Monday 1 November 2010
Organiser: Andrew Stuart
All talks will be in Room B3.02 Mathematics Institute, Zeeman Building
12:30-14:00 Lunch in the Mathematics Institute Common Room
14:00-14:50 Andrew Lorenc (UK Met Office) Data Assimilation for Numerical Weather Prediction
15:00-15:50 Mike Christie (Heriot-Watt, Institute of Petroleum Engineering) Techniques for Calibrating Oil Reservoir Models to Data
- This talk will start by describing the problem of predicting the flow of oil, gas and water in hydrocarbon reservoirs: how the flows are modelled, the data needed to predict flow, and how some quantities needed for prediction are generally unknown. We will then examine ways of calibrating oil reservoir models to observed data, leading to links between the engineering application, mathematics, statistics, and computer science.
16:00-16:30 Tea in the Mathematics Institute Common Room
16:30-17:20 Martin Widmann (School of Geography, Earth and Environmental Sciences, Birmingham) Simulating past and future climate
- In the first part of this talk the differences between numerical weather forecasting and climate simulations will be discussed and examples for simulations of past and future climate will be presented. This will be followed by an overview on attempts to develop data assimilation methods that are suitable for assimilating empirical information about the climate of the last millennium. In the final part statistical downscaling and correction approaches will be introduced. These methods are based on statistical links between large-scale observed or simulated meteorological variables and local meteorological observations, and are applied to obtain estimates for climate variability that are more realistic than the direct model output.