Wednesday 7th May 2014

Statistics and the environment

Lindsay Lee [Presentation] and Jill Johnson [Presentation]

Institute for Climate and Atmospheric Science, School of Earth and Environment, University of Leeds

Statistical analysis to understand the effect of model uncertainty in climate simulations

We will present work done in the School of Earth and Environment at the University of Leeds, applying statistical tools for parametric sensitivity analysis and constraint of computer models simulating important features of the climate. Our focus is the role of aerosols in climate change. These are small particles suspended throughout the atmosphere that affect the exchange of incoming and outgoing solar radiation either directly or by altering clouds. The resulting climate effect of aerosol is persistently one of the dominating uncertainties in the Intergovernmental Panel on Climate Change (IPCC) report with low scientific understanding. Huge investment in observations and more complex models of atmospheric aerosol have improved understanding of aerosol-cloud processes but the uncertainty in model estimates of aerosol concentrations and their effects on clouds have not been reduced. We have used a statistical analysis of a cloud simulator and a global aerosol model to understand the effects of uncertainties introduced by the modelling process. We can use this uncertainty information to target research in the right places and reduce model uncertainty, to simplify models and to quantify the value of observations in different parts of the world. This talk will show practical application of statistical methods including expert elicitation, experimental design, emulation, sensitivity analysis, cluster analysis and history matching.

The meeting will be held in Lecture Theatre 1 of the Roger Stevens Building at the University of Leeds, from 3:30-5pm with refreshments at 3pm in the Level 9 School of Maths foyer