Program

The scientific program will feature sessions with talks on the latest advancements in theory, methods and applications. It will include keynote presentations, invited presentations and contributed papers and posters.

The scientific program of Friday (June 15) will be co-organized with the LEFE-Cerise project and will focus on topics relevant to this project.

A satellite event will be held on June 12, 2018 (Tuesday): a workshop providing an introduction to geostatistical analysis of spatio-temporal data with R, given by members of INRA's RESSTE network. The registration fee for this workshop is 50€.

Short course:

An introduction to geostatistical analysis of spatio-temporal data with R

Organizers: Denis Allard, BioSP, INRA; Liliane Bel, AgroParisTech; Edith Gabriel, Université d'Avignon; Thomas Opitz, BioSP, INRA; Eric Parent, AgroParisTech

When: June 12, 2018 (Tuesday).
More details on the schedule will follow.

Where: Montpellier University.
Details on the exact location will follow.

Participation to the workshop must be indicated in the registration form. Participants are strongly encouraged to bring their own laptop computer with the latest version of R and Rstudio installed. We will provide extensive R code covering the full work-flow. A list of all required R packages will be disseminated to participants in due time.

Summary of workshop content:

We present an overview of (geo-)statistical models, methods and techniques for the analysis and prediction of continuous spatio-temporal processes. We cover the Gaussian process approach, very common in spatial statistics and geostatistics. We leave aside the wide field of Bayesian techniques.

We focus on R-based implementations of numerical procedures and roll up a guiding thread exemplified on a large real air pollution data set. The target variable is the daily mean PM10 concentration predicted thanks to a chemistry-transport model and observation series collected at monitoring stations across France in 2014. We cover the full work-flow from importing data sets to the prediction of PM10 concentrations with a fitted parametric model, including the visualization of data and dependence structures, estimation of the parameters of the spatio-temporal covariance function and model selection. We conclude with some elements of comparison between the most relevant packages that are available today and some discussion for future developments.

The workshop is organized in four successive sessions (each around 1h30):

  • Session 1: Handling and importing large spatio-temporal data using structured objects; projection coordinate systems for geolocated data.
  • Session 2: Visualizing data according to their temporal, spatial or spatio-temporal structures.
  • Session 3: Statistical inference for spatio-temporal models: method of moments; maximum likelihood, pairwise composite likelihoods.
  • Session 4: Prediction and validation.