You are on page 1of 6

Global Trends in Wind Speed and Wave Height I. R. Young, et al. Science 332, 451 (2011); DOI: 10.

1126/science.1197219

This copy is for your personal, non-commercial use only.

If you wish to distribute this article to others, you can order high-quality copies for your colleagues, clients, or customers by clicking here. Permission to republish or repurpose articles or portions of articles can be obtained by following the guidelines here. The following resources related to this article are available online at www.sciencemag.org (this infomation is current as of April 21, 2011 ): Updated information and services, including high-resolution figures, can be found in the online version of this article at: http://www.sciencemag.org/content/332/6028/451.full.html Supporting Online Material can be found at: http://www.sciencemag.org/content/suppl/2011/03/22/science.1197219.DC1.html This article appears in the following subject collections: Oceanography http://www.sciencemag.org/cgi/collection/oceans

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright 2011 by the American Association for the Advancement of Science; all rights reserved. The title Science is a registered trademark of AAAS.

Downloaded from www.sciencemag.org on April 21, 2011

D. to a lesser degree. 48. A. V. 24)] within the time series of monthly mean. Okuyama. Although there is a range of results. T. Ochiai. Int. Recently.5% per year (a 5 to 10% net increase over the past 20 years). B. 7181 (2002).2 m (17). The trend is stronger in the Southern Hemisphere than in the Northern Hemisphere. Miyamoto.org 22 APRIL 2011 451 Downloaded from www. Eschenmoser. including altimeters. Boca Raton. respectively. Soc. however. Angew. P. 130. K. Newburg. (20) carried out systematic calibrations and cross-platform validations of all altimeter measurements over the full 23 years for which data are available. R. J. accepted 4 March 2011 10. C. 111. Am. L. W. Z. for a generous gift of BrF3. P. S. The rate of increase is greater for extreme events as compared to the mean condition. J. CRC Handbook of Chemistry and Physics (CRC Press. Petrov. F. over this period. M. Breslow. D. M. Organometallics 27. Angew. Soc. Yoshikai. M. J. J. point measurements (9). and 99th-percentile values of wind speed and wave height for 2° × 2° regions covering the globe. Takino. V. Ed. C. Y. these data were excluded from the analysis. Lin. R. whereas wind speed is analyzed for the shorter period 1991–2008. J. Melbourne. 23. 19. Australia. Ed. DesMarteau. 31. V. Diaz-Requejo. 120. 22. using a variety of instruments. 2. Fusco. E-mail: ir. A. where there are smaller localized increases in wind speed of approximately 0. Glukhovtsev. At the mean and 90th percentile. M. where E is the total energy of the wave field. Canepa. A. Chem. 18. J. We find a general global trend of increasing values of wind speed and. Chem. X.edu. S.-S.-Y. 25. 3360 (1995).sciencemag. Of these instruments. 22) on a global scale. Acknowledgments: This work was supported by a Grant-in-Aid for Scientific Research (B) (Japan Society for the Promotion of Science). 20. Zhao. wave height. Solvents and Solvent Effects in Organic Chemistry (Wiley-VCH. 21.org on April 21. 15.05 and 1.69 eV) < PhBr (8. 129. Sota. Vinyl Cations (Academic Press. 90th-. Gómez-Emeterio. M. K. Soc. M. There is a clear global increase in wind speed for all three statistics. Soc. C. 91. Montanari. S. Reichardt. 1 to 3 (percentage increase or decrease) and figs. Hs ¼ 4 E. W. U10 . with a root mean square (rms) error of less than 0. Chem. The trend was quantified as a linear function over the duration of the time series. 13. Therefore. The monthly mean. Fiorentino. Numerous calibrations of these altimeters have shown that the instruments can be used ffito meapffiffiffi sure significant wave height. with a rms error of less than 1. J. 29. R. Ochiai. Because the data set spans multiple satellite platforms.-M.young@anu. Ochiai et al. J. Such a result indicates that the intensity of extreme events is increasing at a faster rate than that of the mean conditions. Ochiai. Almost all of these studies are regional rather than global. Stang. L. Lide. many studies show an increasing trend in significant wave height. 117. Z. Mello. Ke. D. Thus. Am. S. K. consistent calibration and validation are critical when investigating long-term trends. Chem. The only significant exception to this positive trend is the central north Pacific. Chem. We thank Central Glass Co. Data from altimeter missions have been used to investigate mean ocean wind and wave climatology (21. As a result. Chem. D. P. 32.06 eV) (14). or satellite observations (16). Phillips. M. E. J. Germany. Synlett 2009. Satellite-based systems provide an alternative to visual or in situ measurements of oceanic wind O Swinburne University of Technology. J. Miyamoto. Ed. Yu. Badiei et al. there exists an almost continuous (there was a break in 1990–1991) record of measures from a total of seven different altimeter missions. N. Commun. 