Statistical Evidence on the Influence of Sunspots and Volcanic Dust on Long-Term Temperature Records

Abstract
Over the years many claims have been made of relationships between climate and volcanic dust veils or sunspots. Sunspot proponents have claimed significant climatic variations on the time scales of 11 years, 22 years, and other multiples of the duration of the sunspot cycle (e.g., King, 1973a,b). Current hypotheses of a 22-year drought cycle in the Great Plains area is a contemporary example of such speculations (Thompson, 1973). Supporters of volcanic dust effects point to reportedly marked temperature drops alter large volcanic eruptions followed by a gradual return to preexplosion levels [e.g., Mitchell (1961) or Lamb (1970) or more recently, Oliver (1976)]. For the case of volcanic dust a plausible physical mechanism has always been at hand: the absorptive and scattering properties of volcanic particles. However, proposed mechanisms for the suggested sunspot-climate link have remained highly speculative, underlining the need for particularly critical assessments when such relationships are suspected. In this article we will attempt such an assessment for both sunspots and volcanic dust by examining all of the continuous, long-term temperature histories of more than 85 years held in the data library of the National Center for Atmospheric Research. By compositing and spectrally analyzing the observed records themselves, as well as comparing them with calculated temperature histories, we attempt to appraise the validity of certain proposed relationships as well as to determine any indications of previously undiscovered ones. Our results suggest that a volcanic signal can be weakly detected but that a sunspot influence cannot be reliably inferred from these data.