Aarde
the data center, consists of all tools for data analysis and synthesis The quality
assessment unit aims at comparing the results with ground truth, e.g. orbits,
regional geoid models, gravity and geoid profiles, and test field data. In addition,
it contains tools for statistical testing and post mission calibration The end product
unit will provide the user with various end products such as gridded geoid heights
and gravity anomalies, geoid slopes, and propagated error estimates. Finally,
the simulator allows to perform full-scale simulations prior and during the mission
Sensor
Level 0: ravmlata
Level 1a: data depacketised and sorted calibration data
Level 1b: calibrated corrected gravity gradients GPS
measurements linear accangular velocity
and acc GOCE orbit
Stochastic models
Simulator
Instrument and
satellite modelling
Environment
Data processor
Preprocessing (data gaps, frame transformation etc
Quick-look analysis, in orbit quality assessment procedures
Gravity field from SGG
Gravity field from GPS
Combined solution
Downward continuation
Regularlzation
Iteration
Combination with terrestrial data
Error propagation
Error-PSfrV
estimation
Error covarlance
analysis
Estimation of
gravity field
parameters
End products
Quality
assessment
Level 2: geopotential coefficients
geoid heights
gravity anomalies
geoid slopes
Comparison with
adopted model
error estimates
Comparison with ground truth
Statistical testing
Post mission calibration
Figure 10: Scheme of the GOCE data center
Full-scale simulations prior to the mission are important in order to investigate
whether or not the mission goals will be achieved depending on the mission and
system design. Full-fledged simulations of the GOCE mission are currently being
done in collaboration with the industrial prime contractor bv the SID consortium,
a cooperation between DEOS, the Dutch Space Research Organisation (SRON)
and the Institute for Astronomical and Physical Geodesy at the "technical University
of Munich (IAPG). The main goal is a realistic description of the quality of the
observations and a proper propagation of the observation errors to any type of
gravity field functional, such as potential coefficients, gravity anomalies, geoid
heights, and geoid slopes. In order to end up with a realistic error budget, the
various error sources have to be identified an a described, e.g. sensor errors (e.g.
a radiometer, star camera, GPS antenna and receiver), control unit errors (e.g.
drag and attitude control), and environmental effects (e.g. orbit, gravity field ond
non-conservative forces). Moreover, the interaction between sensors, control loops,
actuators, and other subsystems have to be taken into account ('closed-loop'
simulation).
The closed-loop simulation (figure 11starts with a given set of gravity gradients
and information about satellite position and orientation, disturbing forces, and a
13
