The sixth FORMOSAT-3/COSMIC Data Users' Workshop was held on 30 October–1 November 2012 in Boulder, Colorado. The purpose of this workshop is to highlight accomplishments in the areas of global positioning system (GPS) radio occultation (RO) operations and algorithm development, meteorology, climate, and ionospheric applications using COSMIC data accessed from the COSMIC Data Analysis and Archive Center (CDAAC). A summary of both the workshop presentations and recommendations is provided with an update of the outstanding issues and potentially new applications that can be explored using COSMIC-2 data.


What: More than 130 people representing 15 nations met to highlight accomplishments in global positioning system (GPS) radio occultation (RO) operations and algorithm development, meteorology, climate, and ionospheric applications using COSMIC data.

When: 30 October–1 November 2012

Where: Boulder, Colorado

Since its launch in 2006, Taiwan's Formosa Satellite Mission 3–Constellation Observing System for Meteorology, Ionosphere, and Climate (FORMOSAT-3/COSMIC) has provided more than 3.2 million global positioning system (GPS) radio occultation (RO) soundings (~1500–2500 soundings per day) to support research and operational numerical weather prediction (NWP). Data processed by the University Corporation for Atmospheric Research (UCAR) COSMIC Data Analysis and Archive Center (CDAAC; see appendix of acronyms) have been widely used by meteorology, climate, and ionospheric communities. As of June 2012, more than 1,900 researchers from 63 countries have become registered users of the data. COSMIC's success has also prompted U.S. agencies to move forward with a follow-on FORMOSAT-7/COSMIC-2 RO mission with Taiwan that will launch six satellites into low-inclination orbits in early 2016 and another six satellites into high-inclination orbits in early 2018. Being developed by the National Aeronautics and Space Administration (NASA)'s JPL, the GNSS RO payload, named TriG (Tri-GNSS), is designed with the capability to track RO signals from the GPS, GLONASS, and Galileo global navigation systems. It is expected to yield up to 12,000 RO profiles per day after the two receiver constellations are fully deployed. The COSMIC-2 soundings will have better signal-to-noise ratios (SNR) than those of COSMIC, allowing profiles closer to Earth's surface and providing higher precision to extend existing applications and promote new ones for the weather, climate, and ionospheric communities. The COSMIC-2 status is available at

The objectives of the sixth FORMOSAT-3/COSMIC Data Users' Workshop were to (i) highlight accomplishments and discuss remaining challenges in the areas of RO data processing and applications and (ii) summarize potential new applications for COSMIC-2 data. Attending the 3-day workshop were government, academic, and private sector participants from Australia, Austria, Brazil, Canada, China, Denmark, Germany, Italy, Japan, Mexico, Spain, Sweden, Taiwan, the United Kingdom, and the United States, as well as 25 graduate students from Taiwanese and American universities and institutes. Presentations were divided into several sessions, spanning RO data processing, meteorology/weather forecast applications, climate applications, ionospheric applications, and COSMIC-2 and future missions. Group discussions were held at the conclusion of each session. The presentations are available online at

This summary highlights some of the presented results and outstanding issues that remain in RO data processing and applications in the “Summary of presentations and session discussions” section. A summary of workshop recommendations on potential applications of COSMIC-2 is presented in the “Recommendations regarding potential applications of COSMIC-2” section.


RO data processing.

CDAAC scientists reported on the status of COSMIC system and inversion algorithms, including updates on COSMIC L2 Civilian (L2C) signal tracking (implemented since January 2012), low Earth orbit (LEO) precise orbit determination (POD) and atmospheric excess phase processing status, and algorithm development for neutral atmospheric data processing. Characterization and reduction of random and systematic errors were the subjects of several lively discussions. B. Schreiner (UCAR) overviewed the current status and future plans in CDAAC. Investigating clock stability, E. Griggs (University of Colorado Boulder) indicated that GLONASS clocks have higher noise than GPS clocks. A simulation study showed that RO bending angles retrieved by the radioholographic method have similar noise-resolution characteristics to geometric optics (C. Ao, JPL). A “variational combination” method was suggested involving combining dual-frequency measurements as observations in one-dimensional variational data assimilation (1D-Var) rather than two variables for ionospheric correction (T.-K. Wee, UCAR). Additionally, a qualitative analysis of the impact of systematic errors induced by the open-loop (OL) Doppler model error was presented (F. Zus, GFZ).

Presentations on advanced applications of RO data included (i) detection of ducting events by analyzing deep RO signals (S. Sokolovskiy, UCAR), (ii) detection and characterization of heavy precipitation via polarimetric RO observations (E. Cardellach, ICE-CSIC/IEEC), (iii) 10-m precision planetary boundary layer (PBL) height detection based on the bending lapse rate (C. Ao, JPL), and (iv) ionospheric reconstruction using 3D assimilation and data interpolating empirical orthogonal functions methods (R. Stoneback, the University of Texas at Dallas). Given the gradually decreasing COSMIC data volume, it was exciting to hear of several new sources of RO data in the near future, including data from the Meteorological Operation B (MetOp-B), the PAZ polarimetric–RO mission, and rising occultations from TerraSAR-X and TanDEM-X. Outstanding issues for RO data processing primarily involve (i) improving RO retrievals in the upper stratosphere and lower troposphere, (ii) developing better quality control procedures, and (iii) exploiting new RO data products.

