(1994)] showed a large-scale anticyclonic gyre in thegeneral vicinity of the gyre shown in Fig. 1C. Howev-er, Reid’s gyre and the one depicted in Fig. 1C havesubstantial differences in their vertical structure andhorizontal substructure.7. Southward flow of North Atlantic Deep Water on theeastern side of the Mid-Atlantic Ridge is supportedby P. M. Saunders [J. Mar. Res. 40, 641 (1982)], S.Gana and C. Provost [J. Mar. Syst. 4, 67 (1993)], andJ. Paillet and H. Mercier (Deep-Sea Res., Part I,inpress).8. Topographic sill depths of ;1000 m are a barrier forflow of this density into the Nordic Seas.9. L. V. Worthington, Johns Hopkins Oceanogr. Stud.6, (1976); R. A. Clarke, H. W. Hill, R. F. Reininger,B. A. Warren, J. Phys. Oceanogr. 10, 25 (1980);M. S. McCartney, Prog. Oceanogr. 29, 283 (1992);W. J. Schmitz and M. S. McCartney, Rev. Geophys.31, 29 (1993).10. R. S. Pickart and W. M. Smethie, J. Phys. Oceanogr.23, 2602 (1993).11. J. Paillet, M. Arhan, M. S. McCartney, in preparation.12. P. B. Rhines and W. R. Holland, Dyn. Atmos. Oceans3, 289 (1979).13. W. R. Holland, Geophys. Fluid Dyn. 4, 187 (1973).14. R. Gerdes and C. Ko¨ berle, J. Phys. Oceanogr. 25,2624 (1995).15. These gyres at intermediate depths are distinct fromthe smaller scale gyres at abyssal depths that aremore clearly associated with the basin topography.16. P. B. Rhines and W. R. Young, J. Fluid Mech. 122,347 (1982).17. W. R. Holland and P. B. Rhines, J. Phys. Oceanogr.10, 1010 (1980).18. S. McDowell, P. Rhines, T. Keffer, ibid. 12, 1417(1982); T. Keffer, ibid. 15, 509 (1985); J. L. Sar-miento, C. G. H. Rooth, W. Roether, J. Geophys.Res. 87, 8047 (1982).19. J. Pedlosky, Ocean Circulation Theory (Springer-Verlag, New York, 1996).20. Potential vorticity is calculated as f1/so3]sn/]z,where f is the planetary vorticity, z is the verticalcoordinate, and sois a constant potential density.The vertical derivative is computed locally over anominal depth of 100 m. In an effort to approximateneutral surfaces [T. J. McDougall, J. Phys. Ocean-ogr. 17, 1950 (1987)], potential vorticity was alsocalculated as f/h, where h is the distance betweentwo locally referenced isopycnals. The differences inthe two methods were insignificant to the results ofthis study; that is, the region and extent of homoge-nization were the same with either calculation.21. M. S. McCartney and L. D. Talley, ibid. 12, 1169(1982).22. J. O’Dwyer and R. G. Williams (J. Phys. Oceanogr.,in press) report, from an analysis of the Levitus dataset, possible regions of homogenization at abyssaldepths in the western North Atlantic.23. An absolute flow field was calculated for s25 36.95by differentiation of a modified [H.-M. Zhang andN. G. Hogg, J. Mar. Res. 50, 385 (1992)] Montgom-ery stream-function field [R. B. Montgomery, Bull.Am. Meteorol. Soc. 18, 210 (1937)] with s35 41.45(at ;3000 m) as the level of no motion.24. D. L. Musgrave, J. Geophys. Res. 90, 7037 (1985).25. Surfaces shallower than s25 36.95 show counter-rotating gyres separated by this instability region inthe western North Atlantic, which suggests the im-portance of eddy flux divergence in the forcing of thegyres.26. J. Pedlosky, J. Phys. Oceanogr. 13, 2121 (1983).27. N. Hogg, Deep-Sea Res. Part A 30, 945 (1983).28. P. B. Rhines and W. R. Young, J. Mar. Res. 40, 559(1982).29. P. Cessi, G. Ierley, W. Young, J. Phys. Oceanogr. 17,1640 (1987); P. Cessi, ibid. 18, 662 (1988).30.W. J. Jenkins and P. B. Rhines, Nature 286, 877 (1980);R. S. Pickart, Deep-Sea Res. Part A 39, 1553 (1992);W. M. Smethie, Prog. Oceanogr. 31, 51 (1993).31. I thank J. Pedlosky for his aid in the interpretation ofthese fields and P. Rhines for his comments on themanuscript. Support from NSF (grant OCE- 9629489)is gratefully acknowledged.20 February 1997; accepted 22 May 1997Maximum and Minimum TemperatureTrends for the GlobeDavid R. Easterling,* Briony Horton, Philip D. Jones,Thomas C. Peterson, Thomas R. Karl, David E. Parker,M. James Salinger, Vyacheslav Razuvayev, Neil Plummer,Paul Jamason, Christopher K. FollandAnalysis of the global mean surface air temperature has shown that its increase is due,at least in part, to differential changes in daily maximum and minimum temperatures,resulting in a narrowing of the diurnal temperature range (DTR). The analysis, usingstation metadata and improved areal coverage for much of the Southern Hemispherelandmass, indicates that the DTR is continuing to decrease in most parts of the world,that urban effects on globally and hemispherically averaged time series are negligible,and that circulation variations in parts of the Northern Hemisphere appear to be relatedto the DTR. Atmospheric aerosol loading in the Southern Hemisphere is much less thanthat in the Northern Hemisphere, suggesting that there are likely a number of factors,such as increases in cloudiness, contributing to the decreases in DTR.The global mean surface air temperaturehas risen about 0.5°C during the 20th cen-tury (1). Analysis has shown that this risehas resulted, in part, from the daily mini-mum temperature increasing at a faster rateor decreasing at a slower rate than the dailymaximum, resulting in a decrease in theDTR for many parts of the world (2, 3).Decreases in the DTR were first identifiedin the United States, where large-areatrends show that maximum temperatureshave remained constant or have increasedonly slightly, whereas minimum tempera-tures have increased at a faster rate (4).Similar changes have been found for otherparts of the world as data have becomeavailable, allowing more global analyses (2,3). However, in some areas the pattern hasbeen different: In parts of New Zealand (5)and alpine regions of central Europe (6),maximum and minimum temperature haveincreased at similar rates, and in India, theDTR has increased as a result of a decreasein the minimum temperature (7). To eval-uate these varying results, we conducted anexpanded analysis on global and regionalscales.Local effects such as urban growth, ir-rigation, desertification, and variations inlocal land use can all affect the DTR (3);in particular, urbanized areas often show anarrower DTR than nearby rural areas (8).Large-scale climatic effects on the DTRinclude increases in cloud cover, surfaceevaporative cooling from precipitation,greenhouse gases, and tropospheric aero-sols (9, 10). Recent studies have demon-strated a strong relation between trends ofthe DTR and decreases in pan evaporationover the former Soviet Union and theUnited States (11),