Lennart Bondesson was a distinguished statistician whose career spanned several Swedish universities, leaving a lasting impact on research and education in mathematical statistics. Known for his deep expertise, generous mentorship, and intellectual curiosity, he contributed significantly to probability theory, sampling methodology, and statistical applications.
In his free time, Lennart enjoyed spending time in nature.
Image Private photo19 April 1944 – 29 June 2024
Lennart Bondesson was born in Helsingborg and grew up in Hova, Västergötland. After completing his secondary education in Mariestad, he studied mathematics, mathematical statistics, and physics at Lund University, earning a bachelor’s degree in 1965. He continued with doctoral studies in Lund under the supervision of Gunnar Blom, which led to a PhD in mathematical statistics in 1974. His dissertation was titled “Characterizations of location and scale parameter families of distributions through optimality properties of certain estimators.”
After a few years at Statistics Sweden (Statistiska Centralbyrån), Agricultural Division, he returned to Lund University as a research assistant (1977–1980), with a brief stay at Stockholm University. He then served as a senior lecturer at Umeå University before becoming professor of forest mathematical statistics at the Swedish University of Agricultural Sciences in Umeå (1983–1994). After five years at Uppsala University (1994–1998), he returned to Umeå University as professor of mathematical statistics (1999–2012), and later as professor emeritus until his passing.
Lennart was part of the golden generation of 6–8 mathematical statisticians born in the early 1940s, all of whom held professorships and contributed significantly to the growth of the field. In Sweden, the Nordic countries, and the broader statistical community, many knew Lennart well and remember his curious, knowledgeable, and often unusual questions and reflections during presentations and discussions.
One of Lennart’s major research interests was the characterization of probability distributions and infinitely divisible distributions. He wrote the book “Generalized Gamma Convolutions and Related Classes of Distributions and Densities” (1992), inspired by Olof Thorin’s work from the 1970s. He contributed many new results in the field and formulated several open problems in his book. Sampling was another area of research where Lennart was active, and several of his PhD students completed their dissertations in this field. He was involved in the Baltic-Nordic-Ukrainian network on sampling methodology.
Lennart was interested in both theory and applications. He made a strong impression as a vibrant teacher and a valued supervisor and consultant to many researchers and doctoral students. During his time at the Swedish University of Agricultural Sciences, his collaborations led to publications in forest genetics, silviculture, and forest inventory.
He was a frequent contributor to articles in the Nationalencyklopedin (NE) and Lexikon 2000, and wrote many statistics compendiums used in teaching. From 2001 to 2003, he served as editor-in-chief of the journal Scandinavian Journal of Statistics.
In his spare time, Lennart enjoyed being in nature. He was a mushroom expert who often took colleagues and friends on mushroom excursions. He was an avid outdoor swimmer year-round, even when the ice had settled, and took long cycling holidays, sometimes all the way from Umeå to Skåne.
With his broad and deep knowledge of probability theory and statistical inference, Lennart was someone many turned to for help and advice. He was generous with his expertise and was a highly appreciated mentor. At a staff day in 2023 at the Department of Mathematics and Mathematical Statistics at Umeå University, he gave a lecture on the history of mathematical statistics in Sweden. Fortunately, it was recorded and remains a living memory of Lennart.
By Peter Anton, Tom Britton, Sara Sjöstedt de Luna, Göran Ståhl and Jun Yu.
Translated from the original Swedish.