Latest Reprocessing and Occultation Prediction Activities at EUMETSAT
Presenter:
Axel von Engeln
EUMETSAT
EUMETSAT
Talk
The presentation will give an overview of radio occultation (RO) related reprocessing and occultation prediction activities at EUMETSAT. Both of these activities are currently provided for the operational GRAS instrument on the EUMETSAT Polar System (EPS), but will include in the future also the RO instrument on the EPS-Second Generation and the GNSS-RO instrument of the Jason-CS/Sentinel-6 programme.
On EPS reprocessing, the GRAS instruments on Metop-A and -B are providing about 600 occultations per day each; the Metop-C launch in 2018 added another about 600 occultations per day. A first GRAS reprocessed data set of Metop-A and -B has been completed and made available in 2017/18 and an updated one to include also data up to 2017 is being prepared. The data is processed to bending angles (level 1B) with the Full Spectrum Inversion on both frequencies and for all altitudes. Long term evaluations show excellent stability of the GRAS instruments in terms of quality and quantity. Validations against ECMWF, COSMIC, CHAMP, ROM SAF reprocessed data will be shown and discussed. The complete reprocessed data set, including a full validation report, is already / will be made available through the EUMETSAT archive. Level 2 data (refractivity, temperature and water vapour profiles), based on these reprocessed data set, are/will be generated by the ROM SAF.
On the occultation predictions, these products are primarily targeted for ground based campaigns, e.g. radio sondes. EUMETSAT provides the predicted occultations for all three Metop satellite, as well as the sub-satellite point. Data is provided at around 6UTC every day and covers predictions for the next 14 days. The accuracy of the occultation predictions is very high, the location and the time of the occultation is accurate to within a few seconds and few km (when no GPS maneuvers is present). The main limitation is caused by occultations not processed to bending angles, e.g. because they are not tracked by the GRAS instrument (due to e.g. a GPS outage/maneuvers), or the data is not of sufficient quality. Several measures have recently been implemented to address these limitations. In total, about 90% of the occultations that are predicted are actually also processed to L1B.
Presentation in PDF:
