Flechtner F.M., Gruber Th., G?ntner A., Mandea M., Rothacher M., Sch?ne T., Wickert J. (Eds.) System Earth via Geodetic-Geophysical Space Techniques

Подождите немного. Документ загружается.

452 G. Michalak and R. König

Fig. 8 (a) Accuracy TerraSAR-X NRT orbits generated by the CHAIN 3. (b) Latency of the

TerraSar-X NRT orbits generated by the CHAIN 3

Near-Real Time Satellite Orbit Determination for GPS Radio Occultation 453

of the dense 30s clocks is time consuming, the length of the hourly generated long

GPS arc was reduced from 24 to 12 h, the short arc remains of 3 h length. This

chain is a kind of mixture of the two above chains, generates LEO orbits with an

acceptable mean latency of 30 min but with the highest accuracy of 5–6 cm. For

TerraSAR-X the latency (90 min) is dictated mainly by the dump data delay (see

Table 1) but will be improved in the future once the data are dumped over the polar

receiving station Ny

Ålesund. This chain is active since February 2007. The accura-

cies and latencies of CHAMP, GRACE-A and TerraSAR-X orbits generated by this

chain are given in Figs. 6, 7, and 8.

3 Summary

Within the German “Geotechnologien NRT-RO” project, GFZ has developed a NRT

orbit processing system for GPS and LEO orbits to support NRT radio occultations.

The system generates NRT LEO orbits every 1.5 h with average position accura-

cies of 6–10 cm validated by SLR over whole arcs. For NRT occultation products

only last 2 h of arcs are used. The accuracy of these orbit parts is computed from

arc overlaps. The LEO orbits of lower accuracy (20 cm for position, 0.20 mm/s for

velocity) are generated with a latency of 13 min when using predicted IGU orbits

and clocks. More precise orbits (6–7 cm for position, 0.06 mm/s for velocity) are

generated with a latency of 30 min when using IGU predicted orbits but 30s clocks

being estimated. In all processing chains the orbit latency can assure the generation

of occultation products with average delay well below 3 h required for NWP sys-

tems. The NRT processing system is designed for easy extension to other LEOs

delivering NRT data, what was demonstrated by inclusion of the TerraSAR-X data

already 3 months after the launch of the satellite on June 15, 2007. The variety of

the approaches to the NRT orbits (three independent chains for each of the LEOs)

enhances the reliability of the system. The system is fully operational and automatic,

requires however non-negligible human activities to account for new unexpected

situations affecting the automatic processing.

Acknowledgments We thank the CHAMP, GRACE and TerraSAR-X teams for their great work

to guarantee the availability of GPS SST data. The described near-real time activities at GFZ are

supported by the German Ministry for Education and Research within the GEOTECHNOLOGIEN

programme (Research project NRT-RO/No. 1262) and also by GFZ.

References

König R, Michalak G, Neumayer K H, Schmidt R, Zhu S Y, Meixner H, Reigber C (2005) Recent

developments in CHAMP orbit determination at GFZ. In: Reigber C, Lühr H, Schwintzer P,

Wickert J (eds.), Earth Observation with CHAMP. Results from Three Years in Orbit, Springer,

Berlin.

Michalak G, Baustert G, König R, Reigber C (2003) CHAMP rapid science orbit determination –

status and future prospects. In: Reigber C, Lühr H, Schwintzer P (eds.), First CHAMP Mission

Results for Gravity, Magnetic and Atmospheric Studies, Springer, Berlin.

454 G. Michalak and R. König

Michalak G., Wickert J, König R, Rothacher M (2007) Precise satellite orbit determination for GPS

radio occultation in near-real time (NRT). European Geosciences Union General Assembly

2007 Vienna, Austria, 15–20 April 2007, Poster EGU2007-A-08740.

Wickert J, Michalak G, Schmidt T, Beyerle G, Cheng C Z, Healy S B, Heise S, Huang C Y,

Jakowski N, Köhler W, et al. (2009) GPS radio occultation: results from CHAMP, GRACE and

FORMOSAT-3/COSMIC. Terr. Atmos. Ocean. Sci. 20, doi: 10.3319/TAO.2007.12.26.01(F3C).

The Operational Processing System for GPS

Radio Occultation Data from CHAMP

and GRACE

Torsten Schmidt, Jens Wickert, and Grzegorz Michalak

1 Introduction

The GPS radio occultation (RO) technique has established as an excellent method

for global and continuous monitoring of the Earth’s atmosphere (Melbourne et al.,

1994; Anthes et al., 2008). During the last 2–3 years most of the weather service

centers worldwide have begun to assimilate RO data in their global and/or regional

weather models (Healy et al., 2007).

The use of analysis results obtained from RO data, as e.g. atmospheric excess

phases, bending angles or refractivity proﬁles, by weather prediction centers

requires an operational data processing system generating and delivering data prod-

ucts automatically within a certain time limit. Such a continuous Near-Real Time

(NRT) data processing system, the CHAMP (GRACE) Atmospheric Processor (in

the following the acronym CAP is used for both, the CHAMP and GRACE process-

ing system), is operating at GFZ Potsdam with RO data from the German CHAMP

and the US–German GRACE satellite missions (Reigber, 1998).

Because of different user demands with respect to the availability of RO prod-

ucts two data processing modes were introduced: a NRT mode and a standard mode.

Currently in the NRT processing mode the daily averaged time delay between occul-

tation measurements and availability of analysis results for the weather service

centres via the GTS (Global Telecommunication System) is less then2hforabout

75%andlessthan3hforabout 90% of all daily products. In the standard process-

ing mode quality checked proﬁles of atmospheric parameters are available with a

latency of about 2 days.

The realization of CAP follows the principle of separating the processing system

into a scientiﬁc and controlling part leading to an independent and more ﬂexible

software development. The CAP is designed to be easily extendable by additional

scientiﬁc modules or input data. Thus, it also allows for an extension to other single-

or multi-satellite radio occultation missions, as e.g. COSMIC or TerraSAR-X.

T. Schmidt (B)

Helmholtz Centre Potsdam, GFZ German Research Centre for Geosciences,

Department 1: Geodesy and Remote Sensing, Telegrafenberg, 14473 Potsdam, Germany

e-mail: tschmidt@gfz-potsdam.de

455

F. Flechtner et al. (eds.), System Earth via Geodetic-Geophysical Space Techniques,

Advanced Technologies in Earth Sciences, DOI 10.1007/978-3-642-10228-8_40,



Springer-Verlag Berlin Heidelberg 2010

456 T. Schmidt et al.

This chapter is focused on the controlling components of the CAP that generate a

continuous data ﬂow between the scientiﬁc software modules and the user. Further,

we concentrate on the description of the NRT part of the processing system.

2 Infrastructure and Input Data

The generation of RO products requires a complex infrastructure. It was devel-

oped with the start of RO measurements aboard CHAMP in 2001, and since then

maintained and extended for, e.g., GRACE (2002) and other satellite missions

(TerraSAR-X in 2007).

The main components of the infrastructure are (Fig. 1):

• the Black-Jack GPS receiver (provided by NASA/JPL) onboard CHAMP,

• three downlink stations receiving CHAMP data (Neustrelitz and Weilheim, both

located in Germany, and Ny-Aalesund in Spitsbergen, Norway) in cooperation

with DLR,

• the ﬁducial low latency and high rate GPS ground network operated in coopera-

tion between GFZ and JPL,

• an orbit processing facility for determination of NRT orbits for the GPS satellites

and CHAMP,

• the CHAMP/GRACE Atmospheric Processor generating CHAMP/GRACE radio

occultation data products, e.g., atmospheric excess phases, proﬁles of bending

angle, refractivity, temperature, and humidity (Wickert et al., 2001), and

• the CHAMP Information System and Data Center (ISDC) archiving and distribut-

ing CHAMP data and analysis results.

Fig. 1 Infrastructure and

input data for GPS radio

occultation data processing

from CHAMP

3 The CHAMP/GRACE Atmospheric Processor

The goal of the CAP is to coordinate the different data streams and to start the var-

ious scientiﬁc applications automatically when all i nput data for an application are

available. The CAP is a modular structured and dynamically conﬁgurable software

The Operational Processing System for GPS Radio Occultation Data 457

Fig. 2 Schematic view of the CHAMP atmospheric processor at GFZ Potsdam

package consisting of a scientiﬁc and controlling part (Fig. 2). Thescientiﬁc soft-

ware modules (e.g. DDIFF, ATMO) calculate the atmospheric excess phases and

vertical atmospheric proﬁles. The controlling components ensure the continuous

data ﬂow of all input data through the scientiﬁc analysis modules and pro-

vide the interface for feeding the RO products into the data center (ISDC) or

directly to the German Weather Service (DWD) who puts the data to the Global

Telecommunication System (GTS) since July 2007.

CAP runs on multiple UltraSparc II/III processor machines and is a combination

of different subroutines and scripts written in C++, Fortran, IDL, Perl, and C-Shell.

4 Standard and NRT Processing

The processing of RO data at GFZ Potsdam is divided into two parts (Fig. 2):

• the standard processing using CHAMP/GRACE data via the dump s ta-

tions Neustrelitz and Weilheim (Germany) and Rapid Science GPS and

CHAMP/GRACE Orbits (RSO) (König et al., 2002),

• the NRT processing using CHAMP/GRACE data via the polar dump station at

Ny-Aalesund, Spitsbergen and NRT Science Orbits (Michalak et al., 2007).

These two modes are necessary due to demands on product availability. The

availability of RO data products, i.e., the time delay between measurement and

delivery of products to the data center or user, depends on two factors: (1) the

458 T. Schmidt et al.

moment at which all input data for the respective applications (Fig. 2) are avail-

able and (2) the duration that CAP needs to process the input data and generate

products. The NRT data user (e.g., DWD) require a data availability within a time

window of less t han 3 h.

(a)

(b)

Fig. 3 Number of monthly CHAMP (a)andGRACE(b) RO data and time delay between mea-

surement and delivering of the data to the DWD (GTS) from the NRT processing since June 2007

The Operational Processing System for GPS Radio Occultation Data 459

The ﬁrst point is fulﬁlled due to the installation of a polar dump station.

This enables a contact to CHAMP about every 90 min. Since 2006 the GPS

and CHAMP/GRACE satellite orbits are available at latest 30–40 min after the

according dump. These NRT orbits cover a3htimewindow.

The second point (duration of data processing) is uncritical, because the process-

ing of one dump with about 10–20 RO events takes less than 5 min.

Figure 3 shows the number of monthly CHAMP/GRACE RO data and the time

delay between measurement and delivering of the data to the DWD (GTS) from the

NRT processing mode since June 2007. At this time the continuous data transfer of

GPS RO products from GFZ to worldwide weather centers has started.

The averaged time delay between CHAMP or GRACE measurements and data

product delivery is less t han 2 h. For about 10–15% of the products a time delay less

than 1 h is reached (Fig. 3).

5 Summary

The CAP enables an operational data processing and the delivery of analysis results

to the data center or directly to the user. Since June 2007 (start of NRT product

delivery) more than 100,000 CHAMP and GRACE RO products, as bending angles

or temperature proﬁles have been generated (about 150 daily, as of January 2009).

Due to the modular structure CAP is suitable for easy extension to other single-

or multi-satellite RO experiments.

In the standard processing mode quality checked products (refractivity and tem-

perature proﬁles) are available with a latency of about 2 days. A near-polar receiving

station at Spitsbergen and a dump-oriented CHAMP/GRACE and GPS orbit produc-

tion cycle made it possible to implement a NRT processing at GFZ Potsdam for the

ﬁrst time. Since mid-2007 GPS RO products with an averaged time delay of less

than 2 h between measurement and availability of the analysis results are generated

and transferred continuously.

Acknowledgments This study was carried out in the Geotechnologien program under the grant

of the German Federal Ministry of Education and Research.

References

Anthes RA, Bernhardt PA, Chen Y, Cucurull L, Dymond KF, Ector D, Healy SB, Ho SP, Hunt DC,

Kuo YH et al. (2008) The C OSMIC/FORMOSAT-3 mission: Early results. B ull. Amer. Meteor.

Soc. 89, 1–21.

Healy SB, Wickert J, Michalak G, Schmidt T, Beyerle G (2007) Combined forecast impact of

GRACE-A and CHAMP GPS radio occultation bending angle proﬁles. Atmos. Sci. Lett. 8,

43–50.

König R, Zhu SY, Reigber C, Neumayer KH, Meixner H, Galas R, Baustert G, Schwintzer P (2002)

CHAMP rapid orbit determination for GPS atmospheric limb sounding. Adv. Space Res. 30(2),

289–293.

460 T. Schmidt et al.

Melbourne WG, Davis ES, Hajj GA, Hardy KR, Kursinski ER, Meehan TK, Young LE (1994) The

application of spaceborne GPS to atmospheric limb sounding and global change monitoring,

JPL Publication, 94-18, Jet Propulsion Laboratory, Pasadena, California.

Michalak G, Wickert J, Koenig R, Rothacher M (2007) Precise satellite orbit determination for

GPS radio occultation in near-real time (NRT), EGU2007-A-08740 abstract.

Reigber C (1998) GPS Atmosphere sounding: An innovative approach for the recovery of

atmospheric parameters, HGF Strategy fund proposal, Potsdam, Germany.

Wickert J, Reigber C, Beyerle G, König R, Marquardt C, Schmidt T, Grunwaldt L, Galas R,

Meehan TK, Melbourne WG et al. (2001) Atmosphere sounding by GPS radio occultation:

First results from CHAMP. Geophys. R es. Lett. 28, 3263–3266.

Assimilation of CHAMP and GRACE-A

Radio Occultation Data in the GME Global

Meteorological Model of the German

Weather Service

Detlef Pingel, Andreas Rhodin, Werner Wergen, Mariella Tomassini,

Michael Gorbunov, and Jens Wickert

1 Introduction: Data Assimilation

Prior to a numerical weather forecast run, the state of the meteorological model has

to be updated using observational information. This operation is called the assimila-

tion step in the process of weather forecasting. Common observational data sources

include in-situ data such as radiosondes, weather station measurements, data from

aircrafts and buoys. In the meantime, satellite remote-sensing measurements play a

major role i n weather prediction. Most remote-sensing observations are radiances

of a given wavelength at nadir. Radio occultation data, however, are a relatively new

source of observational information. Quantities that describe an occultation process,

like bending angles of the individual rays or the refractivity ﬁeld, depend directly on

the temperature and the humidity along the ray’s path. Therefore, radio occultations

can be assimilated in order to include additional information on temperature and

humidity into the model. They have been proven to be a valuable source of infor-

mation in atmospheric research and weather prediction (Wickert et al., 2009; Healy

et al., 2007b).

The information of radio occultations about the earth’s atmosphere is com-

plementary to radiance observations in several senses: whereas temperature and

humidity retrievals derived from radiances have a relatively low vertical resolu-

tion, the r adio occultation observations are sensitive to vertical structures on a much

smaller scale. Unlike radio occultation, radiance measurements generally fail in

cloudy areas (McNally, 2002). Another beneﬁcial property is the lack of fundamen-

tal biases, which is an advantage when taking into account the elaborate procedure

of radiometer calibration, and the nearly uniform global coverage.

At the German weather service, the assimilation of radio occultation data is per-

formed by the three-dimensional variational (3D-Var) assimilation method (Daley

and Barker, 2000), which was recently introduced as the operational global assimila-

tion scheme. The assimilation of observational data is done in a statistically optimal

D. Pingel (B)

Deutscher Wetterdienst, DWD, Offenbach, Germany

e-mail: detlef.pingel@dwd.de

461

F. Flechtner et al. (eds.), System Earth via Geodetic-Geophysical Space Techniques,

Advanced Technologies in Earth Sciences, DOI 10.1007/978-3-642-10228-8_41,



Springer-Verlag Berlin Heidelberg 2010