Dutilleul, Pierre R.L.
Professor
Pierre Dutilleul’s academic background is in Mathematics and Statistics, and his research interests are in statistical inference (estimation and testing) in the temporal and spatial frameworks, in a variety of domains including Environment and Ecology, the soil sciences (e.g., seismology), and phytometry. Accordingly, he is Professor in the Department of Plant Science and Associate Member of the Department of Mathematics and Statistics and of the Bieler School of Environment at McGill University; he is also Adjunct Professor at Université de Sherbrooke. In Google Scholar, Dr. Dutilleul is most known (~900 citations) for his modified t-test for correlation analysis with spatial data, and several extensions of the test have been developed by his group since. Professor Dutilleul is also known for his innovative phytometric research work, in which his group is using a computed tomography (CT) scanner to collect 3-D spatial data on plant structures, and analyzes them statistically; this provided an interview to the Science Magazine and a Radio-Canada Découverte reportage; that work now includes the studies of soil and wood properties. Pierre Dutilleul has authored ~180 peer-reviewed publications and one book (“Spatio-Temporal Heterogeneity: Concepts and Analyses”, Cambridge University Press), and has coordinated from beginning to end the e-book project “Branching and Rooting Out with a CT Scanner” (Nature Publishing Group/Macmillan).
Pierre Dutilleul is also an Associate Member, Department of Mathematics and Statistics, Associate Member, Bieler School of Environment, and Adjunct Professor, Département de mathématiques, Université de Sherbrooke.
Active Affiliations
- 2011 to present: Editor-in-Chief, Environmental and Ecological Statistics, Springer
- 2009 to present: Member, Board of Directors of Centre SÈVE
- B.Sc. (Mathematics)
- M.Sc. (Statistics)
- D.Sc. (Mathematics)(Université catholique de Louvain, Belgium)
- December, 2020: Appears in Stanford University World Rating of Top 2% Scientists List
Spatio-temporal heterogeneity analysis
In statistical sense and simply put, heterogeneity may concern the mean or variance parameter of the distribution of a random variable, or be related to the autocorrelation function of a stochastic process. When the mean or variance parameter value is likely to change in space or time, or both, or when variability is measured from observations that are partially dependent on each other because they are autocorrelated, there is potential for a heterogeneity analysis (Dutilleul, 2011), which should include sound experimental design to begin with (Dutilleul, 1993a). This opens the door to a lot of interesting situations and problems! Dutilleul’s (1993b) modified t-test provides a valid solution to the problem of assessing the correlation between two autocorrelated spatial processes, and was followed by other modified t-tests and modified F-tests in the contexts of multivariate and multi-scale analyses (Dutilleul and Pinel-Alloul, 1996; Alpargu and Dutilleul, 2006; Dutilleul et al., 2008b; Dutilleul and Pelletier, 2011; 2017). Concerning efficient estimation and the decomposition of the variability contained in multivariate spatial datasets, the articles with geostatistical taste of Pelletier et al. (2004) and Larocque et al. (2007) paved the way to a solution based on the fitting of a linear model of coregionalization by estimated generalized least squares and the development of the method of Co-Regionalization Analysis with a Drift (CRAD; Pelletier et al., 2009a; 2009b) eventually. In a spectral instead of geostatistical approach, the method of multi-frequential periodogram analysis (MFPA; Dutilleul, 2001; 2011, Chapter 6) allows the decomposition of a time series, univariate or multivariate, into a number of periodic components, the number of periodic components as well as the period values being estimated in a stepwise procedure. I also have research interests in point pattern analysis, in planetary science (Dutilleul et al., 2009) and animal behavior (Bonnell et al., 2013), and long-term research interests in multidimensional statistics (Dutilleul, 1999; 2018; Manceur and Dutilleul, 2013). Current and recently completed research projects include works in statistical seismology and time series analysis with copulas.
