Search Results

You are looking at 11 - 13 of 13 items for

  • Author or Editor: H. C. S. THOM x
  • Refine by Access: All Content x
Clear All Modify Search
H. C. S. THOM and MARCELLA D. THOM

Abstract

A t-test is given to test monthly temperature normals, and a new test related to the F-test is developed for testing monthly precipitation normals. The theory involving the two errors committed in using such tests is discussed. This theory is used to set the probabilities of type 1 and 2 errors at predetermined levels. The tests are applied to examples from the normals revision program.

Full access
H. C. S. THOM

Abstract

Methods and charts are presented for estimating seasonal degree-day probabilities and quantiles for any location in the United States. Confidence limits for these are given enabling the engineer to judge the accuracy of these estimates and to apply them more effectively. A method for the determination of confidence limits for normals is also given. Examples for Detroit, Mich., and Washington, D. C., are worked out using the methods presented.

Full access
H. C. S. THOM and R. H. SHAW

Abstract

A set of freeze thresholds is defined which replaces the loosely defined element “killing frost.” Freeze series are tested and found to be random. This makes it possible to fit frequency distributions. It is hypothesized that these distributions are normal and a test of significance verifies this. A hypothesis that the variance parameter is constant over Iowa is tested and found to be true. It is shown that spring and fall freeze are independent and this results in a simple form of freeze-free period distribution. The calculation of probabilities of freeze and freeze-free season are explained.

Full access