Docstoc

Martin Lohmann UCAR JCSDA NESDIS COSMIC

Document Sample
Martin Lohmann UCAR JCSDA NESDIS COSMIC Powered By Docstoc
					    JCSDA GPS RO ASSIMILATION

        Error Characteristics

         Martin S Lohmann




1                               Retreat October 2004
    This presentation




    1.   Observation errors (important to specify covariances)

    2.   Assimilation strategies below superrefraction layers

    3.   QC

    4.   BUFR files at NESDIS/NCEP




2                                                                Retreat October 2004
    RO Errors

    1.   Above approx. 20-25 km:
            Errors are dominated by background/ionospheric noise and errors in the 1. Guess
            Errors are not related to the neutral atmosphere
            Dynamic error estimation can be used

    2.   From approx. 5 km to 20-25 km:
            Errors are dominated by along track horizontal variations
            The large scale variations result in representativeness errors
            Small scale variations (turbulence) are measurement errors
            Dynamic error estimation?

    3.   0 to approx. 5 km:
            Errors are dominated by along track horizontal variations and superrefraction
            (tracking errors - CHAMP)
            Dynamic error estimation?
            Superrefraction

    4.   For non-local operators it is straightforward determine the corresponding
         measurement errors. In this case the height ranges indicated above are likely to
         change



3                                                                                   Retreat October 2004
    Dynamic error statistics (20-40 km) vs. previous studies




4                                                        Retreat October 2004
    Refractivity errors for a single occultation




5                                                  Retreat October 2004
    Bending angle errors for a single occultation




6                                                   Retreat October 2004
    Strategy for error estimation below 20 km



       If possible use dynamic error
        estimation to increase impact of
        RO data
                                                 Empirical
                                                 relation
       Look for possible correlation
        between observation errors and       
        some signal property, e.g. FSI
        phase or amplitude fluctuations.
        Also a useful approach for fine-
        tuning of QC
                                                       x
       For the First assimilation attempt
        we will use fixed error estimates
        based on (Kou et al. 2004)



7                                                            Retreat October 2004
    Strategies for handling superrefraction

        Refractivity:                       Local bending angle:

           Refractivity is biased below       Measured bending angles
            superrefraction layers              are unbiased below
                                                superrefraction layers
           Detect height of a likely
            superrefraction layer and
            discard observations or            BUT the problem is ill-
            consider observations as a          conditioned as different
            lower bound below that              model states will correspond
            height                              to the same bending angle
                                                profile
           Likely superrefraction layers
            can be detected from:              Model resolves
             – Measured N – gradient            superrefraction ambiguity
             – Model N – gradient
             – Model-observation bias

8                                                                  Retreat October 2004
    QC Strategies, fine-tuning of QC criteria


       First assimilations will be based on current CDAAC QC

       Profiles with large deviations between background and observations will
        be rejected

       Profiles which pass the CDAAC QC, but fails the NCEP sanity check
        may point out weaknesses in CDAAC QC

       Profiles which pass the NCEP sanity check but are rejected by
        CDAAC QC, will indicate that the CDAAC QC could be to strict

       QC flags are very important



9                                                                   Retreat October 2004
 NCEP RO BUFR-izing


    Currently preparing NCEP BUFR files from CDAAC BUFR files
     (a) Prepare NCEP BUFR tables from CDAAC BUFR tables
     (b) Generate NCEP RO BUFR data files using NCEP BUFR tables
     (c) Test the generated NCEP BUFR FILES for consistency with
         the corresponding CDAAC BUFR FILES


    Intermediate plan to test assimilation of the NCEP BUFR files in the
     NCEP “operational” DAS

    Issues/Questions:
    Are the CDAAC BUFR tables/files in final form ?
      – May be dependent on UKMO
    Are the CDAAC BUFR files already tested for consistency with their parent
     NetCDF files ?


10                                                                     Retreat October 2004
 Summary and outlook

    Above 20 km, error covariances for both optimized bending angles and
     refractivity, including off-diagonal terms, are currently estimated as part of
     CDAAC

    Error Statistics based on these error estimates are found to be in good
     agreement with other studies

    The possibility of using dynamic error estimates below 20 km will be investigated
     in the future

    It is expected that bias caused by superrefraction can be avoided by assimilating
     bending angles instead of refractivity

    CDAAC QC will be fine-tuned based on feed-back from the first assimilations of
     CHAMP occultations

    NESDIS is currently preparing NCEP BUFR files from CDAAC BUFR files



11                                                                           Retreat October 2004

				
DOCUMENT INFO