In the frame of the future satellite mission Meteosat Third Generation (MTG) undertaken by ESA, Thales Alenia Space,
as satellite prime contractor, is responsible for the design, validation and monitoring of the geometric image quality.
All final products delivered by the MTG mission will be geolocated on-ground by the Image Navigation Registration
(INR) process. This process estimates the geolocation of every acquired sample thanks to a Kalman filter based on
observables extracted from the images (e.g landmarks, stars) as well as auxiliary data such as orbit, attitude or scan
The paper presents the high-fidelity engineering tool developed to assess and analyze the future INR performances of the
system. Compared to previous Meteosat generations based on spinning satellites, the 3-axis stabilisation increases the
complexity of the INR model prediction by inducing high-frequency perturbations.
In order to estimate the INR filter behaviour, realistic sets of image observables are simulated. The simulation takes into
account all error sources affecting the pointing knowledge of each MTG instrument such as micro-vibrations, thermoelastic deformations, orbit estimation errors or instrument scan and spacecraft attitude knowledge performances. After simulating the INR process over the images, geometric performances as defined through MTG user specifications are assessed. Thus, the INR behaviour and the overall system performance can be predicted among different operational conditions. It is then possible to analyse the contribution of each perturbation to the final performances and to tune the INR filter with respect to the satellite behaviour.