Empirical climate change studies are more beneficial when they go beyond simply reporting fluctuations in average temperature. ideally dynamic causes of the change should also be addressed. This strategy allows conceptual models of the associated physical-change mechanisms to be developed. The analysis of air masses can play a central role in this approach. Changes in their distribution and characteristics point to modifications in the mid-latitude general circulation. The integrated method air-mass classification scheme was applied to this problem in the north central United States (NCUS) over the 1958-1992 period. Monthly averages for relative frequency, 850-hPa temperature, and 850-hPa dewpoint temperature were calculated for every air mass in each season. There were major changes over the study period during spring and summer and minor changes in the other two seasons. In winter, northern stations showed a slight warming of continental (equivalent to continental polar) air. During spring, tropical (equivalent to maritime tropical) air increased in frequency, but with no change in its average characteristics. In summer, polar air-a mixture of continental polar and maritime polar-was less frequent and tropical air more frequent. Dry tropical (equivalent to continental tropical) air prominently increased temperature, but did not change frequency. During autumn, tropical air was less frequent in the west and more frequent in parts of the eastern NCUS. The spring and summer air-mass variations suggest that western U.S.A. 500-hPa troughs increased in frequency over the study period, representing a subtle but significant structural climate change. Eastern portions of the NCUS are now experiencing more hot and humid days in general and increased thunderstorm activity in the warm season in particular than during the 1960s.
