will speak on
Over recent decades, weather forecasting has evolved from a mainly
qualitative activity to a rigourous quantitative science. The
accuracy of weather predictions has increased steadily, and
continues to improve.
The changing climate will have major implications for humanity. It
is essential that we determine probable future changes with as much
precision as possible. The computer models for modelling climate change
rest on the same mathematical and physical foundation as the models
used for weather prediction.
As the range of prediction increases, the errors grow larger. For
predictions beyond a few days, probabilistic prediction is more
apposite than deterministic forecasting. This is achieved through
the ensemble approach.
In this presentation we will review the mathematical foundations of
modern numerical weather prediction and climate modelling. We will
describe the techniques used for assimilation of observational data
and for integration of the partial differential equations governing
the evolution of the atmospheric flow.
(This talk is part of the IMS September Meeting 2007 series.)