Weather Prediction, Frontogenesis and Ensemble Forecasting

Since the 1950s numerical models atmospheric dynamics have allowed improved weather prediction. Since the 1960s numerical models of the coupled ocean-atmosphere system have been developed to predict climate changes of longer time scales. These models have nonlinear dynamics and have been shown to be sensitive to initial conditions. This sensitivity is not constant, however, but depends on the state of the model. It is described as state-dependent or flow-dependent predictability.

Since the early 1990s operational weather centres have used a technique called ensemble forecasting in an attempt to estimate the flow-dependent predictability of the atmosphere. More recently this type of ensemble approach has been introduced into climate modelling in an effort to establish the sensitivity of these models to parameters and initial conditions.

In the Oxford Environmental Mathematics Group we are investigating methods for constructing and interpreting ensemble forecasts. We also study experimental design and interpretation for climate modelling experiments. Collaborators on these projects include the European Centre for Medium Range Weather Forecasting and climateprediction.com.

People working in this area within OCIAM are

• Dr Lenny Smith

• Dr Irene Moroz

• Dr Patrick McSharry

• Dr John Norbury