Event Details

This workshop will discuss the implementation of novel statistical methods (e.g. multivariate emulators) to improve the computational efficiency of complex environmental models.

Models that are used to simulate different parts of the earth system are essential tools for overcoming environmental challenges (e.g. climate change or air pollution). However, these models are often slow to run, meaning that they are limited in terms of: (i) the level of spatial and temporal resolution they can operate at; (ii) the number of runs that can be carried out for comprehensive quantification of model uncertainty. For example, a typical global atmospheric chemistry-climate model may take 24 hours of computer time to complete a one-year simulation, half of which is used in solving the extensive system of differential equations needed to represent the chemistry. A solution is to replace the slowest part of the model with a simpler statistical model called a Gaussian process emulator. However, carrying this out presents a major challenge: the outputs of the part of the complex model that we are emulating are typically multivariate. The covariances between the different output variables cannot be ignored – as is commonly done in current uses of emulators – because these outputs are used for further calculations in other parts of the model.

The purpose of the workshop is to discuss different approaches for building multivariate output emulators that are computationally efficient. The outcome of the meeting is to have a plan for how this area of research can be developed further; this may include discussion of a possible grant proposal, either from statistical methods perspective or an applications perspective. The workshop is particularly open to statisticians and environmental modellers with an interest in using emulators in an environmental modelling setting. However, we welcome anyone who is interested and may be able to bring a fresh perspective to these modelling challenges. No prior experience of using emulators in practise is required.

For further information prior to registering for this event, please e-mail Amanda Fenwick (a.fenwick@lancaster.ac.uk) or Ed Ryan (edmund.ryan@lancaster.ac.uk).

Ed is a senior research associate at Lancaster University. His current research uses Gaussian process emulators to carry out sensitivity analysis, uncertainty analysis and model calibration of atmospheric chemical transport models. For more information, please see http://www.lancaster.ac.uk/lec/about-us/people/edmund-ryan.

Oliver is Professor of Atmospheric Chemistry at Lancaster University. His research involves developing and applying numerical models of atmospheric processes to learn more about atmospheric composition, chemistry and transport, and to understand how natural and anthropogenic emissions of trace gases affect regional air quality and global climate. For more information, please see http://www.lancaster.ac.uk/lec/about-us/people/oliver-wild.

Please register via the Eventbrite site.