2273 (1969). they substantially influence the mechanisms of air-sea interaction (3). and 99th-percentile trend values for both wind speed and wave height are shown in Figs.98 eV) < PhCl (9. Delaware. Che. Am. B. 12938 (2007). Am. 9705 (2009). FL. Hanack. Donadelli. Sloan. Materials and methods are available as supporting material on Science Online. C. T. 9120 (1999). Subramanian. W. M. Am. 1979). Wilmington. and wind speed. ceanic wind speed and wave height help to control the flux of energy from the atmosphere to the ocean (1) and upper ocean mixing (2). Baran. R. R. wind speeds over the majority of the world’s oceans have increased by at least 0. 33. 124. is much more than temperature. Am. Chem.. New York. C. 1992). This study provided a consistent data set over this extended period. 4897 (1993). providing global coverage of wind and/or waves. 6749 (1989). Young. Renfrow. Am. R. 27. Teflon PFA is a perfluoroalkoxy copolymer resin. M. 90th-. Perez. Resnati. 26. Rappoport. T. changes in wind speed and the surface gravity waves generated by such winds play an important role. the radar altimeter provides by far the longestduration record.au speed and wave height. Urbano. S7 to S9 (SOM) (absolute increase or decrease). 14.1201686 Global Trends in Wind Speed and Wave Height I.-M.-M. Previous attempts to investigate trends in oceanic wind speed and wave height have used ship observations (4–8). Fung. M. Small deuterium KIEs of 1. Japan. often correlated with interannual variations such as the North Atlantic Oscillation. J. A. available from DuPont. Sueda. with the magnitude of the increase being larger for the 99th percentile. Chem.1126/science. Chem. Bach. Yamanaka. we used this data set to investigate whether there have been systematic changes in the ocean wind and wave climate over this SCIENCE VOL 332 period. N. 46. 2003). particularly in the North Atlantic and North Pacific. Chem. Soc. we take particular care to ensure that the trend can be separated from the seasonal component (SOM). 1101 (2007). P. Soc. Zieger et al. Climate.REPORTS 12. Chem. Zieger. 121. scatterometers. Ochiai. Soc. Chen. G. Int. 9961 (2008). Hayashi. Am. We aim here to determine whether there is a statistically significant trend [where trend is defined as a linear increase or decrease in the mean (23.25% www.-H. 2118 (2008). M. The present analysis uses recently developed satellite altimeter data sets to carefully investigate such trends on a global scale.* S.. 30. Au. The mean and 90thpercentile wind speed trends are relatively similar. We used a 23-year database of calibrated and validated satellite altimeter measurements to investigate global changes in oceanic wind speed and wave height over this period. Since the launch of GEOSAT in 1985.org/cgi/content/full/332/6028/448/DC1 Materials and Methods SOM Text Figs. Ionization potentials of halobenzenes increase in the order PhI (8. R. T. the wave height analysis considers the period 1985–2008. 2011 . Nakamura. Babanin Studies of climate change typically consider measurements or predictions of temperature over extended periods of time. Curci. The trend was expressed for each 2° × 2° region as the annual percentage increase or decrease relative to the mean condition and in absolute terms. N. M. In the analysis.5 m/s (17–21). 159 (2009). J. 47.and 99th-percentile wind speed data from the GEOSAT altimeter were of questionable quality (SOM). 511 (2010). The analysis revealed that 90th. Chem. *To whom correspondence should be addressed. S1 to S12 Tables S1 and S2 References 14 December 2010. 17. 4126 (2008). Weinheim. Am.sciencemag. 115..25 to 0. Soc. C. 10528 (1998).07 for the hydride abstraction reaction of the bridgehead adamantane C–H bond with tert-butyl cation and hydroperoxonium ion were reported (3). J. and synthetic aperture radar. M. Careful ship observations (4–6) also show waves locally generated by the wind (hereafter referred to as wind-sea) and swell behaving quite differently and that there exist quite different trends in wind speed and wave height. Okada. D. Nakanishi. Victoria. Huang. Asian J. Because the seasonal cycle typically is large. 16. 28. care must be exercised in determining trend information from the data set [supporting online material (SOM)].. 24. Che. Chem. A. Tada.sciencemag. W. Over the oceans. numerical modeling (10–15). N. Supporting Online Material www. D. In the present study. Soc.

