Development of Regional Extreme Model Atmospheres by Regression Methods.
NAVAL POSTGRADUATE SCHOOL MONTEREY CALIF
Pagination or Media Count:
A group of stations in the North American Arctic region have been analyzed for statistical determination of temperatures at mandatory pressure levels. For each station the temperature at a key level, peculiar to that station, has forced in at the first step and retained at each subsequent step, in the development of stepwise regression equations which predict temperatures at the mandatory levels. In general, eight-step predictions, in terms of interlevel thicknesses, were found to give optimum specification of the desired temperatures. The best estimate of the regional atmosphere, which is conditionally dependent upon the existence of an extreme 1 probability of the forcing level temperature, is obtained with a high degree of confidence. Author