COMPLEMENTARITY OF ERS-1 AND TOPEX POSEIDON ALTIMETER DATA IN ESTIMATING THE OCEAN CIRCULATION - ASSIMILATION INTO A MODEL OF THE NORTH-ATLANTIC/

Citation
E. Blayo et al., COMPLEMENTARITY OF ERS-1 AND TOPEX POSEIDON ALTIMETER DATA IN ESTIMATING THE OCEAN CIRCULATION - ASSIMILATION INTO A MODEL OF THE NORTH-ATLANTIC/, J GEO RES-O, 102(C8), 1997, pp. 18573-18584
Citations number
26
Categorie Soggetti
Oceanografhy
Journal title
JOURNAL OF GEOPHYSICAL RESEARCH-OCEANS
ISSN journal
21699275 → ACNP
Volume
102
Issue
C8
Year of publication
1997
Pages
18573 - 18584
Database
ISI
SICI code
2169-9275(1997)102:C8<18573:COEATP>2.0.ZU;2-R
Abstract
This study examines the possibility of improving ocean circulation est imates by using TOPEX/POSEIDON (T/P) altimetric measurements in conjun ction with those obtained from ERS 1. An evaluation is carried out by assimilating these data sets into a circulation model of the North Atl antic. Experiments with simulated data show that the simultaneous use of both data sets results in clearly enhanced results, even for the es timate of the deep circulation, which is a key issue in ocean circulat ion modelling. However, the benefit is less clear when using real alti meter data, which could result from the fact that the spectral structu re of the orbit error is not taken into account in the assimilation pr ocedure. Comparisons of the model results with the altimeter measureme nts and with numerous independent in situ velocity measurements demons trate that the estimates of the surface and subsurface circulations ar e still improved with a dual assimilation. However, this gain seems we aker than could be expected from experiments with simulated data. More over, comparisons with deep velocity measurements show no improvement at all, which probably indicates that measurement errors and nonmodell ed physics are not well taken into account in the assimilation procedu re. It is also observed that the combined use of the two data sets may help in limiting some systematic shortcomings of the numerical tools in exploiting the data. Finally, it is worth noting that the ERS 1 orb it characteristics almost systematically lead to better results than T /P ones for the range of temporal and spatial scales addressed in thes e experiments.