Summary
Non-linear models are often required in environmental applications, for example to incorporate seasonal effects. A wide variety of useful parametric forms is available, while nonparametric methods have the potential to offer further flexible extensions to any modelling situation. The aim of this paper is to incorporate this flexibility into non-linear models by allowing appropriate terms to vary smoothly over time. This uses the general structure of additive, semiparametric and varying coefficient models, within a non-linear setting. Data on water quality from the River Clyde are used as an example.
Similar content being viewed by others
7 References
Bowman, A.W., Azzalini, A., (1997). Applied Smoothing Techniques for Data Analysis Oxford University Press
Fahrmeir, L., Tutz, G. (2001). Multivariate statistical modelling based on generalized linear models Springer Verlag, New York, 2nd edition.
Gallant A.R. (1987). Non-linear statistical Models Wiley Series in Probability and Statistics
Green, P., Jennison, C. and Seheult, A. (1985) ‘Analysis of field experiments by least squares smoothing’, J.R. Statist. Soc. B, 47, pp. 299–315
Harvey, A. (1989). Forecasting structural time series models and the Kaiman filter Cambridge University Press
Hastie, T.J. and Tibshirani, R.J. (1990). Generalised Additive Models. Chapman and Hall, London.
Hastie, T.J. and Tibshirani, R.J. (1993). ‘Varying-coefficient Models (with discussion)’. J.R. Statist. Soc. Ser.B vol. 55, pp. 757–96.
Hirst, D.J. (1998). ‘Estimating Trends in Stream Water Quality With A Time-Varying Flow Relationship’. Austrian Journal of Statistics vol 27,1&2,39–48
Venables, W., Ripley, B. (1997). Modern Applied Statistics with S-Plus Springer Verlag, New York
Acknowledgements
We would like to thank Brian Miller from the Scottish Environment Protection Agency for supplying the Clyde data and providing helpful information on its context. Andrew McMullan also gratefully acknowledges the financial support of the Engineering and Physical Sciences Research Council through a research studentship.
Part of the work for this paper was developed for a presentation at the Euroworkshop on Statistical Modelling, funded by the European Union under grant EWStatModel, HPCF-CT-2000-00041, Event number 2. Financial support for attendance at this event is gratefully ackowledged.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
McMullan, A., Bowman, A.W. & Scott, E.M. Non-Linear and Nonparametric Modelling of Seasonal Environmental Data. Computational Statistics 18, 167–183 (2003). https://doi.org/10.1007/s001800300139
Published:
Issue Date:
DOI: https://doi.org/10.1007/s001800300139