Lise Manchester Award 2024

Jonathan Jalbert
Christian Genest
Luc Perreault
Lise Manchester Award

Jonathan Jalbert (Polytechnique Montréal), Christian Genest (McGill University), and Luc Perreault (Institut de recherche d’Hydro-Québec) are the recipients of the 2024 Lise Manchester Award. This award is given every other year by the Statistical Society of Canada to commemorate the late Dr Lise Manchester’s abiding interest in using statistical methods to study matters of relevance to society. The award recognizes excellence in statistical research that helps guide public policy in Canada.

Jonathan (PhD, 2016, Université Laval and INP ENSE3, Grenoble) is an Associate Professor of Mathematics and Industrial Engineering at Polytechnique Montréal. He uses extreme-value theory, spatial statistics, and Bayesian methods to analyze large data sets arising from climatic models in order to assess flood risks, determine the vulnerability of infrastructures to climatic risks, and update the dimensioning criteria of public structures in nonstationary climate conditions.

Christian (PhD, 1983, University of British Columbia) is Professor and Canada Research Chair in Stochastic Dependence Modeling at McGill University. Well known for his pioneering work in extreme-value theory and copula modeling, he has long been interested in applications of statistical methods to insurance, finance, and hydrology. In recent years, he has turned his attention to the design of Bayesian hierarchical models for the frequency estimation of rare events.

Luc (PhD, 2000, ENGREF Paris and INRS-ETE, Québec) is a Senior Researcher and Project Manager at Hydro-Québec’s Research Institute in Varennes. His expertise includes time series analysis, extreme-value theory, and computational Bayesian methods. He has published extensively on issues relating to frequency analysis, probabilistic forecasting, statistical post-processing of ensemble weather forecasts, and regime switching models for hydrometeorological data.

Christian, Luc, and Jonathan started collaborating when the latter joined McGill as a postdoctoral fellow in 2016. Together with colleagues and students, they have since worked together on various projects involving the analysis and modeling of extreme meteorological events in the context of climate change. Some of their joint work includes an estimation of the return period of the devastating 2011 flood in the Richelieu Valley, the development of a new model for the statistical post-processing of ensemble weather forecasts, and a predictive analysis of baseflow in the Rivière des Prairies, which is a major source of drinking water for the Montréal metropolitan area.

The award highlights more specifically the creative nature and impact on Canadian public policy of their recent paper published in the Journal of Agricultural, Biological and Environmental Statistics (vol. 27, 2022). In this article, Jonathan, Christian, and Luc describe a hierarchical Bayesian strategy for spatial modeling of extreme precipitation anywhere in Canada which takes into account some orographic variables and a reconstruction of historical meteorology from the Canadian Regional Climate Model (CRCM).

Based on complex equations that represent the primary physical interactions in the atmosphere, ocean, ice and Earth’s surface, the CRCM provides synthetic data on temperature, precipitation, humidity, wind, etc. on a fine grid in space and time. The innovative use of this information, which involved major technical challenges, allowed Jonathan, Christian, and Luc to produce spatially coherent parameter estimates for generalized extreme-value distributions for heavy precipitation and, in turn, valid intensity-duration-frequency (IDF) curves for any location in Canada.

At present, engineers coast to coast use the IDF curves produced by Environment and Climate Change Canada (ECCC) to design infrastructures such as urban drainage systems. However, these curves are available only at data collection sites. The Canadian network of meteorological stations being sparse, except in highly populated areas, users are then forced to resort to IDF curves corresponding to sites that are located sometimes tens or hundreds of kilometers away. In addition, the curves are neither spatially coherent nor accounting for climate change. The approach developed by Jonathan, Christian, and Luc has the potential to resolve these issues and has already been integrated into Hydro-Québec’s operations. In partnership with ECCC, a project is currently underway to enhance this methodology and extend its use.

A nascent version of this approach had already been used by Jonathan and Christian, together with PhD student Nicholas Beck and Mélina Mailhot, professor of actuarial science at Concordia University, to quantify the magnitude of extremes surges on the Atlantic Coast of Canada. This work, which was used by a large Canadian insurance company to determine the premiums of its first coastal flood protection, received the Best Presentation Award at the 11th International Conference on Extreme Value Analysis and the Wiley-TIES Award for the best article published in Environmetrics in 2020.

The citation for the award reads: 

The citation for the award reads: “To Jonathan Jalbert, Christian Genest and Luc Perreault, for their sustained efforts to develop and implement complex models for extreme weather phenomena, and in recognition of the influence on Canadian public policy of their most recent innovation enabling interpolation of precipitation extremes over vast domains through the incorporation of auxiliary data from climate model runs.”

The nomination package was assembled by François Bellavance (HEC Montréal).