Recent and New GNSSRO Missions: Quality Assessment and Comparative Data Assimilation Study

The Joint Center for Satellite Data Assimilation (JCSDA) supports its partner agencies in the effort to assimilate Radio-Occultation (RO) observation from Global Navigation Satellite Systems (GNSS). The objectives are twofold:

1) Adding into operations, observations from new platforms as they become available, and

2) Developing and transitioning advanced assimilation capabilities.

JCSDA’s efforts toward 2) are described in a companion talk (Zhang et al.). This talk focuses on 1) and reports on the evaluation of the data from the PAZ and MetOp-C missions. We use the PAZ data processed by CDAAC for which the two polarized signals have been combined to produce bending angles, similar to other existing missions. The MetOp-C observations were provided by Eumetsat. For each platform, we computed the mean bias and root mean square errors of the observations minus background (OmB) using the operational NOAA global 6-hour forecast over the month of January 2019. We compared those statistics against the ones obtained from the currently operational GNSSRO platforms (COSMIC-1, MetOp-A&B, Tandem-X, Terra-X and Kompsat-5). The results show that PAZ data have a very low bias with performances comparable to those of Kompsat-5. The MetOp-C statistics are very similar to those of MetOp-A, with a significant bias in the lowest troposphere. For both platforms, we conducted a month long of 6-hourly cycling assimilations/forecasts and generated score cards, a measure of the benefit of the assimilation relative to the current operational forecast. The PAZ score card shows an improvement of the geopotential height anomaly correlation for days 1-3 forecast in the Southern hemisphere. The assimilation of MetOp-C data results in a temperature bias reduction above 10hPa in the Northern hemisphere for all forecasts. We intend to repeat similar studies with the coming COSMIC-2 mission.