BEGIN:VCALENDAR VERSION:2.0 PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4// BEGIN:VEVENT UID:20260906T105215EDT-39463clkSP@132.216.98.100 DTSTAMP:20260906T145215Z DESCRIPTION:Title: Accommodating outlying observations in environmental spa tio-temporal processes.\n\nAbstract: In the analysis of most spatiotempora l processes in environmental studies\, observations present distributions that are not normal. Commonly\, some transformation is applied to the data and inference is performed at the transformed scale. Commonly\, the trans formation will have an impact on the description of the uncertainty at fut ure instants of time or unobserved locations of interest.\n In this talk I will discuss some of the projects I have been involved with in the last fi ve years that relax the assumption of normality of spatiotemporal processe s after some suitable transformation of the data. In particular\, I will f ocus on a recent proposal that models the variance law of multivariate dyn amic linear models. The proposed approach adds flexibility to the usual Mu ltivariate Dynamic Gaussian model by defining the process as a scale mixtu re between a Gaussian and log-Gaussian processes. The scale is represented by a process varying smoothly over space and time which is allowed to dep end on covariates. Analysis of artificial datasets show that the parameter s are identifiable and simpler models are well recovered by the general pr oposed model. The analyses of two important environmental processes\, maxi mum temperature and maximum ozone\, illustrate the effectiveness of our pr oposal in improving the uncertainty quantification in the prediction of sp atio-temporal processes.\n DTSTART:20230201T203000Z DTEND:20230201T213000Z LOCATION:Room 1140\, McGill College 2001\, CA\, QC\, Montreal\, H3A 1G1\, 2 001\, avenue McGill College SUMMARY:Alexandra Schmidt\, McGill University URL:https://www.mcgill.ca/mathstat/channels/event/alexandra-schmidt-mcgill- university-345738 END:VEVENT END:VCALENDAR