Euclidean Distance Based Permutation Methods in Atmospheric Science

  • Authors:

  • Paul W. Mielke, Jr.;Kenneth J. Berry

  • Affiliations:

  • Department of Statistics, Colorado State University, Fort Collins, CO 80523-1877, USA. mielke@lamar.colostate.edu;Department of Sociology, Colorado State University, Fort Collins, CO 80523-1784, USA. berry@lamar.colostate.edu

  • Venue:
  • Data Mining and Knowledge Discovery
  • Year:
  • 2000

Quantified Score

Hi-index 0.00

Visualization

Abstract

The majority of existing statistical methods inherentlyinvolve complex nonmetric analysis spaces due to their least squaresregression origin; consequently, the analysis space of suchstatistical methods is not consistent with the simple metricEuclidean geometry of the data space in question. The statisticalmethods presented in this paper are consistent with the data spacesin question. These alternative methods depend on exact andapproximate permutation procedures for univariate and multivariatedata involving cyclic phenomena, autoregressive patterns, covariateresidual analyses including most linear model based experimentaldesigns, and linear and nonlinear prediction model evaluations.Specific atmospheric science applications include climate change,Atlantic basin seasonal tropical cyclone predictions, analyses ofweather modification experiments, and numerical model evaluations forphenomena such as cumulus clouds, clear-sky surface energy budgets,and mesoscale atmospheric predictions.