Title: Investigating spatial scan statistics for multivariate functional data
Authors: Camille Frevent - University of Lille (France) [presenting]
Mohamed Salem Ahmed - University of Lille (France)
Sophie Dabo - University of Lille (France)
Michael Genin - University of Lille (France)
Abstract: In some research fields, such as environmental surveillance, pollution sensors are deployed in a geographical area. In a context where these sensors measure simultaneously the concentrations of many pollutants at regular intervals over a long period of time, environmental experts may search for environmental black spots, that can be defined as geographical areas characterized by elevated concentrations of pollutants. To this end, the development of spatial scan statistics for multivariate functional data indexed in space is very relevant. The proposed methods are derivated from a MANOVA test statistic for functional data, an adaptation of the Hotelling $T^2$-test statistic, and a multivariate extension of the Wilcoxon rank-sum test statistic. The performances of the methods are investigated through a simulation study and they are applied on multivariate functional data to search for spatial clusters of abnormal daily concentrations of air pollutants in the north of France in May and June 2020.