Computational Statistics in Climatology
Scientific descriptions of the climate have traditionally been based on the study of average meteorological values taken from different positions around the world. In recent years however it has become apparent that these averages should be considered with other statistics that ultimately characterize spatial and temporal variability. This book is designed to meet that need. It is based on a course in computational statistics taught by the author that arose from a variety of projects on the design and development of software for the study of climate change, using statistics and methods of random functions.
What people are saying - Write a review
We haven't found any reviews in the usual places.
advection velocity analyzed anomalies ARMA autocorrelation function autoregressive models climate time series climatological coefficients computed considered correlation function corresponding covariance function covariance matrix determined diffusion equation digital filters diurnal cycle domain equal example Figure finite differences forecast accuracy forecast error formulas Fourier transform frequency instrumental variable interval latitude bands least squares linear regression main diagonal methodology multivariate non-zero nonstationary normalized standard deviation normalized variance number of observations observed and simulated observed data obtained parameter estimates periodogram point estimates Polyak polynomial degree possible presented random process regions regressive filters sample scheme shows simulated data smoothing spatial averaging spatial-temporal spectral and correlation spectral density spectral estimates spectrum stationary stationary process statistical dependence statistical structure statistically significant steps ahead stochastic models surface air temperature Table temperature time series tion two-dimensional univariate values variability variations white noise zero
Page vi - Because the spectrum estimation methodology is based on the choice of smoothing window, the book opens with a consideration of digital filters (Chapter 1). Questions of averaging (as the dominant statistical procedure in climatology) and fitting simple linear models are given
Page ix - climate fluctuations for both point gauges and spatially averaged data. In many cases, the closeness of climate fluctuations to white noise or to first-order multivariate AR models is discussed.