A Search for a Solar Influence on the Skill of Weather Forecasts

Abstract

An attempt is made to find a relationship between solar variability and weather using the skill scores of an objective weather forecasting procedure as the basis of analysis. The hypothesis being tested rests on the nature of a numerical weather prediction model as an initial value problem and on the fact that the model does not include, either explicitly or implicitly, the effects of solar variability which occur between the time of the initial input data and the time of forecast verification. Therefore, solar disturbance, if it affects the weather, is expected to result in forecasts which are less skillful than average and the lack of solar disturbance, in more skillful forecasts. Geomagnetically disturbed and quiet days are used as measures of solar disturbance and solar quiet, respectively. In addition, the interplanetary magnetic sector structure is also used as a measure of organized solar activity. Although some of the results may be interpreted as indicating a real solar effect, that are internal inconsistencies which lead perforce to the conclusion that the case for a sun-weather effect is not proven.