|Title||Global observing system experiments in the ECMWF assimilation system|
|Publication Type||Technical memorandum|
|Secondary Title||ECMWF Technical Memorandum|
|Authors||Bormann, N, Lawrence, H, Farnan, J|
This study summarises results from observing system experiments with the ECMWF system, conducted over two seasons covering a total of 8 months. The experiments investigate the forecast impact of withholding selected observations from the assimilation system compared to using the full observing system. The observing systems considered are: conventional observations, microwave radiances, data from hyperspectral infrared instruments, bending angles from GPS radio occultation, as well as AtmosphericMotion Vectors.
Results show that conventional observations and microwave radiances are presently the main drivers of headline scores, with infrared sounders adding further robustness for a wide range of geophysical variables. GPS radio occultation measurements give significant impact in the upper troposphere/ lower stratosphere, mainly on temperature, but also other variables, and the data have a clear influence on the mean state in these regions. Atmospheric Motion Vectors add benefits for tropospheric wind, particularly in the tropics and at the short range. The strong impact of the microwave satellite radiances is aided by the availability of an unprecedented number of instruments, providing good spatio-temporal coverage.
The observing systems considered have considerable effects on mean analyses, especially the conventional observations, resulting both from the direct assimilation of the observations as well as interactions with the variational bias correction.