GPS radio occultation with CHAMP: Initial results - COSMIC

GPS radio occultation with CHAMP: Initial results J. Wickert et al., Proc. Beacon Satellite Symposium, Boston, June 4-6, 2001.

-1-

GPS radio occultation with CHAMP: Initial results J. Wickert, Ch. Reigber, G. Beyerle, R. König, C. Marquardt, T. Schmidt, T. Meehan1, L. Grunwaldt, R. Galas GeoForschungsZentrum (GFZ) Potsdam, Division Kinematics & Dynamics of the Earth, Telegrafenberg, 14473 Potsdam, Germany 1

Jet Propulsion Laboratory (JPL), California Institute of Technology, 4800 Oak Grove Dr., Pasadena, CA 91109, USA

Abstract: The German CHAMP (Challenging Minisatellite Payload) satellite (launched on July 15, 2000) is a geoscience mission aiming for the determination of Earth’s gravity and magnetic field and for GPS-based atmospheric sounding. The atmospheric profiling experiment onboard CHAMP was started in February 2001 and is considered as an important step to improve the GPS (Global Positioning System) radio occultation technique as a tool for global limb-sounding of Earth’s atmosphere. CHAMP carries the latest generation of Jet Propulsion Lab’s GPS flight receiver and has a more directional occultation antenna compared to the GPS/MET “proof of concept” experiment. By the end of May 2001 more than 7000 occultations were analysed. First statistical comparison of a set of 438 vertical profiles of dry temperature, derived from CHAMP measurements, with corresponding global weather analyses were performed. The observed temperature bias is less than ~1 K above the tropopause and less than 0.5 K in the altitude interval from 12 to 20 km at latitudes >30°N. About 55 % of the compared vertical profiles reached the last kilometer above Earth’s surface. Analysis of GPS radio occultation signal in the lower troposphere applying wave optics techniques indicates contribution of a secondary component, which is interpreted as reflected GPS signal from Earth’s surface.

The GPS radio occultation experiment was successfully started on February 11, 2001 [Wickert et al., 2001b]. Thus, CHAMP succeeds the pioneering GPS/MET (Global Positioning System/Meteorology, Ware et al., 1996) ‘proof-of-concept’ project. GPS/MET successfully demonstrated the potential of GPS-based limb sounding from Low Earth Orbiting (LEO) satellites as a technique for global and all weather atmospheric sounding with high accuracy and high vertical resolution [Kursinski et al, 1997, Rocken et al., 1997, Melbourne et al., 1994]. CHAMP and the Argentine SAC-C satellite (launched on November 21, 2000) carry the latest generation of Jet Propulsion Laboratory’s GPS flight receiver (BlackJack) and a more directional occultation antenna (see Fig. 1) with improved gain (+5 db compared to GPS/MET). From this instrumental configuration improved signal quality and application of advanced signal tracking techniques with respect to GPS/MET experiment were expected before starting the occultation experiment [Yunck et al., 2000].

1 Introduction The German CHAMP (CHAllenging Minisatellite Payload) satellite (Fig. 1) was launched on July 15, 2000. The mission objectives are the determination of Earth’s gravity, magnetic field and global sounding of Earth’s atmosphere by GPS radio occultation technique [Reigber et al., 2000a, 2000b]. CHAMP was launched with a COSMOS rocket from Russian cosmodrome Plesetsk (62.5°N, 40.3°E) into an almost circular (eccentricity=0.004), near polar (inclination=87.2°) orbit with an initial altitude of 454 km. Contact: [email protected]

Fig. 1. The German CHAMP satellite (view from antivelocity direction with GPS occultation antennas)


2 Data analysis and processing In preparation of CHAMP’s GPS occultation experiment a complete infrastructure for occultation data generation, processing, transfer and archiving was established. The main components are the GPS receiver onboard the CHAMP satellite (provided by JPL) and the ground infrastructure, which is shown in Fig. 2. It consists of the downlink station at Neustrelitz, Germany (53° N, 13° E) including the Fiducial Network

Precise Orbit Determination

Downlink Station & Raw Data Center

Occultation Processing System

Archives

Data Center

User

Fig. 2. Processing and analysis of CHAMP occultation data [Wickert et al., 2001a]

Raw Data Center, the fiducial GPS ground network, the Precise Orbit Determination facility [Neumayer et. al, 2000], the Occultation Processing System and for archiving and distribution the CHAMP Information System and Data Center. The downlink station and the Raw Data Center are operated by the German Aerospace Center (DLR). The other components are maintained by GFZ, whereby the GPS ground network is operated in cooperation between GFZ and JPL.

