Kuo_GPSARC_Retreat_122710.pptx - COSMIC

Document Sample
Kuo_GPSARC_Retreat_122710.pptx - COSMIC Powered By Docstoc
					GPSRO Data Processing and Science
      Applications at UCAR
              Bill Kuo
            UCAR COSMIC
•   Status of FORMOSAT-3/COSMIC
•   Planning for FORMOSAT-7/COSMIC-2
•   Missions of Opportunities
•   GPSRO data processing at UCAR
•   GPSRO science applications
•   Possible areas for collaboration
             Successful COSMIC Profiles
Neutral Atmosphere ~ 2.5M
Ionosphere EDPs ~ 2.6M
          Current COSMIC Spacecraft/Payload Status
            FM1              FM2                  FM3               FM4        FM5          FM6

Space-     Nominal   - Lost 1 of 2 solar   - Frozen solar array   Battery    Lost       -
craft                arrays                drive                  concern?   contact
                     - Average ~50 %       - Average ~50%                    Sep 26,
                     duty cycle            duty cycle                        2010
                     (Currently ~ 70%)     - Lost contact Aug                Returned
                                           1, 2010                           Nov 10,

GOX        -POD-01   -                     -                      -POD-01    - Lost     POD-00,03
Payload    low SNR                                                low SNR    POD-00     low SNR
                                                                  - OFF

            FORMOSAT-3/COSMIC sounding number went below 1000 in Nov
            2010, and now a little over 1000 per day.
Continuation of GPSRO measurements
• FORMOSAT-3/COSMIC has a mission life of
  five years. Gradual degradation of the
  constellation is to be expected after 2011.
• It is essential to have the follow-on mission
  (F7/C2) to continue and enhance the GPSRO
• We should also look at the possibility of using
  other international research missions for
                A Possible Design for

                     System Requirements
• GNSSRO Level 1 Requirements Documents (L1RD) status
    • L1RD Completed and signed 5 May 2010
• L1RD assessment underway
   • Initial budget/partnerships for COSMIC-2 do not fully support all L1RD
        • C-2 was designed to be a replication of C-1 – except X2 satellites
             • All at 72 degrees
             • 2 northern tracking stations with current latency (around 60
                min average)
        • L1RD requires 2 inclinations – 72 degrees and 24 degrees
        • L1RD requires 45 min average latency
• Partnering with the AF for SSAEM (Space Situational Awareness
  Environmental Monitoring) sensors – 2 secondary payloads on 6 satellites
   • AF purchase of rockets will close budget shortfall – allow us to meet all
      L1RD threshold requirements

          First Launch in mid 2014, Second launch in mid 2016
                   F7/C2 Current Activity
•   Mission Definition Review – successfully completed in August 2010
•   TriG Payload
     –   SRR complete in August 2010
     –   PDR planned for November 2010
     –   Antenna design kicked off for COSMIC-2
     –   Procurement strategy in draft
•   Air Force is proceeding with partnership
     – Payloads contracts work
     – Discussions with STP for the Minotaur 4 – received ‘11 funding
     – AF provided draft MOA under review at NOAA
•   NSPO is moving forward quickly on spacecraft procurement
     – RFI released in May 2010
     – 5 RFIs under review
     – Plans to release RFP for 12 spacecraft by January 2011
•   NOAA working ground planning
     – Discussions with KSAT (Kongsberg Satellite Services, Norway) on ground station options
     – Working with UCAR on proposal to “operationalize” CDAAC processing software to
       install at NSOF
       GNSS RO Possible
Missions of Opportunity (MOOS)
    (Excluding NASA Decadal Missions)
                                GNSS RO Possible
                             Missions of Opportunity

Mission           Launch-    GNSS RO                  Orbit (alt/inc/ LT)   # Occs    Operational/Real
                  Duration   Payload                                        Per day   -Time
F7/C2             2015       JPL TriG (GPS,Galileo)   800km/72,24°/-        >8,000    Yes
METOP-B           2012       GRAS (GPS)               817km/98.7°/09:30LT   ~600      Yes
OceanSat-2        2009       ROSA (GPS)               720km/98.3°/12:00LT   ~500      No

