|Title||Data usage and quality control for ERA-40, ERA-Interim and the operational ECMWF data assimilation system|
|Series/Collection||ERA Report Series|
|Authors||Tavolato, C, Isaksen, L|
|Event Series/Collection||ERA Report|
|Place of publication||Shinfield Park, Reading|
This report investigates and interprets time series of data usage, quality control decisions and departure statistics for ERA-40, ERA-Interim and the operational ECMWF data assimilation system. This is done with special emphasis on the performance of the (variational) quality control for conventional data. While ERA-40 and ERA-Interim are mainly affected by the changes in the available observation datasets, the changes in the use of observations are also due to developments within the assimilation system. ERA-40 used 3D-Var whereas ERA-Interim used 4D-Var, so the intercomparison shows elements of the impact of the assimilation system upgrade. Another interesting comparison is done of the ERA-Interim and the operational assimilation system. Whereas ERA-Interim uses a fixed 4D-Var assimilation system over the whole period, operations updated the 4D-Var assimilation system several times during the period investigated. Not all those updates relate to changes in the data usage, but some differences can be seen and evaluated. These differences are presented with selected examples, as well as tables of data usage statistics to give a general overview.