date: Sat, 19 Jun 1999 22:34:40 +0100 (BST) from: Simon Tett subject: [schlesin@uiatma.atmos.uiuc.edu: Commentary on your Nature paper] to: k.briffa@uea.ac.uk Keith -- I guess a little in confidence but what is the quality of the paleo data for the "65-70" nrth atl oscillation the Michael Schlesinger is going on about. Simon ------- Start of forwarded message ------- Date: Sat, 19 Jun 1999 15:07:07 -0500 From: Michael Schlesinger Subject: Commentary on your Nature paper To: sfbtett@meto.gov.uk Cc: schlesin@uiatma.atmos.uiuc.edu, Natasha@uiatma.atmos.uiuc.edu, jfbmitchell@meto.gov.uk Content-type: MULTIPART/ALTERNATIVE; BOUNDARY="Boundary_(ID_nWE4dxCiUjFpUfoOD6FgRg)" - --Boundary_(ID_nWE4dxCiUjFpUfoOD6FgRg) Content-type: text/plain; charset=iso-8859-1 Content-transfer-encoding: quoted-printable Dear Simon: Below is a Scientific Correspondence that I and Natalia Andronova have submitted to Nature commenting on your recent paper therein. Our commentary is not critical of your paper, rather, it proposes a possible alternate explanation for the warming of the first half of this century, a possibility that you did not exclude in your paper, as we note in the commentary. Cheers, Michael Schlesinger _________________________________________________________ Cause of the Warming During the First Half of the Twentieth Century - Implications for Climate Sensitivity and Detection Tett et al. [1] find that the increase in global-mean near-surface temperature during the first half of the twentieth century may be due to variations in the sun's irradiance. While this supports the earlier findings of Kelly and Wigley [2], Schlesinger and Ramankutty [3] and others, there are two reasons for being cautious about this finding. =46irst, there are no reliable measurements of solar irradiance prior to the beginning of satellite observations in 1978. Second, an oscillation with a period of about 65-70 years has been found in the near-surface temperature over the North Atlantic which appears to explain the observed global-mean warming during the first half of the twentieth century. Below we briefly expand on these two factors and highlight the importance of their resolution in terms of climate sensitivity and detection. Ground-based observations of solar irradiance, such as those of Abbott from 1908 to 1952, are unable to detect irradiance variations less than about 1% [4], while the much more precise solar-irradiance observations by satellites began only in 1978 [5]. Thus we know observationally that solar irradiance did not vary by more than about 1% over the four '11-year' sunspot cycles observed by Abbott, and that solar irradiance did vary by about 0.1% over the two sunspot cycles since 1978 [6]. We do not know observationally, however, whether solar irradiance before 1978 varied by more than it did after 1978. Tett et al. [1] used putative solar-irradiance variations constructed using 'surrogates' by Hoyt and Schatten [7] (the fraction of penumbral spots, solar-cycle length, equatorial rotation rate, decay rate of the solar cycle, and mean level of solar activity) and Lean et al. [8] (sunspot areas and locations, He 1083 nm emission, group sunspot numbers, and Ca emissions from the sun and sunlike stars). The resulting constructions of solar-irradiance variations are not in agreement with each other. Thus both cannot be completely correct, while both could be wrong. In contrast, there is a growing body of evidence that the warming during the first half of the twentieth century was the result, at least in part, of an oscillation in the surface temperature of the North Atlantic Ocean with a period of about 65-70 years, a possibility not excluded by Tett et al. [1]. This oscillation was found by Schlesinger and Ramankutty [9] in their analysis of the instrumental near-surface temperature record, first from its manifestation in the global-mean temperature, and second from the geographical distribution of temperature, the latter showing the oscillation to be located only over the North Atlantic and its surrounding continental margins. Subsequent analyses of much longer paleoclimate proxy data (tree rings, ice cores, ice melt, lake varves, coral, historical data) have shown the existence of this oscillation during the pre-instrumental period, going back as far as the paleodata allow, to the year 500 [10-12]. Moreover, simulations by the coupled atmosphere-ocean general circulation model of the Geophysical Fluid Dynamics Laboratory have shown such an oscillation in the surface temperature of the North Atlantic, initially with a period of 40-50 years [13], but as the length of the simulation was increased from 300 to 2000 years, with a longer period of about 40-80 years [14]. A version of this model with doubled horizontal resolution and 50% greater vertical resolution simulates the oscillation with a period of about 60-80 years [Delworth, personal communication]. It is important to determine which of these two explanations for the early twentieth century warming is correct, if either. Our