RO applications for meteorology/weather forecasting.

Louis W. Uccellini, director of the National Centers for Environmental Prediction (NCEP) (at the time of this meeting), outlined ongoing efforts to accelerate the transition of new research and operational observing capabilities (advanced microwave, hyperspectral infrared, GPS RO, GOES-R) into NCEP's operational NWP system. He noted that GPS RO has achieved the third highest impact per observation in NWP systems, suggesting COSMIC-2 will have substantially more impact than COSMIC. Several presentations assessed the value of assimilating GPS RO data in different NWP systems (i.e., the Arctic system reanalysis, Taiwan official weather operation systems, etc.). Observations and predictions of atmospheric phenomena including tropical cyclones, extreme precipitation, and atmospheric rivers were also presented. With the advanced OL tracking technique, it was demonstrated that COSMIC data are able to penetrate deep into the lower troposphere and detect water vapor variations in the tropical atmospheric boundary layer. Several presentations demonstrated the utility of RO data for analyzing and predicting tropical cyclone forecasts.

Participants agreed that several questions should be addressed for RO meteorology/weather forecasting applications, including (i) How do negative refractivity biases from RO retrievals impact NWP?, (ii) Can bending-angle assimilation reduce the impact of negative refractivity bias?, and (iii) How to conduct optimal vertical thinning of RO data to maximize its impact on NWP?

Climate applications.

A theme common to many of the presentations was the need to identify the causes of the remaining structural uncertainties among different operational centers, which have been examined in part by Ho et al. (2012). Presentations on the current climate applications using GPS RO data focused on the following issues:

  • RO data quality: RO climate applications depend on accurate RO inversion processing and stable data quality. S.-P. Ho (UCAR) discussed the operational inversion algorithms and processing methods used by different operational centers. He summarized the differences and standard deviations of the individual centers relative to the intercenter mean in order to quantify the structural uncertainties in the variables derived through the RO retrieval chain.

  • Using RO data to study modes of variability: Z. Zhen (UCAR) presented results for characterizing the evolution of the Madden–Julian oscillation using COSMIC bending angle, refractivity, and moisture profiles. C.-Y. Huang (NCU) reported results for characterizing the global precipitable water in ENSO events revealed by COSMIC measurements. Several people presented results for quantifying tropical tropopause variability and its link with deep convective temperature and moisture signals observed in GPS RO data {W. Randel [National Center for Atmospheric Research (NCAR)], T. Birner [CSU], and L. Pan [NCAR]}. G. Stephens (JPL) highlighted some key climate challenges that could be resolved using GPS RO data, noting that the high vertical resolution RO data are adequate to resolve changes in the tropopause and boundary layer, as needed to understand important climate processes.

  • Using RO data to assess/calibrate other datasets: S.-P. Ho, A. Reale [National Oceanic and Atmospheric Administration (NOAA)], and B. Sun (NOAA) validated NOAA satellite products and radiosonde data using CHAMP, COSMIC, and GRACE temperature and moisture profiles. S.-P. Ho also used GPS RO temperature profiles to calibrate multiple years of microwave sounder temperature measurements in the lower stratosphere.

Remaining issues for GPS RO climate applications include (i) quantification of the uncertainty of accuracy in terms of upper boundary (above 30 km), (ii) continuation of RO trends research (Ho et al. 2012) to quantify and understand the issues of structural uncertainty among operational centers, and (iii) quantification of the accuracy for RO derived water vapor for climate studies.

Applications in the ionosphere.

The RO technique can monitor the ionospheric state in terms of both slant total electron content (TEC) and amplitude scintillation S4 index along GNSS ray paths as well as profiles of retrieved electron density under the spherical symmetry assumption. RO has several advantages over other ionospheric survey methods, including providing whole profile, global coverage with high vertical resolution, and accuracy. Various aspects of RO applications in the ionosphere were demonstrated during the workshop. A. Burns (NCAR) and L. Qian (NCAR) reported the climatology of COSMIC electron density and compared that with the NCAR-TIEGCM. C.-Y. Huang and Y. H. Chu (NCU) demonstrated the utility of RO SNR for mapping the ionospheric sporadic E layer. B. Carter [Royal Melbourne Institute of Technology (RMIT University)] did a systematic study on the COSMIC S4 index. B. Zhao (IGGCAS) reported an interesting feature of ionospheric F3 layer from COSMIC data.

