Weather Prediction and Data Assimilation
We do research to improve estimates of past, present, and future atmospheric and oceanic states. Given initial conditions, we use approximations to the governing equations to predict weather and climate. The models are built with millions of lines of code, running on some of the world’s fastest supercomputers. We consider both coarser resolution global models and higher-resolution cloud-resolving regional models.
The largest improvements in weather forecasting over the last decade have been attributable to data assimilation improvements. Data assimilation schemes generate initial conditions for forecasts by using observations to correct short-range forecasts. The corrected states provide initial conditions for subsequent forecasts.
However, more research is needed. For example, clouds are known to be a leading cause of prediction error. Cloud observations could help fix these errors, but few cloud observations are assimilated. Furthermore, the dream of allowing observations to teach model parameterisations of clouds has yet to be achieved.
Coordinator
Craig Bishop
Academic staff
We are currently unable to retrieve the Staff contact information requested.
Graduate researchers
We are currently unable to retrieve the Staff contact information requested.