Modern phytometry
My research activities in this domain started before 2003, with the development of a fractal dimension estimation procedure to quantify the structural complexity of plant branching patterns and thereby improve plant light interception models (Foroutan-pour et al., 2001), but they were boosted with the creation of the CT Scanning Laboratory for agricultural and environmental research on the Macdonald Campus of McGill University in Fall 2003, thanks to an NSERC Major Equipment grant (PI: Dutilleul) and the portion of a CFI grant (PI: Fortin) for the equipment of a computer room. The CT scanning equipment was recently renewed and expanded thanks to two CFI 9 grants, and there is now a macro-CT scanning section (CFI 9 grant PI: Geitmann) and a micro-CT scanning section (CFI 9 grant PI: Ghoshal).
Since the creation of the facility, my research group, generally in collaboration with other research groups for the applications, developed new procedures for the graphical, quantitative and statistical analyses of CT scanning data in a broad range of domains other than the medical one for which the macro-CT scanning equipment was originally designed. This research first included pioneering work (Dutilleul et al., 2005; 2008a; Lontoc-Roy et al., 2006; Han et al., 2008) and then, diversified and sophisticated works (Lafond et al., 2012; 2015; Dutilleul et al., 2014; 2015; Subramanian et al., 2015; Han et al., 2017), where “fractal” has become a keyword in two forms: mono-fractal and multi-fractal. Collaborating research groups are from McGill, U. de Sherbrooke, U. Laval, U. de Montréal, UQÀM, Canadian governmental departments (Agriculture and Agri-Food, Natural Resources), and Spain, Scotland outside Canada.
In Spatio-Temporal Statistics, I continue to “develop and apply”: develop, when there is a need for a new statistical method or model; apply, to extract and distill the fuzzy information hidden in a dataset.
- I continue to work with Prof. Yves Carrière (The University of Arizona) and American entomologists. The results obtained in our latest collaborative research project, with contributors from a resilience foundation and a multinational company, were published in PNAS (Carrière et al., 2020). I was in charge of the statistical analysis of spatio-temporal data, based on a mixed model for covariate analysis.
- My sabbatical leave in the Department of Statistics at UC Berkeley in 2013 gave me the opportunity to meet with Prof. Roland Bürgmann (Earth and Planetary Science) and Dr. Christopher Johnson. Among other things, our fruitful collaboration produced three articles (Dutilleul et al., 2015; 2020; 2021). The MFPA method plays a key role in our work and we were able to resolve several periodicities in Central California earthquake catalogs that reveal external periodic forcing.
- My adjunct professorship in the Département de mathématiques at U. de Sherbrooke allows me to co-supervise Master’s students with Prof. Taoufik Bouezmarni; two thesis projects were completed on the topic of time series analysis with copulas and two are under way/in preliminary stage. A first publication can be reported (Bégin et al., 2020).
- My long-term research collaboration with Prof. Joann Whalen (Natural Resource Sciences, McGill) reached a new milestone with the development of the spatio-temporal concept of persistence and its quantification in soil science in the Ph.D. thesis project of Tian Tian (Tian et al., 2021).
The next two projects are at the interface with Modern Phytometry
- With Prof. Louis-Paul Rivest and Dr. Nishan Mudalige (U. Laval), we are re-analyzing statistically some of the tree branching patterns presented in Dutilleul et al. (2015).
- Prof. Kunio Shimizu (Keio U.), Prof. Imoto Tomoaki (U. of Shizuoka) and I are collaborating on the development of a new statistical approach to modeling tree growth from wood CT scanning data.
In Modern Phytometry, research collaborations with Centre SÈVE members continue to occupy an important place. See, for example:
- the recent results obtained in the Biochar CT Scanning project with Prof. Donald Smith (Smith) and Dr. Ondrej Masek (The University of Edinburgh) and published or to be published (Han et al., 2020; Srocke et al., 2021);
- the start of the FRQNT Projet de recherche en équipe with Prof. François Belzile (U. Laval), after the completion of our Centre SÈVE Nouvelles Initiatives project;
- the ongoing Iceberg Lettuce CT Scanning project, conducted with Dr. Mamadou Lamine Fall (AAFC) and Ph.D. candidate Azza Larafa.