Relationships between mean and standard deviation of air temperature: implications for global warming

• Published in: Climate research 2002-11, Vol.22 (3), p.205-213
• Main Author: Robeson, Scott M.
• Format: Article
• Language: English
• Subjects: Climate change; Climate models; Global climate models; Global warming; High temperature; Inverse relationships; Probability distributions; Standard deviation; Statistical variance; Temperature distribution
• ISSN: 0936-577X; 1616-1572
• DOI: 10.3354/cr022205

Using data from the contiguous USA, the relationship between mean and standard deviation of daily air temperature was estimated on a monthly timescale from 1948 to 1997. In general, there is either an inverse relationship or a weak relationship between mean and standard deviation. For both daily maximum and daily minimum air temperature, the inverse relationship is spatially coherent for one-third to two-thirds of the contiguous USA for most months. The inverse relationship also is fairly strong, with typical reductions in standard deviation ranging from 0.2 to 0.5°C for every 1°C increase in mean temperature. A weaker, direct relationship between mean and standard deviation occurs in some northern states, primarily during spring and fall months. Using the predominant inverse and weak relationships as historical analogs for future climatic change suggests that interdiurnal variability of air temperature should either decrease or remain unchanged under warming conditions. Although the variability of air temperature may decrease or remain unchanged at most locations in the contiguous USA, the probability of extremely high air temperatures should still increase, depending on the magnitude of changes in mean air temperature and the nature of the variance response.