Subsampling confidence intervals for parameters of atmospheric time series: block size choice and calibration

Abstract

Problems of practical implementation of the computer intensive subsampling methodology are addressed by Monte Carlo simulations of a situation typical for atmospheric time series. The motivating data were collected under Lake-Effect Snow Studies Project in the winter of 1983–1984 over Lake Michigan. Certain enhancements of subsampling methodology are suggested specifically on the issue of optimal block size choice.

Keywords:

• Subsampling
• Nonparametric estimation
• Time series
• Confidence intervals
• Atmospheric data analysis

Acknowledgements

The research of Alexander Gluhovsky was funded by the National Science Foundation grant no. ATM-9909009 and by the James S. McDonnell Foundation.