Inversion methods

This work has been funded by EPSRC/MTD and NERC. Solutions for a linearised problem have been developed and successfully compared with wave buoy measurements over a wide range of wave conditions during the NURWEC2, SCAWVEX, EuroROSE, SHOWEX experiments and WAVENET trials experiments. In the EuroROSE project a real-term version of the program was successfully implemented and an operational version has been tested during the WAVENET trials.

Solutions have also been found to the non-linear problem but for a number of reasons this has proved to be less useful than was hoped. The main problem, and one that will keep us busy for many years to come, is that the integral equation that is being inverted does not provide a good description of the measured backscatter in high sea conditions i.e. exactly the times of most interest for many applications.

We have found ways of getting round this problem – so that we can get good estimates of many wave parameters over a wide range of conditions – but we would like to develop a more general theory.

In collaboration with Professor Clive W Anderson, SCEOS and Probability and Statistics, we developed a technique to extract short wave directional parameters providing, amongst other things, robust estimates of wind direction.

We have developed innovative statistical techniques for evaluating the results of our inversion techniques and for comparing the measurements with wave buoys and wave models. A PhD student, funded by the EU and the Portuguese government, worked on detailed comparisons with the SWAN wave model developed at Delft University of Technology, the Netherlands.

More recently an EPSRC PhD student developed a neural network inversion method which shows some promise. In addition she contributed to work on bistatic radar modelling and inversion.