Key points are not available for this paper at this time.
Northern Hemisphere surface-synoptic and radiosonde data from November 1961 through February 1962 are used to develop diagnostic relationships between surface-observed variables at a single station and the dewpoint spread at the 850-, 700-, 500- and 400-mb levels above that station. The approach consists of two steps: 1) the isolation within a decision-tree framework of those cases for which individual surface-observed variables yield highly reliable estimates of upper-level humidity, and 2) the application of a statistical technique (Regression Estimation of Event Probabilities) to the remaining cases to derive equations yielding probabilities of occurrence of specified categories of dewpoint spread. This approach yields useful diagnostic information with a quantitative measure of reliability.
Chisholm et al. (Thu,) studied this question.