-2-

The analysis is based on a double difference method (see Fig. 3, combining GPS ground and CHAMP data to correct for satellite clock errors, Wickert et al., 2001a) to calibrate the atmospheric excess phase. The atmospheric parameters are derived from atmospheric excess phase using geometric optics approach and the Abel inversion technique [e.g. Hocke, 1997; Melbourne et al., 1994]. The analysis of CHAMP’s first occultation data is described in more detail by Wickert et al. [2001b]. 3 Initial Results Fig. 4 shows the Signal to Noise Ratio (SNR) of a representative CHAMP Occultation (number 2 on February 11, 2001). In spite of the activated antispoofing (A/S) mode of the GPS system CHAMP’s flight receiver yields SNR, which is comparable or even better than the corresponding signal strength of GPS/MET) during A/S-off (L1 twice as strong as GPS/MET, L2 similar to GPS/MET). The GPS/MET data analysis was primarily focused to A/S-off periods (“prime-times”). Data processing of A/S-on data required more effort in data analysis, due to significant more noise on L2 data. Consequently from about 50200 occultations, recorded by GPS/MET, only 9050 (estimated from Rocken et al., [1997]) measurements were analysed. Due to the significant better signal strength for A/S-on, observed in the first occultation measurements, data processing will not be limited due to anti-spoofing mode. First statistical comparisons of dry temperature profiles, derived from CHAMP measurements and ECMWF (European Centre for Medium-Range Weather Forecasts) analyses temperature were performed. A set of 438 selected vertical profiles of dry temperature, measured during April 19-21, 2001 was compared with corresponding ECMWF data in

Fig. 3. Analysis of CHAMP occultation data: Double Difference Geometry [Wickert et al., 2001a]

The analysis software for deriving the vertical profiles of atmospheric parameters, as refractivity, pressure, temperature and water vapor is part of a dynamically configurable system for operational data product generation [e.g. Wehrenpfennig et al., 2001] to process the occultation data in a fully automated way.

Fig. 4. Signal to Noise ratio (SNR) of CHAMP Occultation No. 2 (February 11, 2001), SNR for L2 was available with 1 Hz frequency only. [Wickert et al., 2001c]


the height interval between 5 and 25 km. The occultations were separated into three subsets, high latitudes (northern hemisphere), low latitudes and high latitudes (southern hemisphere). The mean deviation (CHAMP-ECMWF) and their standard deviations with 0.2 km sampling are depicted in Fig. 5. In each subset the CHAMP measurements at the tropopause (between 16 and 19 km at low latitudes, about 11 km in the other regions) are systematically colder than the analysis. This warm bias of the analysis is due to higher vertical resolution of the radio occultations compared to the analysis which was available on standard pressure levels only. This has also been observed for GPS/MET data [Rocken et al., 1997]. Above the tropopause, however, the agreement between CHAMP and the analysis is better than 1 K with a standard deviation of ~1.5 K or less. In the northern hemisphere the mean difference is less than 0.5 K between 12-20 km altitude. Below the tropopause the presence of water vapor leads to negative bias of the dry temperature profiles compared to ECMWF temperature. The negative bias, as expected, is strongest at low latitudes due to the high absolute humidity in this region. About 55% of the selected profiles reached the last kilometer above the Earth's surface. Uncorrected multipath effects in the lower troposphere will lead to errors in the refractivity and consequently in the derived temperature and humidity profiles. The application of advanced wave optical retrieval techniques shows great potential to solve this

-3-

problem [e.g. Gorbunov et al., 2000, Pavelyev et al., 1996] One of these methods is the radio holographic approach [Pavelyev et al., 1996]. Observed signal (phase and amplitude) is combined with a simulated signal, derived from raytracing calculations to form a radio hologram. The radio hologram is evaluated by spectral analysis to identify contributions of direct (main) received signal and eventually secondary components due to multipath wave propagation. Fig. 6 shows the temporal evolution of the radio hologram power spectral density, derived from CHAMP occultation event number 38 on February 16, 2001. The data corresponding to the last about 20 seconds of the occultation are shown. The red colors in Fig. 6 indicate strong signals, whereas the blue colors are related to weak signals. A strong component at zero frequency shift is evident. During

Fig. 6. Temporal evolution of the radio hologram power spectral density derived from CHAMP occultation No. 35 on February 16, 2001.

the last 8 kilometers a secondary component appears. This component shows a characteristic frequency dependence over time. This phenomenon was already observed in GPS/MET measurements [Beyerle and Hocke, 2001]. They interpreted this secondary component as contribution from rays reflected from ocean or ice surface. The hypothesis is supported by ray tracing calculations simulating reflected signals [Beyerle and Hocke, 2001]. Fig. 5. Statistical comparison between dry temperature profiles, derived from CHAMP measurements and temperatures from 6-hour ECWMF analyses during April 19-21, 2001. The differences (CHAMP-ECMWF) are plotted for (a) 166 profiles in the northern hemisphere (>30°N), (b) 118 profiles at low latitudes (30°N to 30°S) and (c) 154 profiles in the southern hemisphere (