaz Gol (KBG) Bay. The comparison of altimeter-derived Caspian annual mean
sea level with tide gauge data from different stations generally shows good agreement, but with some
discrepancies (up to 7 cm in 1995). There are apparent inconsistencies between different tide gauges - some
show increases when others are dropping - and although part of this difference might be due to geographical
variability, the differences seen at long temporal scales are likely to be due to problems with the gauges.
It is therefore difficult to infer water balance with such problematic in situ data. 10 years of T/P altimeter
data were studied to derive the maps of amplitude of the annual signal. The altimetry processing is challenging
with the problems of ice and the dry tropospheric correction when working on lakes. The value of the dry
tropospheric correction is altitude dependant, but T/P GDRs icorrectly compute the correction at sea level,
while for other satellites the correction is calculated strictly following the topography - i.e. to the lake bed.
There are also land movements, which should be (and are not) accounted for. From an oceanographic point
of view, in the Caspian Sea the situation is more complex than for the Black Sea as there are also some
cyclonic eddies. Tests with a TOPEX radiometer-derived wind speed show it to be unreliable, normally much
higher than model or altimeter-derived wind speed. A Mog2d high frequency model for the Caspian Sea at
7 km resolution offshore, reducing to 1.5 km at the coast, is under development to compute tides to be used
for the tidal correction in the X-TRACK processor. The derived standard deviation of SSHAs along some
Jason tracks (using 1 Hz data) shows some weak oceanographic signals close to the coast (plus the artefacts
due to islands). There is a need for in situ data for validation and to confirm the regional wind response in
the model. Concerning the analysis of the synoptic dynamics induced by atmospheric forcing and Volga River
discharge, a Mean Sea Surface Model GCRAS06 was created, which is not influenced by interannual
changes of the Caspian Sea level (Figures 5 and 6).
White and Barents Seas
|
Figure 7
|
[44] The physical characteristics of the Barents Sea and exchange with the nearest seas have been studied.
The average depth of the basin is
> 200 m and the deepest point is
>
600 m. The White Sea is located at the
southernmost part of the Barents Sea. Many factors influence the hydrodynamic regime of these seas,
including the tidal regime (up to 8 m amplitude in some places). In RADS these tides can be corrected with
FES99 or GOT00 (the latter is the choice recommended by RADS authors). But both models have
0.5o grid - too coarse to properly resolve the White Sea. GC suggests that in this area we should use the HRCRF
(Hydrometeorological Research Centre of the Russian Federation) tidal model. This differs from GOT by up
to 4 m in some places, for instance at the entrance of the White Sea (Figure 7). Another issue in this region
is the Earth's crust uplift, which can be up to 4 mm yr
-1. Storm surges are also important and may reach
2 m in the strongest events. The coverage of this area by different satellites was discussed - T/P and Jason
only marginally touch it. GFO is good for the White Sea, while ERS and Envisat cover all the White and Barents
seas. The ice-free period is Apr/May to Oct/Nov for the White Sea. There was consensus that this is an
appropriate area to show the possible improvements due to the adoption of regional tidal models as opposed
to global models. Correlations of sea level from altimetry and tide gauge data can be fairly high ( > 0.9) for
the Barents Sea using ERS data. For Geosat in the Barents Sea the correlations are noticeably lower, probably
due to orbit errors and more inaccurate corrections. Correlations are extremely high when ERS-2 and Envisat
are used in combination with two Norwegian tide gauges for which there are long, high-quality time series
|
Figure 8
|
(Figure 8). These high correlations might be explained by tide effect, which are very significant in the
Barents Sea. One might be recommended to remove the tides both from satellite data and from tide gauges data
and then make a comparison. Still for the White Sea, but this time with ERS-2 data, which cover the whole basin,
the correlations are significant. The White Sea is also affected by tidal rips, a modification of roughness
when tidal currents converge in a narrow channel, which is very difficult to model. Finally, also in the White
Sea there is a lot of scatter between in situ and altimeter/model data. To improve the correlation, additional
tuned processing should be attempted (filtering, averaging).

Citation: Lebedev, S., A. Sirota, D. Medvedev, S. Khlebnikova, S. Vignudelli, H. M. Snaith, P. Cipollini, F. Venuti, F. Lyard, J. Bouffard, J. F. Cretaux, F. Birol, L. Roblou, A. Kostianoy, A. Ginzburg, N. Sheremet, E. Kuzmina, R. Mamedov, K. Ismatova, A. Alyev, and B. Mustafayev (2008), Exploiting satellite altimetry in coastal ocean through the ALTICORE project, Russ. J. Earth Sci., 10, ES1002, doi:10.2205/2007ES000262.
Copyright 2008 by the Russian Journal of Earth Sciences
Powered by TeXWeb (Win32, v.2.0).