Abstract
An objective method for determining probabilities of surface temperature extremes is described herein. Least-squares linear regression equations have been developed to estimate temperatures that would be equaled or surpassed 1, 5 and 10% of the hours at any given location during the warmest and coldest months of the year. These equations are based on an index calculated from three generally available parameters: 1) mean monthly temperature (usually July and January), 2) mean daily maximum temperature for the month and 3) mean daily minimum temperature for the month.
