1949Transactions American Geophysical UnionRequires access

Prediction of sunspot numbers

A. G. McNish, J. Virginia Lincoln

Open publisher page 130 citations

Abstract

A formula for predicting smoothed annual sunspot numbers is developed. A first approximation to the prediction of a future value in a cycle is the mean of all past values for that part of the cycle. This estimate can be improved by adding to the mean a correction proportional to the departure of earlier values of the cycle from the mean cycle. These correction factors are determined by the method of least squares. Sunspot data for 1834 through 1943 are used, and statistical reasons for rejecting data of earlier years are presented. Reliability of the method is discussed on the basis of the standard deviations of predictions made for each of the years of the ten cycles in question, using all the prediction formulas developed. The prediction of three‐month means is also discussed.

About this research paper

What this paper is about

A formula for predicting smoothed annual sunspot numbers is developed. A first approximation to the prediction of a future value in a cycle is the mean of all past values for that part of the cycle. This estimate can be improved by adding to the mean a correction proportional to the departure of earlier values of the cycle from the mean cycle. These correction factors are determined by the method of least squares. Sunspot data for 1834 through 1943 are used, and statistical reasons for rejecting data of earlier years are presented. Reliability of the method is discussed on the basis of the standard deviations of predictions made for each of the years of the ten cycles in question, using all the prediction formulas developed. The prediction of three‐month means is also discussed.

Why it matters

OpenAlex reports 130 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

A formula for predicting smoothed annual sunspot numbers is developed. A first approximation to the prediction of a future value in a cycle is the mean of all past values for that part of the cycle. This estimate can be improved by adding to the mean a correction proportional to the departure of earlier values of the cycle from the mean cycle. These correction factors are determined by the method of least squares. Sunspot data for 1834 through 1943 are used, and statistical reasons for rejecting data of earlier years are presented. Reliability of the method is discussed on the basis of the standard deviations of predictions made for each of the years of the ten cycles in question, using all the prediction formulas developed. The prediction of three‐month means is also discussed.

Key concepts: Sunspot number, Standard deviation, Mathematics, Sunspot, Statistics, Reliability (semiconductor), Mean value, Applied mathematics

Related papers

Back to paper searchBrowse research topicsOriginal source
Prediction of sunspot numbers — Research Paper | ScholarLens