You must have a javascript-enabled browser and javacript and stylesheets must be enabled to use some of the functions on this site.


The use of data-quality information for optimal scientific application of altimetric data

Annalisa Martini(1) , Carolina N. Loddo(1) , Yannice Faugere(2) , Batoula Soussi(2) , Ouan-Zan Zanife(2) , and Joel Dorandeu(2)

(1) Serco S.p.a., Via Sciadonna, 24, 00044 Frascati, Rome, Italy
(2) CLS, 8-10 rue Hermes, Parc Technologique du Canal, 31526 Ramonville St-Agne, France


The objective of this paper is to show how the information related to the instrument status and the complex processing chain is fundamental to use in an optimal way the altimetric data. Instruments are designed as a compromise between scientific requirements and technical feasibility. Data processors are designed as a compromise between extraction of maximal information from the instrument and the constraints in processing and dissemination resources. This latter compromise should drive the data quality expected from an instrument. In practice, data quality is variable due to several sources of quality degradation. Optimal application of these data to scientific studies requires knowledge of the data quality. This paper explains important transient error sources that frequently affect data quality. These are related to instrument anomalies and degradations, near-real-time calibration quality variations, and data processor issues. The paper outlines the various sources of information that have been set up to provide the data user with quality information, in particular many dedicated records in the products themselves, and several documentation resources available on line. It also describes methods of applying this information, for example by filtering using specific data fields. For some parameters, error information in product files is not yet adequate, and improved information is available in separate documents. Finally the off-line consolidation of products allows corrections to near-real-time orbital data and the usage of more up-to-date auxiliary informations, and have therefore a significant quality advantage over near-real-time data. This paper addresses the most important differences between the two.



                 Last modified: 07.10.03