KOMPSAT-5         2010       IGOR+ (GPS)              685km/98.5°/06:00LT   ~500      No

Megha-Tropiques   TBD        ROSA (GPS)               867km/20°/-           ~500      No

SAC-D             TBD        ROSA (GPS)               657km/98.5°/10:15LT   ~500      No
TanDEM-X          20         IGOR (GPS)               514km/97.4°/18:00LT   ~500      No
PAZ               2012       IGOR+ (GPS)              510km/97.4°/-         ~500      No
EQUARS            2012       IGOR (GPS)               750km/20°/-           ~500      No
CNOFS             2008       BlackJack (GPS)          853/405km/13°/-       ~250      Best effort
SAC-C             2000       BlackJack                715km/98.5°/10:15LT   ~200      Best effort
                       Missions of Opportunity
            SAC-C (Satélite de Aplicaciones Cientificas – C)

•   Argentinian CONAE mission launched Nov 2000
•   715km altitude, 98° inclination, 10:15 LT
•   JPL BlackJack, Open Loop, four single patch antenna
•   Near real-time data provided by Germany’s GFZ and CONAE
•   CDAAC providing 140-180 occultations per day to NOAA
•   ~ 50-60% success from tracked profiles
•   UCAR working with Tom Meehan and CONAE
     • configure GPS receiver to track rising occultations all day (now only 10-
        18 UTC)
     • reduce negative impact of aging oscillator
   Near Real-time SAC-C/COSMIC
           Global stats

SAC-C                   COSMIC
                          Missions of Opportunity
• TerraSAR-X, German mission, Launch Jun 2007
   -   BRE IGOR
   -   GFZ providing NRT ~200 setting profiles to GTS (UCAR assisted GFZ in debugging
• Tandem-X, German mission, Launch Jun 2010
   -   BRE IGOR
   -   Identical orbit as TerraSAR-X, useful for scientific studies
• OCEANSAT-2, Indian mission, Launch Sept 2009
   -   ROSA receiver
   -   Yaw-biased attitude, 35 degrees
   -   Commissioning phase, no data available
• KOMPSAT-5, Korean mission, IGOR+, Launch 2010?
   -   BRE IGOR+
• PAZ, Spanish mission, IGOR+, Launch in Jan 2012
   -   IGOR+
   -   Tom Meehan says firmware must be modified to provide data useful for RO. This
       effort is not funded
                                COSMIC Operational Weather Centers
                          Getting COSMIC Results toProcessing

                             TACC                                    JCSDA

RTSs:                               UCAR/Unidata’s        N                     ECMWF
Alaska                         C    LDM
                                    WGET Community
                               D                                                 CWB
                                     1500-2000 WMO
                                                          S     GTS
 Input Data                    A     BUFR Files
 - COSMIC data
 - GPS ground data                   per day with         D                     UKMO
 - GPS NDM Bits                A     Latency ~ 75-90min
 - GFS Forecast                                           I
                               C                                                  JMA
 - Bernese Config files
                                                                             Canada Met.
                  Science & Archive

   CDAAC reliability estimated > 99.5%, Latency ~ 75-90 min
                      Main CDAAC Functions

• RO Payload Operation
   – Configuration control (firmware and tracking configuration)
   – Scientific/technical guidance for commanding, operation
   – Near real-time monitoring, trouble-shooting, and Q/C data analysis
• Data Processing and Analysis
   –   Level0 unformatting and QC
   –   GPS ground processing (ZTD, site estimation, clock estimation)
   –   LEO POD, and atmospheric excess phase
   –   Absolute TEC generation
   –   Inversions (neutral atmosphere and ionosphere)
   –   Retrievals (1DVAR)
   –   NWP and correlative data handling
   –   Product QC and analysis
               CDAAC Functions (cont)
• System Operation and Monitoring
   - H/W, O/S and NFS filesystem
   - System fail-overs
   - CDAAC operations
• Input data stream monitoring
   – RTS downlink
   – GPS Bit-grabber operation and monitoring
   – GPS ground data (IGS, NRCan, EUMETSAT, COSMIC sites)
   – IGS and IGU orbits, clocks and EOP (Earth Orientation Parameter)
   – Bernese configuration
   – NCEP GFS, ECMWF, radiosonde, ionosonde
• Data Management, Dissemination, and Archival
• Support Data Users
                Best Effort Monitoring