48/36% (mean).50% at the 99th percentile. At every buoy location. the exceptions being the areas of weak positive trend noted above. 15/8% (mean). In order to validate the trends observed in the altimeter data. (20). wave height. again consistent with the altimeter results. the wind speed trend becomes more positive. the same analysis was applied to the 12 deep-water buoys used by Zieger et al. whereas the weaker trends in the equatorial regions are not.REPORTS per year and some areas where there is a weak negative trend. the same general features observed in the altimeter trends are repeated in the buoy data.org Downloaded from www. as does much of the equatorial regions of all oceanic basins. Fig. The magnitude of the trend also increases at extreme values. which showed a weak increasing trend for the mean and 90th percentile. It should be noted that the areas of weak trends observed for the mean and 90th percentile are largely not statistically significant. Large regions of the north Pacific and north Atlantic show a weak negative trend (0. Also shown (Figs.25% per year at the 90th percentile and 0. 60/47% (99th percentile).50% per year. this trend now being statistically significant.org on April 21. together with the corresponding altimeter values for 2° × 2° regions centered on each of the buoy locations. 1 to 3) are regions of the globe where the calculated trend is statistically significant at the 95% level (SOM). 53/40% (90th percentile). the following percentages of points produced trends that were statistically significant at the 90/95% confidence levels. The mean wave height trend (Fig. indicating that extreme wind speeds are increasing over the majority of the world’s oceans by at least 0.sciencemag. there is no clear trend for mean monthly wave height across the buoys. is the trend statistically significant (SOM). The equatorial and tropical regions of all oceanic basins show a neutral trend for wave height in both cases. 1) shows a relatively neutral condition. particularly in the north Pacific. As for the altimeter observations. 3) wave height trends are progressively more positive. now shows a stronger trend of approximately 0. Averaged over the full globe. and 99th-percentile time series. respectively: wind speed. At the 99th percentile. which is consistent with the altimeter results. It is clear that there is variability between buoys in the same geographic region and between buoy and corresponding altimeter trend values. The results are shown in Table 1. The calculated wind speed trend for the majority of the globe is statistically significant. Color contour plots of mean trend (percent per year). Wind speed is shown at the top and wave height at the bottom. the Southern Hemisphere has a consistent weak positive trend of approximately 0.e.25% per year). The region in the central north Pacific. 1. this trend being weaker than for wind speed. In almost none of these regions. 20/12% (90th percentile). the buoys produce a more positive trend for wave height at extreme values. Points that are statistically significant according to the Seasonal Kendall test are shown with dots. however. the stronger positive trends at high latitudes are statistically significant. However. with the higher latitudes (greater than T35°) of both the hemispheres showing positive trends of approximately 0. For the 99th-percentile wave height.sciencemag.25% per year. 2) and the 99th-percentile (Fig. moving to more extreme conditions). the wind speed trend is positive for each of the monthly mean. 2011 . The 90thpercentile (Fig. The wave height trend becomes more positive moving from the mean to the 99th percentile (i. 90th-. 68/55% (99th percentile).. However.75% per year. However. 452 22 APRIL 2011 VOL 332 SCIENCE www.

