Much of my philosophy and activities are defined by my research, or at least are in harmony with it.

My main interest and background is in developing methods and tools for time series analysis and in general - analysis of uncertain dynamic and non-linear systems. The general thread in the apparently wide-ranging application areas I am involved in, is also modelling of complex natural dynamic systems while not denying the underlying uncertainties. I am also interested in the effect of scale (both spatial and temporal) on the modelling process and model structure.

The other fundamental element in these modelling exercises is their Data-Based Mechanistic character, whereby the model structure is not assumed, but identified using objective statistical criteria. Complexity of the model that is identifiable and useful depends strongly on the available data - we can only model the modes of behaviour of the system under investigation that are present and 'visible' in the data. This approach to modelling, proposed and introduced by Peter Young, is what brought me to Lancaster in the first place, and I am delighted to have been able to develop my own contributions to it during the last nearly three decades.

Development of new methods of non-linear systems modelling and control with the current threads being (for details see the publications on my University maintained page):

1. -Multivariable State Dependent Parameter methods
   

2. -Spectral decomposition of signals and data, mostly relating to environmental processes
   

3. -Rainfall-Runoff and catchment dynamics, including the original Reverse Hydrology approach developed in the last few years
   

4. -Detection of change in environmental data
   

5. -Analysis of trends, cycles and interdependencies in long term environmental records ; these methods have been recommended in the 2007 UNEP 'Air' Report.
   

6. -Ongoing development of CAPTAIN Toolbox - a Matlab library developed over the years.