G. Crowley (ASTRA), M. Butala (JPL), G. Bust (JHU APL), T. Matsuo (CU), and A. Chartier (UB) reported their progress in using the COSMIC RO data in ionospheric data assimilation, showing great impact on ionospheric nowcasting and potential for enhancing short-term forecasting. X. Yue (UCAR) reported his preliminary results in ionospheric reanalysis by assimilating all the available ground- and LEO-based GNSS data during 2002–12.

Some outstanding issues in RO observations of the ionosphere were also discussed. The Abel inverted electron density can contain significant errors due to the spherical symmetry assumption in areas of large horizontal inhomogeneity, especially in the low-latitude and low-altitude regions. Since the raw slant TEC can be difficult to use directly for the scientific studies, it is still worthwhile to improve electron density retrievals. In addition, more robust methods are needed to localize irregularities along the GNSS ray to better serve space weather monitoring. Shorter data delays (~<5 min) are needed for accurate ionospheric weather nowcasting and forecasting because of the ionosphere's rapid evolution.


It is anticipated that the dramatic increase in the number of atmospheric and ionospheric observations from the COSMIC-2 RO mission will greatly benefit the research and operational communities. The workshop made the following recommendations:

  • Meteorology/weather forecast applications: Deeper penetration of the tropical lower troposphere with high accuracy and high density will provide improved moisture observations under clouds, which is very important to a variety of studies including tropical cyclone analysis and prediction and global and regional PBL structures. With the gradual decrease in COSMIC RO data, its impact on NWP has been decreasing. The significant increase in data from COSMIC-2 should dramatically boost the impact of RO data on global and regional NWP well beyond that achieved by COSMIC.

  • Climate applications: With its substantially better diurnal and spatial coverage, COSMIC-2 will support and enable innovative research for (i) studying hydrological processes over the tropics, (ii) evaluating climate model veracity by assessing their representation of key climate processes, (iii) calibrating other satellites measurements and building of long-term climate records, (iv) improving the accuracy and quality of global reanalysis, and (v) studying improved resolution of dynamical scales (gravity waves, etc.) and modes of variability.

  • Ionospheric applications: The vast amount of data that will be provided from COSMIC-2 will allow the first opportunity to make accurate ionospheric nowcasts and short-term forecasts via sophisticated data assimilation systems analogous to those in meteorology. Mapping global irregularities, including sporadic E layer and F layer and scintillations in near–real time, is anticipated as well. Additionally, plasmasphere imaging and research could be done based on a substantial increase in the number of POD observations.


The Sixth FORMOSAT-3/COSMIC Data Users' Workshop was supported by the National Science Foundation (NSF), the National Space Organization (NSPO) Taiwan, the National Science Council (NSC) Taiwan, and the Taipei Economic and Cultural Office in Houston, Texas (TECO-Houston). Contribution from C. Ao was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration.



    Atmospheric and Space Technology Research Associates


    COSMIC Data Analysis and Archive Center


    Challenging Minisatellite Payload

  • COSMIC-2

    Constellation Observing System for Meteorology, Ionosphere, and Climate Mission 2

  • CSU

    Colorado State University

  • CU

    Colorado University at Boulder

  • ENSO

    El Niño–Southern Oscillation


    Formosa Satellite Mission 3–Constellation Observing System for Meteorology, Ionosphere, and Climate

  • GFZ

    German Research Centre for Geosciences


    Globalnaya Navigatsionnaya Sputnikovaya Sistema (or Global Navigation Satellite System)

  • GOES-R

    Geostationary Operational Environmental Satellite R-Series

  • GNSS

    Global Navigation Satellite Systems

  • GPS

    Global positioning system


    Gravity Recovery and Climate Experiment


    Institute of Space Science-Spanish National Research Council/Institute of Space Studies of Catalonia


    Institute of Geology and Geophysics Chinese Academy of Sciences


    Johns Hopkins University Applied Physics Laboratory

  • JPL

    Jet Propulsion Laboratory

  • L2C

    L2 Civilian

  • LEO

    Low Earth orbit

  • MetOp-B

    Meteorological Operation B

  • NCU

    National Central University in Taiwan

  • NWP

    Numerical weather prediction

  • OL

    Open loop

  • PAZ

    Satellite mission for polarimetric radio occultation techniques using GNSS system, and potential application in atmospheric precipitation

  • PBL

    Planetary boundary layer

  • POD

    Precise orbit determination

  • RO

    Radio occultation

  • SAR

    Synthetic aperture radar

  • SNR

    Signal to noise ratio

  • TanDEM-X

    TerraSAR-X add-on for digital elevation measurement

  • TEC

    Total electron content

  • TerraSAR-X

    Terra SAR operating in the X band


    Thermosphere–Ionosphere–Electrodynamics General Circulation Model

  • TriG


  • UB

    University of Bath


and Coauthors
Reproducibility of GPS radio occultation data for climate monitoring: Profile-to-profile inter-comparison of CHAMP climate records 2002 to 2008 from six data centers
J. Geophys. Res.