During the early studies of ESA's Hipparcos space astrometry mission, and while still a graduate student in 1976, he was "recruited" to the project by Erik Høg and thereafter played a crucial role in various aspects of the mission definition and in the data analysis. He set out the overall principles of the astrometric data reduction aiming to combine and solve together the attitude, the system and the astrometric parameters of the stars. This crucial '3-step procedure' was used successfully by the two consortia (NDAC and FAST) later entrusted by ESA with the Hipparcos data processing.
The principle of reconstructing space astrometric positions from one-dimensional observations carried out in the innovative Hipparcos sky scanning mode was fully original (and frequently questioned outside the project), and at the very limit of available computational power even by the end of the mission in 1997. The numerical principles had to be demonstrated, together with the solution’s statistical properties. Already by the end of 1976, Lindegren had produced a set of definitive technical notes and simulations showing how to obtain a 'rigid sphere' with all astrometric parameters from a scanning satellite. Innovation, efficiency, completeness, clarity, and mathematical rigour have been the hallmarks of his many and varied fundamental contributions to space astrometry since that time.
From 1990 Lennart Lindegren led the Consortium NDAC (Northern Data Analysis Consortium) sharing with FAST (led by Jean Kovalevsky) the data processing of Hipparcos.[2] In addition to the overall scientific coordinating responsibilities, he developed many of the approaches and algorithms related to the mission: his innovation, insight, and mathematical rigour impacted the optical and focal plane design, the instrument calibration, the scanning law, the attitude determination (and the associated 'dynamical smoothing'), the double star analysis (as observed via a signal modulated by a grid), the effects of chromaticity and thermal load fluctuations, the optimum combination of the NDAC and FAST catalogue solutions, and the link to an extragalactic reference frame.
He was member of ESA's Hipparcos Science Team for the entire duration of the project (1976–1997).
Erik Høg has written:[3] "A new era of my life began on 1 September 1973 when I returned to Denmark with my family of five, after 15 years in Hamburg. I had obtained a tenure at the Copenhagen University where I was going to work on the construction of automatic control of the meridian circle in Brorfelde. Very soon, however, I heard of a young student at Lund Observatory who worked alone on modernizing the old meridian circle there. I went to Lund and 'found' Lennart. A few years later, Andrew Murray, my old colleague and member of the Hipparcos science team, would say: 'Erik, the best you have ever done for astronomy was to find Lennart!' and I agreed". Later Høg writes: "Of his numerous papers I will only mention two. He wrote a paper on 'Photoelectric astrometry',[4] a subject I had proposed, where he systematically discussed the performance of methods for precise image location from observations. It remains a classical paper. The second paper to mention is about the rigidity of the celestial coordinate system obtained by the one-dimensional observations in a scanning satellite as TYCHO/Option A/Hipparcos. The question was asked in 1976 as mentioned above, but it took years before we had the answer which was affirmative. The study was led by Lennart and contains his brilliant mathematical analysis of the simulations, but he modestly left the position as first author to another person."[5]
