The radio occultation experiment aboard CHAMP (Part I): Operational data processing, technical aspects, atmospheric excess phase calibration and ionospheric profiling Jens WICKERT(1), Rolf KÖNIG(1), Torsten SCHMIDT(1), Christoph REIGBER(1),Norbert JAKOWSKI(2), Georg BEYERLE(1), Roman GALAS(1), Ludwig GRUNWALDT(1), Tom K. MEEHAN(3), and Tom P. YUNCK(3) (1) GeoForschungsZentrum Potsdam (GFZ), Division 1: Kinematics and Dynamics of the Earth, Potsdam, Germany. (2) German Aerospace Center (DLR), Institute of Communications and Navigation, Neustrelitz, Germany. (3) Jet Propulsion Laboratory (JPL), Pasadena, USA.
1. Introduction CHAMP (CHAllenging Minisatellite Payload), the German geoscience satellite, was launched on July 15, 2000 (Reigber et al., 2002) and is in orbit and excellent condition now for about 2.5 years. The measurements of CHAMP are used to determine Earth's gravity and magnetic field and to derive precise information about the vertical temperature, humidity and electron density distribution on a global scale using the innovative GPS radio occultation technique (Jakowski et al., 2002, Wickert et al., 2001). The properties of the calibration-free atmosphere limb sounding technique with GPS (e.g. allweather-capability, high accuracy and high vertical resolution) offer great potential for atmospheric and ionospheric research, improvement of numerical weather forecasts, space weather monitoring and climate change detection (e.g. Anthes et al., 2000; Hajj et al., 2000; Kuo et al., 2000; Kursinski et al., 1997). The potential of this, relatively new, sounding technique was demonstrated for the first time by the pioneering U.S. American GPS/MET (GPS/Meteorology) – Experiment (Ware et al., 1996). CHAMP, together with the U.S. Argentinean SAC-C satellite (launched on November 21, 2000), succeeds GPS/MET. The measurements of both satellites and the improved (in relation to GPS/MET) infrastructure for GPS radio occultation data reception, transfer, analysis and provision brought significant progress for the GPS radio occultation technique. Here we focus on the current status of the CHAMP occultation experiment and various aspects of the occultation infrastructure and data processing at GFZ Potsdam.
Figure 1. Number of daily CHAMP atmospheric occultations (duration > 20 s) for 2001 and 2002 (as of October 13, 2002, when the number of 100,000 occultations was reached). The total height of the columns corresponds to the number of daily measurements. The red and blue color indicate occultations with calibration failure and insufficient data quality, respectively. The height of the green columns corresponds to the number of vertical atmospheric profiles provided to the CHAMP data center at GFZ. ECMWF (European Centre for Medium Range Forecasts). Since then within 549 days in 2001 and 2002 a total of 112,066 occultations (duration >20s) were recorded as of December 11, 2002. Since March 10, 2002 (update of the flight receiver software) about 260 measurements daily are performed (Fig. 1).
2. Status of the CHAMP occultation experiment CHAMP’s GPS radio occultation experiment was activated on February 11, 2001. 7 occultation measurements during an one hour period were recorded by the GPS receiver (JPL’s “Blackjack”) onboard the satellite. They were analysed by GFZ’s automatic occultation processing system (Wickert et al., 2001a) and validated with meteorological analyses from Figure 2. Number of daily ionosphere occultations for 2002 (Up to December 8, 2002, black columns). The height of the green columns corresponds to the number of vertical electron density profiles successfully retrieved in DLR Neustrelitz from the radio occultation measurements.
*Corresponding author address: Jens Wickert, GeoForschungszentrum Potsdam (GFZ), Division 1, Kinematics & Dynamics of the Earth, Telegrafenberg, 14473 Potsdam, Germany, e-mail: [email protected]
the GPS receiver onboard the CHAMP satellite (provided by JPL) and the ground segment (see Fig. 4). It consists of the near polar downlink station at Ny Alesund, Spitsbergen (79.0° N, 11.5° E), the fiducial GPS ground network (“High rate and low latency network”, currently consisting of ~40 stations), the Ultra rapid Precise Orbit Determination facility, the operational occultation processing system and, for archiving and distribution, the CHAMP Information System and Data Center (ISDC). A second downlink station at DLR Neustrelitz, Germany (53.1° N, 13.1° E) serves as backup. The GPS ground network is operated in cooperation between GFZ and JPL, the other components are maintained by GFZ. Figure 3. Predicted altitude scenario for the CHAMP mission (December 2002). The scenario strongly depends on the real solar activity, which cannot be completely simulated in advance. An orbit manoeuvre with orbit lifting was successfully on June 9, 2002. A second successful orbit lifting (~20 km) has been carried out on December 10, 2002. Currently a lifetime of longer than 2006 is expected. Vertical profiles of atmospheric parameters were derived using the operational processing system for about 70% of the measurements. More detailed numbers of analysed measurements and provided atmospheric profiles will be given after the first reprocessing campaign in December 2002. During this campaign all atmospheric profiles are derived using a significantly improved inversion software (see also Chap. 6). The first 189 ionospheric radio occultations were performed on April 11 and 12, 2001. In total 73,311 measurements were recorded during 2001 and 2002. 45,555 electron density profiles were derived (Status by December 8, 2002). After the flight-receiver software update on March 10, 2002 continuously on average 260 atmosphere and about 200 ionosphere daily events were recorded (Figs. 1 and 2). Since the CHAMP mission is expected to last at least until 2006 (Fig. 3), an unprecedented long-term-set of GPS occultation data is expected.
