[Paper Review] The astrometric core solution for the Gaia mission. Overview of models, algorithms and software implementation

L. Lindegren, U. Lammers+4 more|Dec 18, 2011

Geophysics and Gravity MeasurementsEarth and Planetary Sciences41 references125 citations

TL;DR

This paper presents the mathematical and algorithmic framework for the Gaia astrometric core solution, a global least-squares estimation of ~500 million parameters including positions, parallaxes, proper motions, satellite attitude, and instrument calibration. The AGIS software system successfully demonstrated feasibility on simulated data, achieving sub-microarcsecond accuracy and scaling to real mission requirements with ~60 days of computation on a 10 Tflop/s machine.

ABSTRACT

The Gaia satellite will observe about one billion stars and other point-like sources. The astrometric core solution will determine the astrometric parameters (position, parallax, and proper motion) for a subset of these sources, using a global solution approach which must also include a large number of parameters for the satellite attitude and optical instrument. The accurate and efficient implementation of this solution is an extremely demanding task, but crucial for the outcome of the mission. We provide a comprehensive overview of the mathematical and physical models applicable to this solution, as well as its numerical and algorithmic framework. The astrometric core solution is a simultaneous least-squares estimation of about half a billion parameters, including the astrometric parameters for some 100 million well-behaved so-called primary sources. The global nature of the solution requires an iterative approach, which can be broken down into a small number of distinct processing blocks (source, attitude, calibration and global updating) and auxiliary processes (including the frame rotator and selection of primary sources). We describe each of these processes in some detail, formulate the underlying models, from which the observation equations are derived, and outline the adopted numerical solution methods with due consideration of robustness and the structure of the resulting system of equations. Appendices provide brief introductions to some important mathematical tools (quaternions and B-splines for the attitude representation, and a modified Cholesky algorithm for positive semidefinite problems) and discuss some complications expected in the real mission data.

Motivation & Objective

• Address the challenge of processing ~1 billion astrometric observations from Gaia to produce a high-accuracy celestial reference frame.
• Develop a global iterative solution that simultaneously estimates astrometric parameters, satellite attitude, and instrument calibration.
• Ensure numerical robustness and computational feasibility for processing up to 100 million primary sources with high precision.
• Design a scalable software system capable of handling the massive data volume and complex interdependencies in Gaia's data.
• Provide a foundation for the astrometric reference frame and instrument calibration essential for all subsequent data processing in the Gaia mission.

Proposed method

• Formulate the astrometric core solution as a global least-squares adjustment of ~500 million unknowns, including source positions, parallaxes, proper motions, attitude, and calibration parameters.
• Implement an iterative solution process divided into distinct blocks: source, attitude, calibration, and global updating, with auxiliary processes like frame rotation and primary source selection.
• Use quaternions for efficient and stable satellite attitude representation and B-splines for smooth temporal interpolation of attitude and calibration parameters.
• Apply a modified Cholesky algorithm to solve the positive semidefinite system of normal equations arising from the least-squares problem.
• Structure the solution to allow incremental refinement, starting from a priori estimates close to the final solution to reduce required iterations.
• Validate the approach using simulated data for 2 million primary sources, demonstrating numerical stability and convergence.

Experimental results

Research questions

• RQ1 How can a global least-squares solution be efficiently and robustly implemented for the simultaneous estimation of 500 million astrometric and instrumental parameters?
• RQ2 What mathematical models and numerical algorithms are required to ensure sub-microarcsecond accuracy in the astrometric solution for 100 million primary sources?
• RQ3 How can the computational complexity of processing Gaia’s massive data stream be managed within feasible time and resource constraints?
• RQ4 What role do quaternions and B-splines play in enabling stable and accurate representation of satellite attitude and calibration over time?
• RQ5 To what extent can the iterative solution converge reliably when starting from approximate initial estimates, given the scale of the Gaia mission?

Key findings

• The AGIS software system successfully implemented the global iterative solution, demonstrating numerical validity and robustness on simulated data with 2 million primary sources.
• The solution achieved an RMS uncertainty in the attitude parameter γ of approximately 3×10⁻⁵, consistent with formal error estimates and the true parameter value.
• Processing 135 iterations on a 14-node IBM cluster (0.65 Tflop/s) took nearly 6 days, corresponding to ~3×10¹⁷ floating-point operations.
• Scaled to the full mission with 10⁸ primary sources, the total computational workload was estimated at ~1.5×10¹⁹ flop, requiring ~60 days on a 10 Tflop/s machine.
• The solution is feasible due to convergence from a close initial estimate, potentially reducing the number of required iterations.
• All fundamental components of AGIS are in place, and the system is expected to meet Gaia’s ambitious astrometric accuracy goals despite real-world data complexities.