Christopher's Official Project title:
PhD in Adaptive Processing and Forecasting of Weather Radar Data
I'm looking at ways of using weather radar data for flood forecasting, mostly in real-time (ie forecasting floods during the storm itself). Rainfall intensities measured by weather radars can be quite inaccurate, so I shall nail down the 'sheet' of radar rainfall intensities by using a network of raingauges. All the data is supplied to the office in real-time, by courtesy of the Environmental Agency, over a data link from Warrington, Chesire.
Forecasts are made mainly by using applied model theory - Kalman filtering and Integrated Random Walk (IRW) techniques - and most of the work is done using the Matlab computer package with in-house model analysis programs. I have also built a whole family of transfer function models for the rainfall-runoff relationships, including splitting the observed runoff into quickflow, or 'flash' flow, and baseflow. It is possible to forecast both of these quantities using the Kalman Filter or IRW programs. Hopefully my work will lead to a forecasting model that is flexible and accurate enough to replace the current models used for actual flood forecasts in NW England.
Back to my homepage
[ University | CRES | Systems and Control ]