• Monitor the system regularly throughout the day M-
  F 8am-8pm

• CDAAC Ops team monitors the system 3 times per
  day on weekends and holidays

• Available by email and cell phone

• CDAAC reliability estimated > 99.5%
                     COSMIC/CDAAC Status

• CDAAC 3.0 released late November 2010
• Post-Processing continues …
   - Pushing COSMIC to public website
   - Metop/GRAS 2010 on website
• COSMIC recently producing ~1,000 occultations/day
• CNOFS producing ~125 occs/day
   - Interpolating 1Hz L2 to 50 Hz
   - Working with Paul Straus to modify CORRISS firmware
• SAC-C producing 125-150 occs/day
   - Working with CONAE/JPL to update firmware
   - Finalizing agreement with GFZ to provide SAC-C support for 2011
              F7/C2 Data Processing Center
• Reliable and low latency input data streams (GNSS ground
  network, LEO data from RTSs, ..)
• Primary and Backup Data Processing Center (DPC)
• Development system (GNSS capable)
• Staging system to test processing changes
• Communication access between systems (DPCs, SOCC, RTSs)
• Monitoring of near real-time operations
• Maintenance and on call technical support
• Operator Training
• Science payload processing
• Post-Processing and archiving
                         NOAA Operational DPC
• Primary DPC at NOAA/NSOF (NOAA Satellite Operations
   –   Single/Dual string (operational call)
   –   24/7 monitoring
   –   Virtualization of processing S/W
   –   SW Staging/testing string
• Backup DPC at off-site location
   –   Single/Dual string (operational call)
   –   Hot/Cold Backup (operational call)
   –   Return to Service (operational call)
   –   < 24/7 monitoring (operational call)
   –   Tested yearly, documentation
               Post-Processing and Archive Plan
• Post-processing requires use of up-to-date software and algorithms
• Technical and scientific expertise are required to monitor processing
  and validate data analyses
• UCAR plans to post-process F-7/C-2 data
• Archive via NOAA CLASS system
   – Level0 and higher level products
   – Must find host at Data Center (NGDC, NCDC, ..).
       • Process started
• UCAR will archive real-time and post-processed data on NCAR HPSS
  (High Performance Storage System)
• Taiwan DPC archive
            CDAAC System changes

  Better system for managing code for
production/ research
  Restructure portions of software
  Add more test suites
  Clean up unused code
  Improved QC, bad flags, error estimates
  Multi-GNSS capability, new observables!
  Low latency processing
  Better systems management
  Need to develop F-7/C-2 Level0-Level1 processing
  Improved monitoring scripts
  Improved documentation
             Neutral Atmospheric Inversions
• Restructure ROAM (Radio Occultation Atmospheric
• Add GNSS capability, new observables
• Improve wave optics processing
   – looking for better filtering approaches
   – looking for alternative methods (like recently introduced WDF)
• Improve bending angle optimization
   - testing methods with reduced weight of climatology
   - validation by independent data sources in the stratosphere
• Improve 1DVAR retrieval code and documentation
• Perform additional validation studies (e.g. integrate ROPP
  package, RO-Trends+)
   - Requires investigation and understanding the sources of the
   atmPhs                 1) Mission-dependent; 2-4) Mission-independent.

                          1) Input positions, velocities, raw phase and amplitude,
1) Phase Connection       clim. model.
                          Processing removal of NDM, connection of the phase,
            conPhs        down-sampling to one rate.
                          Output positions, velocities, connected phase,
                          amplitude, HSL, lat, lon, height of TP.
2) Bending Angle
Generation (WO,GO)        2) Input output from 1
                          Processing retrieval of WO (Phase Matching and FSI
            benPrf        or CT2) and GO bending angles for L1, L2, (L5)
                          Output GO and WO bending angles for L1,L2,(L5)
3) Bending Angle
Correction, Connection    3) Input output from 2
                          Processing ionospheric correction (incl. additional
            bcnPrf        smoothing of L4), connection of GO and WO bending
                          Output ionosphere free connected bending angles
4) Bending Angle
Optimization, Inversion   4) Input output from 3, clim. model, atm. model.
                          Processing optimization of bending angles,
                          inversion of N,T,P.
   atmPrf                 Output N,T,P.
                 Ionospheric Processing

