Version 1 (modified by lalibert, 8 years ago) (diff)


Quality Assurance Issues

Scientific quality assurance is an important part of WP3. With the CDB, it is possible to create diagnostics that uses and combine high frequency and high resolution data efficiently. Already, during the beta testing phase, data quality issues were observed but only when the data was combined in the latter stages of the diagnostic computation. Similarly, the bugs with the Climate Data Operators (CDO) were discovered because of its different behavior on different models. This means that, while model outputs were standardized using CMOR, differences in the data structure persist that can lead to unexpected results when using common tools like CDO.

With the suite of diagnostics to be included in the CDB, it will be possible to monitor quality of datasets in the archive. The initial version of the CDB does not however describe a method to report the data quality issues encountered in the archive. In future versions, a reporting mechanism should be implemented that would hopefully improve the standardization of model outputs in the archive.