4. Operational data processing A dedicated, modular structured, scientific analysis software for GPS radio occultation data of the neutral atmosphere was developed for CHAMP (Wickert, 2002). The modules are continuously improved. The software calculates the atmospheric excess phase for each occultation event (see Chap. 5) and vertical atmospheric profiles (see Chap. 6). The analysis modules are part of a dynamically configurable system for operational data product generation (Schmidt et al., 2002; Wehrenpfennig et al., 2001). This system ensures the continuous data flow through the scientific analysis modules and provides the interface for making available the input data (GPS ground and satellite data from CHAMP and the Fiducial Network, the GPS and CHAMP orbit ephemeris and meteorological analysis data from ECMWF) and for the provision of the atmospheric data products to the archive system (ISDC at GFZ Potsdam). If scientific modules for ionosphere data
3. Infrastructure for GPS occultation data analysis The infrastructure for the GPS radio occultation experiment was installed within the framework of the CHAMP mission and the German HGF (Helmholtz Society for Research) strategy fund project GASP (GPS Atmosphere Sounding Project, see Reigber (1998)). Within GASP a pre-operational atmospheric data product delivery is aimed at, which requires a maximum time delay of not more than 3 hours between measurement and product delivery to weather services. The main components of the operational infrastructure for data generation, transfer, analysis and archiving are:
Figure 4. Overview of the infrastructure for measurements, data reception, transfer, analysis and distribution of CHAMP’s GPS radio occultation experiment.
near-polar receiving station at Ny Alesund. There is a ground contact about every 86 minutes with a minimum duration of ~7 minutes (Wickert et al., 2001b). The time delay due to the data transfer to the processing center at GFZ can be neglected. The time delay for providing the ground station data (hourly RINEX files) via CHAMP’s ISDC is currently between 30 to 70 min. The delay can be further reduced since the data are transmitted from the ground station sites as files covering 15 min intervals. These files are transferred to the network processing centers with maximum time delay of 15 min (Wickert et al., 2003a, Galas et al., 2001). The described operational availability of the GPS satellite and ground data with low latency allows for the demonstration of rapid orbit and occultation processing. Since April 2002 the GPS and CHAMP satellite orbit ephemeris are available every 3 hours (Ultra rapid Science Orbit, USO; Reigber et al., 2002). The accuracy of the USO is comparable to the CHAMP standard product, the Rapid Science Orbit (RSO), which is delivered on a daily basis with ~14 hours latency (König et al., 2002). Satellite Laser Ranging (SLR) residuals for CHAMP USO and RSO are shown in Fig. 5. The average RMS of the difference to IGR orbit solution (provided by the International GPS Service, IGS) of the GPS satellites is ~8 cm for the RSO and ~12 cm for the USO. Due to the availability of Ultra rapid Science Orbits the demonstration of Near-Real-Time provision of precise GPS occultation data products became feasible and is demonstrated at GFZ Potsdam since April 2002. The results of a study, performed with CHAMP data from April 28, 2002 are shown as an example in Fig. 6. For each occultation event a delay between measurement and provision of the temperature profile of 3-6 hours was reached. The comparison shows (Fig. 6) that the quality of the near-real-time data products is comparable to that of the standard products (generated using RSO, which are provided on a daily basis with ~16 hours latency.
Figure 5. Comparison of Satellite Laser Ranging (SLR) residuals for Rapid Science Orbits (RSO) and Ultra rapid Science Orbits (USO) for the CHAMP satellite, derived by GFZ’s precise orbit determination facility. The accuracy of both orbit products is comparable (Average SLR residual for RSO 6.8 cm; for USO 8.4 cm). analysis are included, the operational processing of CHAMP’s ionospheric occultations is possible (Jakowski et al., 2002). These data may be used to derive space weather information on the global state of the ionosphere (Jakowski et al., 2001, Wehrenpfennig et al., 2001). Beside the operational occultation data analysis, the timeliness of the input data provision is the crucial factor for operational provision of atmospheric data products. I.e. CHAMP’s occultation data, the Fiducial Network data and the precise orbit ephemeris of CHAMP and the GPS satellites have to reach the atmosphere processing unit with low latency. For the CHAMP occultation infrastructure it was demonstrated, that a delay between measurements and provision of the data products at ISDC of 3-6 hours for each measurement can be reached and further reduction of that delay is possible. The first requirement is continuous availability of CHAMP’s GPS data for orbit determination and occultation processing. This is achieved by using the
5. Atmospheric excess phase calibration The standard method for the derivation of the atmospheric excess phase is a double difference technique to correct for satellite clock errors. GPS measurements (50 Hz precise phases L1 and L2) from the occulting and an reference GPS satellite are combined with simultaneously recorded GPS ground station data from the Fiducial Network (see also Fig. 4). Details of the excess phase calibration are given e.g. by Hajj et al. (2002), Wickert (2002), Wickert et al. (2001b) and Schreiner et al. (1998). The termination of the Selective Availability (SA) mode of the GPS (resulting in reduced amplitude of GPS clock variations; see e.g. Wickert, 2002) on May 2, 2000, 04:05 UTC had several advantageous consequences for the excess phase calibration, which for first time was demonstrated with CHAMP’s occultation data.
Figure 6. Comparison (bias and standard deviation) of two sets of 182 vertical dry temperature profiles, derived from CHAMP occultation measurements on April 13, 2002. The profiles are derived using a) RSO and b) USO. The comparison indicates 2 nearly identical data sets with no remarkable mean deviation up to 40 km and negligible standard deviation of