• Absolute TEC uncertainty under investigation
   - DCB estimation, Code/phase leveling uncertainty
• Electron Density Profiles
   - Improving EDPs (Electron Density Profiles) by using information
     on horizontal gradients with DA
• Improve scintillation products
• Add GNSS capability, new observables
• Perform additional validation studies
   - With Paul Straus of Aerospace, JPL
  Parallel efforts between U.S. and
           Taiwan for F7/C2
Data impact Data processing     Operation NWP
            R&D Center          DPC       Operation
COSMIC      CDAAC               NSOF/NOAA NCEP
NCAR        R&D                 DPC
            Multiple missions
TTFRI       GPSARC              CWB/NSPO     CWB
GPSARC      R&D                 DPC (TACC)   M.I.C.
CWB         Backup
 Possible Collaboration between U.S.
              and Taiwan
• GPSRO data processing:
  – GPSRO data processing research, inversion, and
    algorithm improvement
  – Operational GPSRO processing
• GPSRO Science Applications:
  – Use of GPSRO data in operational NWP
  – Systematic evaluation of the impact of F3/C and F7/C2
    data on typhoon and flood prediction
  – Ionospheric research and space weather
  – Climate applications, trend detection, post-processing
                             Data Assimilation Retrieval of Electron Density
                             Profiles from Radio Occultation Measurements

Comparison of standard COSMIC Abel retrieval and                         Simulation of retrieval errors for standard COSMIC
  data assimilation retrieval with Ionosonde data                          Abel retrievals and data assimilation retrievals






  Yue, X., W. S. Schreiner, Y.-C. Lin, C. Rocken, Y.-H. Kuo, and B.         Geomagnetic latitude and altitude variations of
  Zhao, 2010: Data Assimilation Retrieval of Electron Density
  Profiles from Radio Occultation Measurements. J. Geophys.                 electron density during noon time (LT=13)
  Res – Space Physics, 2010JA015980, (under review).
 Definition of geopotential
                                                 : geopotential [m2 s-2]

              g,zdz  g,zz
                z                                g: acceleration due to gravity [m s-2]
             0                                   Note: it depends on  and z
                                                 : latitude [deg.]
                                                 z: geometric height [m] (CDDAC COSMIC RO)
  Geopotential height
                                            Z: geopotential height [m] (NWP, WRF, Meteorology)
                                           g0: standard gravity at mean sea level [m s-2]
      Z              z                     Note: it doesn’t depend on  and z
         g0                                  Definition: the acceleration of a body in free fall at
                                            sea level at a geodetic latitude of about 45.5°
                              Textbook of meteorology approximate that g doesn’t
                              change with height and latitude, and then Z is almost close
                            to z, but we cannot use the approximation in RO world.
Plot in next slide:
     • At sea level, difference between Z and z is zero because
          – Same geopotential height at sea level (geoid, g is same at reference height 0)
     • Biases between z and Z depending on latitude and height
          – Due to difference of g and g0
                                                          Ideal sphere Earth
                                  g0                      In meteorology
                      g ,h
                                         Center of mass
                                                             Real Earth

                  Geoid height, 0m

                         g,z < g0

                                                                        g,z < g0

                            g,z > g0

•                                                          
    In z (geometric height) –Z (geopotential height) plot above, positive value in equator shows g in equator
    is smaller than g0.
•                
    Negative value near pole in lower atmosphere is because g in pole is larger than g0, but positive value in
    higher atmosphere (15km~) is, again, because g in higher altitude is smaller than g0
       O-B Statistics in Original and Modified WRFVAR



Positive biases in original WRFVAR are disappeared in modified WRFVAR

    00h                                 12h

        GPSv_NCEP:operational with GPSRO
      GPSh_NCEP:height-correction with GPSRO
          NOGPSv_NCEP:without GPSRO
    24h                       48h

           • H-correction is comparable with NO
           • For forecast, H-correction has slightly
            smaller RMS than NO GPSRO at
            higher levels.
           • H-correction has smaller bias at model
            top.                                  34

Shared By: