GNSS Radio Occultation: progress report

1. GNSS Radio Occultation:progress report Lidia Cucurull Joint Center for Satellite Data Assimilation JSDI funded project: Further development of the assimilation of Global Positioning System (GPS) radio occultation (RO) observations UMBC, Maryland, May 2009

2. Outline • Recent work • Assimilation code • Latency issues • Early evaluation of new instruments • Ongoing activities (research-2-operations) • Future plans

3. Assimilation code • More accurate forward operator for refractivity • Three term expression • Analysis of different sets of refractive indexes • Update of the quality control procedures • More observations (in particular in tropical latitudes) • Optimal observation error characterization • Smoother normalized differences • No empirical tuning • Changes resulted in an improvement in model skill in SH (mass fields) and reduction of the low- and high-level tropical wind errors. • These changes are under current pre-operational testing at NCEP for Fall 2009 implementation • Detailed description of the changes and results can be found in Cucurull 2009, submitted to MWR.

4. (O-B)/O_err original (ops) QC & error modified QC & error NH TR Many more Observations !! Very few observations Errors too small SH

5. Recent impacts with COSMIC • expx (NO COSMIC) • cnt (operations - with COSMIC) • exp (updated RO assimilation code - with COSMIC) COSMIC provides 8 hours of gain in model forecast skill at day 4!!!!

6. COSMIC latency • There was a latency problem with the COSMIC data at NOAA • As a consequence, many observations arrived too late to be assimilated in the operational system • As a collaborative effort between UCAR, NESDIS and NWS, this issue was resolved in January 2009 and many more observations were within the assimilation window after the fix • COSMIC data is being pulled from TNC/TOC instead of DDS server since 3 March 2009.

7. Preliminary evaluation of GRAS (>30N, August 2008) refractivities • The JCSDA identified the significant effects of a bug in the encoding of the GRAS BUFR files at DMI. • The bug was particularly important in local regions where the differences between the geoid and the ellipsoid heights are significant. • The bug assigned a wrong height to the observations and introduced a bias in the refractivity profiles. • The results were reported to DMI and they fixed the bug in the BUFR files on 11 Nov 2008. • The same bug affected the GRACE-A and CHAMP BUFR files ( GFZ fixed the bug on 14 Nov 2008).

8. Ongoing activities • UCAR recently identified a bug in the processing of the COSMIC observations that modifies the statistics of the data (bias and standard deviation in stratosphere). We are currently testing the impact of the new data to assess if additional modifications in the analysis code are needed. • Evaluating impact of Metop/GRAS and GRACE-A data. If results look good, the data will be ready for operational assimilation (the implementation needs to fit NCO’s schedule). • Working on getting SAC-C (Argentinean mission) sample data for evaluation and possible transition to real-time -currently delivered in non-real time- and TerraSAR-X (German mission) -data in real-time but not distributed to the external community yet. • Proposal submitted to evaluate ROSA (Italian RO receiver) on Oceansat-2

9. Ongoing activities (cont’d) • Collaborative project with CWB (Taiwan) to accelerate the use of COSMIC into the NCEP’s regional data assimilation system (NAM). (JCSDA project). • Co-mentoring with UCAR/CDAAC a SOARS (NSF-funded) student to improve the use of COSMIC data to retrieve PBL heights for assimilation into the real-time mesoscale analysis system (RTMA). (NCEP/EMC project) • Got funding from NESDIS/OSD for a 1-year (2-year?) post-doc to add the GNSS RO OSSE capability into the Joint OSSE project. (JCSDA project) • Test possible satellite configurations to design optimal constellation for operational global weather forecasting (UCAR will focus on the regional component). • Use of RO for validation of re-calibrated radiances and evaluation of the impact of RO observations on the bias correction of nadir sensors into the CFSRR system (Re-analysis). (NESDIS project; PI Cheng-Zhi Zou, submitted) • Coordinating work with Dr. Kursinski on the external grant awarded via competitive Federal Funding Opportunity (FFO).

10. Improving the Impact of GPSRO Data Assimilation at NCEP Scope of Research: E. R. Kursinski, A. Otarola University of Arizona F. Xie, C. Ao Caltech/JPL • Improve the impact of GPS RO particularly in the lower troposphere • Two areas of emphasis • Super-refraction • Occurs at very sharp PBL top over oceans • Causes refractivity to be systematically underestimated • Improve GPS RO error covariance • Create humidity dependent error covariance • Examine representativeness error

11. Ongoing activities (cont’d) • Working with OSD on a follow-on capability beyond COSMIC - as NESDIS Program Scientist for GNSS RO . This involves NESDIS, NWS and other NOAA line offices. • NOAA is moving forward on a replacement to COSMIC • “COSMIC 2” (conceptual) • 12 satellites • Homogeneous global coverage • GPS and Galileo signals (minimum) • Higher gain antennas for lower SNR • 8000+ soundings per day • 30 minutes data latency - from sounding to data delivery to the weather/ionospheric centers

12. Future plans • Monitor current GPS RO observations (modifications in the processing of the data might require adjustments in the assimilation algorithm) • Evaluate future RO sensors (as early as possible) to enable operational assimilation of the new data as soon as possible. • Delivery of a perl-based software to monitor GPS RO statistics in real time at NCEP (in collaboration with D. Hunt, UCAR) • Operational assimilation of COSMIC (and other RO sensors) into NAM (in collaboration with Yen-Chih Shen, CWB) • Improve the assimilation code • Transitioning to 1D bending angle • Explore 2D forward operators • Support the design & requirements for a COSMIC follow-on mission • Collaboration project with Spain IEEC for launch of a COSMIC-type receiver for RO and modified antennas (multi polarized) to detect precipitation