(28) have shown that the wave climate in the Southern Hemisphere is influenced by the Southern Annular Mode. Indeed.org SCIENCE VOL 332 22 APRIL 2011 453 Downloaded from www. 27) indicate that for wind-generated waves. and the results show little correlation between wind speed and wave height trends in these areas. Figure S6 (SOM) shows the trend in mean monthly wind speed calculated from NCEP (National Centre for Environmental Protection) global reanalysis results (25). for instance. The present results do show causal relationships between the trends in wind speed and wave height. it is Fig.g. differing trends in wind-sea (a proxy for wind speed) and swell are plausible. it has been shown that swell is influenced by not only the intensity of generating meteorological systems but also their frequency (4. It does not necessarily follow that the observed trends are a result of. In contrast. Hence. The present analysis is aimed at determining whether there is a linear trend over the period of the observations (approximately 23 years). the wave climate in tropical regions is dominated by remotely generated swell. However. it is highly likely that such long-term oscillations will significantly influence the global ocean wind and wave climate. and hence one would expect a stronger correlation between wind speed and wave height. Observations of local wave generation (26. Similar interannual variations have also been shown to be correlated with wave heights in the North Atlantic (5. A detailed analysis of the observed trends would require information on the distribution of wind-sea and swell together with swell propagation direction and a basin-specific analysis. significant wave height is approximately proportional to wind speed.4 across large areas of the Southern Ocean. The Southern Ocean is dominated by strong westerly winds blowing across large oceanic fetches. Such data are not directly available from the altimeter. 1. As expected. 2011 . interannual-to-decadal variations of the high-latitude wind belts have been observed. 11. Color contour plots of the 90th-percentile trend (percent per year). 10. 99th percentile).sciencemag. A regression analysis between the monthly mean altimeter significant wave height and the Southern Annual Mode Index showed a weak correlation. and Hemer et al. S9) are consistent with this hypothesis. For extreme conditions (e. 3 and fig.. in this region. Hence. global warming.org on April 21. The present analysis cannot absolutely determine why there is a stronger trend in wind speed than in wave height or the underlying physical processes responsible for the observed positive trends in both quantities at extreme values. the situation is more complex in that wave height in the open ocean is a mix of locally generated wind-sea and remotely generated swell. one might initially expect that the spatial patterns for wind speed and wave height would be similar.REPORTS A further validation check was performed by comparing the altimeter global trend values with numerical model results. The present results (Fig. waves tend to be generated by local storm events. there are well-correlated positive trends in both wind speed and wave height. with a correlation coefficient up to 0. Hence. This result is qualitatively consistent with Fig. The same is true in the high latitudes of the Northern Hemisphere. 2. Wind speed is shown at the top and wave height at the bottom. with wind speed and wave height showing similar positive trends for 99th-percentile conditions. www. 14). Although trends in wind-sea will be correlated with trends in the local wind speed.sciencemag. 13. Points that are statistically significant according to the Seasonal Kendall test are shown with dots. 6). Because the present data set is only two decades long.

A. 5667 (2006). Cox. D. 2313 (2001). Res.26 5. or simply the upward portion of a multidecadal oscillation. P. which could be extrapolated into the future. X.70 2. Jannink.02 6.45 0. 109 (2006).7 70. 29.20 2. Katsaros.27 1.40 Altimeter trend (cm/s/year) 99th 10. there is a clear statistically significant trend of increasing wave height at high latitudes and more neutral conditions in equatorial regions. 10. K. S.02 0.0 137. 6.42 13.00 7.6 46. Climate 28.6 46. 3. R.00 1.4 66.63 1. Comparison of trend estimates for buoy and altimeter data.59 –0. Clim.14 4. Geophys.48 2.89 4. S.40 0. 27. V.sciencemag. Lett.00 1. with no clear statistically significant trend for mean monthly values.4 53. Berry. J.21 0.94 13.59 2.61 4. J. I.16 12.02 4. G. 31.36 0. B. K. D.6 75.1 32. Clim.92 2. Sterl.71 –0. Dyn.7 93. Babanin.51 –0.3 42.84 –0.29 1.8 162. J.58 0.24 0. Grigorieva.8 38. however. E.96 6.45 3.42 1. S. Drennan.47 0.24 –0.02 0.9 25.24 0. Points that are statistically significant according to the Seasonal Kendall test are shown with dots.08 1.org on April 21. 4. Gulev.8 38.90 3. A. C. Swail.43 2.65 –0.34 1.41 0.08 1. Geophys. Hasse.63 0.59 4.00 0. L20605 (2006). L. Gulev. Bold values are statistically significant at the 95% level (bold and underscored) and at the 90% level (bold) where two significance tests were passed (the normal distribution and the homogeneity test) (SOM).64 2. Komen.sciencemag. M.55 0.71 0. A.3 42. 77.8 162.01 4.99 2.25 –0. K.07 4.93 1.52 4.0 130.50 12.9 57. 8. A. Phys.5 17.31 –0.03 2. Int.47 2. Latitude (oN) 25. Int. 106.54 0. R.95 –0.88 3.73 10. V.4 53.00 0.43 3. 26.43 –0.1 23. 2275 (1996). V. Res.12 1. Chem.61 –2.98 1. 2087 (1997).5 131.44 0. Res. Lett.5 17. Grigoriev. with the locations grouped by geographic region.76 10. 747 (2008). Geophys.01 0.org Downloaded from www. 189 (2009). Komen.82 3. Gulev.24 –0.6 75. Thomas. E. the results are more complex.00 –0.3 157. L24302 (2004).46 2. Cotton. Res. K.14 0.66 7.9 25.03 10. Geophys.50 –0. R.1 23. At more extreme conditions. Cotton. 3. A.50 6.00 454 22 APRIL 2011 VOL 332 SCIENCE www.46 2.50 2. R.33 7. 2. Res. Bull.86 3. Geophys.1 40. Gulf of Mexico North Atlantic North Pacific Hawaii 42001 42002 44004 44011 41002 46001 46002 46005 46006 46035 51001 51002 25. The top panel shows wind speed and the bottom panel shows wave height. J.13 0. Gulev. V. Meteorol.54 0. 19.64 –0.42 1. Region Buoy no. Kent. Earth 23.41 –0.41 4.11 0. Phillips. Sterl.51 2.06 –0. Swail. A.00 2. G. J. Ganopolski. V. 12. 19.04 0. 5.06 1.57 4.99 2. 587 (1998).4 148.57 2.5 177. M.34 0. Wind speed is shown at the top and wave height at the bottom.56 5.88 –0. Soc. Only a longer data set will be able to separate these possibilities. B. indicate that over the past two decades there has been a consistent trend toward increasing wind speeds. Table 1. For wave height.53 0.3 157.4 66.5 41.99 –0. 7. Swail. J.REPORTS Fig.11 –0. Color contour plots of the 99th-percentile trend (percent per year).41 1.16 0.42 1. L. Donelan.0 137.7 93. The present analysis does. 1091 (1999).33 9.54 10.5 41.1 32. Am.27 not possible to distinguish between a steadily increasing or accelerating trend.1 40. Wang.36 0.8 0. 103.00 0.8 Buoy trend (cm/s/year) Mean 90th 1.00 0.84 1.42 3.79 3.83 3. W.90 4. S.1 89.4 148.62 10. Bouws. C.0 130.1 Longitude (oW) 89. W. A.46 3. A. Climatol. References and Notes 1.69 10.06 0. 2011 .95 –0.06 0.74 5.5 131. 33. J.24 5.21 7. V.05 0. Ocean Model. V.9 57. 9.50 0.5 177.77 0.7 70. K. 5477 (1998).42 0. D.15 5.51 0. 11.66 Gulf of Mexico North Atlantic North Pacific Hawaii 42001 42002 44004 44011 41002 46001 46002 46005 46006 46035 51001 51002 99th Mean 90th 4.54 –0. T.50 2.25 8. J. Babanin.67 –0. R.

A Review of Modern Approaches to the Problem of Trend Extraction (Technical Report RRS2008/03. 19. Soc. as changes in evaporation and precipitation will influence the oxygen isotopic composition of seawater and. N. 251 (2003). Washington. T. TEX86. B. 2549 (2009). 15). Technol. D. I. Young. Today. WY 82071. A stronger and more immediate connection between marine surface waters and the atmosphere heightens the sensitivity of this oceanic layer to changes in regional hydrology (i. M. Kalnay et al. Kutztown.0 per mil increase due to increased continental ice volume. 19. Greenhouse gas concentrations were up to five times as high as present-day levels (1. accepted 4 March 2011 Published online 24 March 2011. Bianconicini. 23) and long life spans of these species mean that their bioapatite d18O values dampen or negate .g. J.1 (2002). A. S. <100 m depth). 28. Census Bureau. Cont. Holland. S1 to S9 Table S1 References 1 September 2010.S. Kushnir. S.. meridional gradients in d18Osw are a product of latitudinal differences in evaporation SCIENCE VOL 332 www. Alternatively. 2Program in Ecology. Wang. 14. variation in skeletal bioapatite d18O values simply reflects changes in the oxygen isotopic composition of the waters inhabited by these consumers (19–21). 2011 Latitudinal Gradients in Greenhouse Seawater d18O: Evidence from Eocene Sirenian Tooth Enamel and precipitation and the isotopic fractionation of water vapor by Rayleigh distillation during transport in the atmosphere (16 ). I. Atmospheric circulation pat- T 1 Department of Geology and Geophysics. Int. Appl. 27. R. and evidence that these wetter.0‰ greater than bottom waters).1126/science. Esterby. 25. I. 24. This research was supported by an Australian Research Council (ARC) Linkage Grant (LP0882422). In addition to characterizing the tropical greenhouse climate. 22. Kvamstø.g. J.and high latitudes) (9. 6). Meteorol. upwelling. precipitation and evaporation) (14. Atmos. Oxford. 98. This event defines the Eocene-Oligocene boundary and heralded a major shift in Earth’s climate state from greenhouse to icehouse conditions. USA.275 (1993).org/cgi/content/full/science. Queffeulou. J. R. P. R. J. Z. We gratefully acknowledge the support of the ARC and our industry partner RPS MetOcean. K. 495 (2004). 437 (1996). 11–13). dentin. F. Although postburial alteration is a potential complication for stable isotope analysis of fossil vertebrate material. enhanced evaporation. in isotopic equilibrium with seawater. UK. E. Clim. 3Department of Physical Sciences. Hemer. (2)]. Church. 8 (suppl. The warmer temperatures of the Eocene may have promoted a more vigorous hydrologic cycle. Vikebø. Clim. A. latent heat to the poles. Hydrol. Young. Supporting Online Material www. Young. bone. Young. 27. Oxygen isotopic fluxes track moisture fluxes and. Results from an atmospheric general circulation model support this interpretation and suggest that Eocene low latitudes were extremely wet. 2008). annual moisture balance substantially wetter than that of today. G. Current understanding of atmospheric water vapor content in the Eocene has primarily focused on specific time intervals (e. J. UK. A). USA. V. I. which is much more resistant to the processes of diagenesis than are other bioapatites (e. J. sampling bioapatite from large marine mammals can also reduce or remove local environmental effects that are an inherent part of coastal environments (e. 107. T. G. Dagum. Bull. U k′37. Swail. restriction of analyses to tooth enamel. T. One means of acquiring more information on the Paleogene hydrologic cycle comes from the examination of latitudinal changes in the oxygen isotopic composition of ocean surface waters (d18Osw. J. R. L. similar information on precipitation levels at low latitudes is. where high global temperatures would have led to an increase in the concentration of water vapor in Earth’s atmosphere through intensified subtropical evaporation (7–9). Laramie. he Eocene epoch represents an extremely warm interval in Earth’s climate history. Int. Sewall3* The Eocene greenhouse climate state has been linked to a more vigorous hydrologic cycle at mid. S.and high latitudes. Climatol.e. K. R. WY 82071.2* and Jacob O. I. that of foraminiferal tests. G. From this temperature peak. USA. Deutsch. 