Assessing the potential effects of environmental hazards on both human health and the environment requires models based on realistic assumptions and the utilisation of data from multiple sources. These data may be misaligned in both space and time and will have different data generation mechanisms that will lead to different error structures, each of which may vary over both space and time.

Baysian hierarchical models provide a flexible framework in which to combine information from multiple sources and allow uncertainty from both the inputs and the modelling process to be acknowledged in a coherent manner. The use of the complex models together with `big data' will necessitate consideration of the computation required to perform inference. In this course, we will consider;

\begin{itemize}

\item the need for spatio-temporal modelling in environmental research,

\item Baysian hierarchical models as a framework for combining information from multiple data sources over space and time,

\item methods for performing Baysian interference with high dimensional data, including approximate Baysian inference using Integrated Nested Laplace Approximations (INLA),

\item examples of the use of data from multiple sources in addressing substantiative questions in environmental research.