A METHOD FOR PREDICTING THE PROBABLE FREQUENCY OF DAILY MINIMUM TEMPERATURES FROM SUMMARIZED DATA
ARMY NATICK LABS MA EARTH SCIENCES LAB
Pagination or Media Count:
A method for assessing the relative frequency and level of daily minimum temperatures in various parts of North America in winter months is developed in this paper from the summary 10-year records of 10 widely separated weather stations representative of the numerous variables that are involved in the occurrence of low temperatures. The method is based on the following four items usually found in climatic summaries a absolute minimum temperature b mean daily minimum temperature c mean daily maximum temperature d lengggth of record. The temperature and frequency estimates are achieved by use of a nomographic device with a grid of converted temperature values representing 36 varying degrees of asymmetry of temperature distribution, and a series of predictive curves designating the frequency of occurrence in days per month. The efficiency of the method is demonstrated by tests of internal consistency, also by application to 12 widely separated test stations in the United States, and to 6 Handbook stations and by varying length of summary records 10 to 70 years at a single station, Washington, D. C.