FIELD (MAP) STATISTICS
J. L. Stanford, and J. R. Ziemke, Iowa State University department
of Physics and Astronomy, Ames, 50010
Field
(Map) Statistics. (CHAPTER 16 of STATISTICAL METHODS FOR
PHYSICAL SCIENCE, J. L. Stanford and S. B. Vardeman, eds., Academic
Press, 1994, Volume 28 in the series METHODS OF EXPERIMENTAL PHYSICS),
pages 457-479.
Abstract
Monte Carlo methods have been described for testing the statistical
significance of fields of a statistic. Examples were given using
several actual data time series from current atmospheric physics
research. The tests can be important when, as is often the case, the
measured data fields contain interdependencies that may give the illusion
of nonrandomness. Three different schemes are discussed for generating
the noise series used in Monte Carlo tests, and the data examples
utilized illustrate how the use of different schemes affect the outcome
of the statistical tests.