Extreme events are known to have major impacts on the environment and society. The focus of this research is to develop appropriate models that quantify changes in extreme and related events.
References:
Ghosh, S., Beran, J., Innes, J.(1997)
Nonparametric conditional quantile estimation in the presence of long memory. Student, vol. 2, No. 2. 109-117.
(The authors model the variability in some tree ring data and global climate data.
Global climate data source: Climate Research Unit, Univ. East Anglia)
Draghicescu, D., Ghosh, S. (2000)
Different aspects of quantile estimation for transformations of Gaussian processes. Abstracts, Invited talk, Annual symposium on Forecasting, Lisbon. 21-24 June 2000.
Ghosh, S., Draghicescu, D. (2002)
Predicting the distribution function for long-memory processes. International Journal of Forecasting. (Application to precipitation data). Data source: SMA.
For many observed time series, only partial knowledge may be available about the structure of the underlying probability model. An appropriate semiparametric model may be obtained by combining the existing knowledge in terms of parametric and nonparametric components.
References:
Beran, J., Ghosh, S. (1998)
Root-n-consistent estimation in partial linear models with long-memory errors. Scandinavian Journal of Statistics. Vol. 25, No. 2, 345-357.
(The authors examine global warming in the two hemispheres. Data source:
Climate Research Unit, Univ. East Anglia)
References:
Beran, J., Ghosh, S., Sibbertsen, P.(2003)
Nonparametric M-estimation with long-memory errors. International Journal of Forecasting. (Application to wind-speed data).
Beran, J.,Feng, Y., Ghosh, S., Sibbertsen, P.(2002)
On robust local polynomial estimation with long-memory errors. International Journal of Forecasting. (Application to wind-speed and financial data).
References:
Ghosh, S. (1996)
A new graphical tool to detect non-normality. Journal of the Royal Statistical Society, Ser. B.
Ghosh, S. , Beran, J. (2000) Comparing two distributions: The two sample T3 plot. Journal of Computational and Graphical Statistics.
Ghosh, S., Ruymgaart, F. (1992)
Applications of empirical characteristic functions in some multivariate problems. Canadian Journal of Statistics.
Splus codes:
Study of various spatial processes is important in many application areas in the environmental sciences. Our research on this topic includes collaboration with C. Frei (formerly at ETHZ; now at Meteotest, Zurich) on the spatio-temporal development of heavy precipitation and other events in Switzerland. Other areas of aplication include forestry and landscape ecology, where evaluation of large scale spatio-temporal trends are of interest.
References:
Ghosh, S.,Landmann G., Pierrat J.C., Müller-Edzards C. (1997)
Spatio-temporal variation in defoliation. In: C. Müller-Edzards, W. de Vries, J.W. Erisman (eds), Ten years of monitoring forest condition in Europe. Studies on temporal development, spatial distribution and impacts of natural and anthropogenic stress factors. Technical background report. Geneva and Brussels, United Nations Economic Commission for Europe / European Commission, pp. 35-50.
Ghosh, S., Wildi, O. (xxxx) Statistical analysis of landscape data: space-for-time, probability surfaces and discovering species.
Ghosh, S. (xxxx) Random fields based explanation for power-laws in species-area curves.