1996). E. The offset between tropical d18O values from sampled Eocene sirenian tooth enamel and modern surface waters is greater than the expected 1. I. X. Mass. Challenor. M. 16. knowledge of the Eocene hydrologic cycle is critical to accurately calculating paleotemperatures from planktonic foraminifera. 17. Furnes. 20. Zieger. Furevik. 16. however. Res. J. Woolf.edu terns and a reduced latitudinal temperature gradient [e. G. 20. 475 (2010). 9. Wind Generated Ocean Waves (Elsevier.0‰ lower than those for bottom waters).. (10)] would have limited rainout along a longitudinal trajectory and facilitated the meridional transport of water and. the d18O values of ocean surface waters can provide insight into hydrologic cycle changes. and annual global temperatures during the Early Eocene Climatic Optimum (EECO) at ~50 million years ago (Ma) were as much as 12°C higher than modern values (3–5). R.. Clementz1. The rate of global temperature deterioration increased markedly at the end of the Eocene as the volume of continental ice on Antarctica rapidly increased.edu.sciencemag. W. Likewise. P. which precipitate their calcite shells. University of Wyoming. Hydrogh. V. J. Cotton. Oxford.1197219 Mark T. Cardone. J.sciencemag. 15.. DC. Greenwood. 10. reduced evaporation and higher precipitation levels at high latitudes result in lower salinities and d18Osw values (more than −2.REPORTS 13. Hunter. 26. M. conditions gradually declined with only a single prolonged interruption in this trend by a respite to warmer conditions during the Middle Eocene Climate Optimum (MECO) (4. In contrast. *To whom correspondence should be addressed.g. 10. Hasselmann. thence. 26. E-mail: mclemen1@uwyo. R.org 22 APRIL 2011 455 Downloaded from www. Young. thence. J. 23. 127 (1996). Mg/Ca. limited. thus producing a warmer and more humid early Paleogene climate state.sciencemag.. (C11). which may be costly and time consuming to obtain. Thus. 30. P. U. Ocean. S. The net result is that whereas ocean bottom waters (>1000 m depth) maintain salinities and d18O values that are more or less uniform globally. thus. Evidence of global climate change through the Cenozoic has primarily come from geochemical and isotopic analysis of marine microfossils. P. d18Osw values (Fig.org on April 21. one might expect that intensification of the hydrologic cycle during the Eocene would affect the magnitude of this meridional gradient in d18Osw values and should be recorded in marine isotope records. R. Geod. Eocene hyperthermal events) and regions (mid. PA 19530.. Ocean Res. 10. Vinoth. Atlas of the Oceans: Wind and Wave Climate (Pergmon.. 1A) and surface-water salinities (Fig. F. Cane. or scales). McElroy. D. 1999). Enhanced evaporation at lower latitudes leads to higher salinities and d18Osw values (as much as 2.g. 29. Geophys. A. 2107 (1997).1197219/DC1 SOM Text Figs. Am. can substantially improve the utility of this method. 17. The large home ranges (22. Laramie. This increased offset could result from suppression of surface-water d18O values by a tropical. J. Shelf Res. Kutztown University. Process. Climatol. Young. sewall@kutztown. University of Wyoming.g. stable isotope analysis of bioapatite from marine mammals is more appropriate for reconstructing d18Osw values (18). Alexandrov. 1B) can vary significantly with latitude (17 ). 235 (1994). J. 18. 2). As the d18O values of carbonates is also influenced by temperature. 23. 77. 21. 12 (1973). R. Y. Reistad. or tests. Zwiers. 2368 (2004). 931 (1999). Mar. As mammals can maintain a constant core body temperature (~37°C). reconstruction of Cenozoic d18Osw from planktonic microfossils would require an independent measure of temperature [e. G. more humid conditions extended to lower latitudes or throughout the entire Eocene is so far lacking. or continental runoff). Res. Geophys.