LocationIfGI CIP pool

Objective: Explore spatial and temporal data representations
General description: Look at how different research communities have approached the representation of spatio-temporal data. (references)
Required back-ground knowledge:  Good working knowledge of R, experience of handling spatial or temporal data, possibly in R.
Software / R packages required: (to follow);

9:00–10:30 Representation in time and space - entities, support, processes,
separability, modelling
Summary: Observations in time present different representational issues from those in space. Temporal processes (for example longitudinal data analysis, time series analysis, irregular time series) are well-known, but mixed spatio-temporal processes are not easily accessible. We need to keep the questions of support and particularly change of support in mind. In addition to these issues, ontologies (in the GIScience meaning) are of key importance in using richer data, because different sciences approach observations in different ways. We also need to look at spatial survival models briefly. (slides, copies of papers)

10:30–11:00 Coffee break;

11:00–12:30 Spatio-temporal data analysis in R packages - status and outlook
Summary: Spatio-temporal data is represented - as we have seen - in many different ways, within and across sciences. Some areal data may be stacked by annual aggregates, other data is acquired from animal tracking, epidemics develop in time and space, environmental monitoring stations report data for temporal aggregates, etc. Some data sets have longitudinal or panel characteristics, others have just one observation in space and time. (slides)

12:30–14:00 Lunch break;

14:00–15:30 exercises (data and script), (package list), (package installation script), Summary: Areal panels, monitoring interpolation, surveillance, tracks, point patterns, ...

15:30–15:45 Coffee break;

15:45–17:00 exercises.