maximum-likelihood estimation of the climate sensitivity - the change in the earth's equilibrium near-surface temperature in response to a radiative forcing equivalent to that for a doubling of the pre-industrial CO2 concentration, =90T2x - indicates that if the solar irradiance varied as suggested by Hoyt and Schatten [7] or Lean et al. [8], then the value of =90T2x required to best represent the instrumental near-surface temperature changes is about 40% smaller than if there was no variation in solar irradiance [15]. Clearly, whether or not this is so can have a significant impact on decisions about climate-change policy. Conversely, if the more correct answer is the 65-70 year oscillation in North Atlantic near-surface temperature, then any coupled atmosphere-ocean model that either fails to simulate this oscillation or simulates it too weakly would underestimate the internal variability of the climate system [16]. This in turn would overestimate the signal-to-noise ratio for the observed near-surface temperature changes during the second half of the twentieth century, and thereby contribute to a possibly spurious claim of the detection of climate change and its attribution to anthropogenic causes. It is therefore of paramount importance to learn whether the observed warming during the first half of the twentieth century was caused by variations in the sun's irradiance, the 65-70 year oscillation in North Atlantic surface temperature, their combined effect, or something as yet undiscerned. Michael E. Schlesinger and Natalia G. Andronova Department of Atmospheric Sciences University of Illinois @ Urbana-Champaign 105 S. Gregory Avenue Urbana, IL 61801 USA e-mail: schlesin@atmos.uiuc.edu 1. Tett, S.F.B., Stott, P.A., Allen, M.R., Ingram, W.J. & Mitchell, J.F.B. Nature (1999). 2. Kelly, P.M. & Wigley, T.M.L. Nature 360, 328-330 (1992). 3. Schlesinger, M.E. & Ramankutty, N. Nature 360, 330-333 (1992). 4. Fr=F6hlich, C. in The Solar Output and its Variation (ed. White, O.R.) 93-109 (Colorado Associated University Press, Boulder, 1977). 5. Willson, R.C. in The Sun as a Variable Star, International Astronomical Union Colloquium 143 Proceedings (eds. Pap, J., Fr=F6hlich, X., Hudson, H. & Solanki, K.) 54-62 (Cambridge Univ. Press, New York, 1994). 6. Fr=F6hlich, C. & Lean, J. Geophys. Res. Lettr. 25, 4377-4380 (1998). 7. Hoyt, D.V. & Schatten, K.H. J. Geophys. Res. 98, 18,895-18,906 (1993). 8. Lean, J., Beer, J. & Bradley, R. Geophys. Res. Lettr. 22, 3195-3198 (1995). 9. Schlesinger, M.E. & Ramankutty, N. Nature 367, 723-726 (1994). 10. Mann, M.E., Park, J. & Bradley, R.S. Nature 378, 266-270 (1995). 11. Mahasenan, N., Watts, R. & Dowlatabadi, H. Geophys. Res. Lettr. 24, 563-566 (1997). 12. Shabalova, M.V. & Weber, S.L. J. Geophys. Res. , In press (1999). 13. Delworth, T., Manabe, S. & Stouffer, R.J. J. Climate 6, 1993-2011 (1993). 14. Delworth, T.L., Manabe, S. & Stouffer, R.J. Geophys. Res. Lettr. 24, 257-260 (1997). 15. Andronova, N.G. & Schlesinger, M.E. J. Geophys. Res. , (in preparation) (1999). 16. Barnett, T.P. J. Climate 12, 511-518 (1999). - --Boundary_(ID_nWE4dxCiUjFpUfoOD6FgRg) Content-type: text/enriched; charset=iso-8859-1 Content-transfer-encoding: quoted-printable Dear Simon: Below is a Scientific Correspondence that I and Natalia Andronova have submitted to Nature commenting on your recent paper therein. Our commentary is not critical of your paper, rather, it proposes a possible alternate explanation for the warming of the first half of this century, a possibility that you did not exclude in your paper, as we note in the commentary. Cheers, Michael Schlesinger _________________________________________________________ TimesCause of the Warming During the First Half of the Twentieth Century - Implications for Climate Sensitivity and Detection TimesTett et al. [1] find that the increase in global-mean near-surface temperature during the first half of the twentieth century may be due to variations in the sun's irradiance. While this supports the earlier findings of Kelly and Wigley [2], Schlesinger and Ramankutty [3] and others, there are two reasons for being cautious about this finding. First, there are no reliable measurements of solar irradiance prior to the beginning of satellite observations in 1978. Second, an oscillation with a period of about 65-70 years has been found in the near-surface temperature over the North Atlantic which appears to explain the observed global-mean warming during the first half of the twentieth century. Below we briefly expand on these two factors and highlight the importance of their resolution in terms of climate sensitivity and detection. Ground-based observations of solar irradiance, such as those of Abbott from 1908 to 1952, are unable to detect irradiance variations less than about 1% [4], while the much more precise solar-irradiance observations by satellites began only in 1978 [5]. Thus we know observationally that solar irradiance did not vary by more than about 1% over the four '11-year' sunspot cycles observed by Abbott, and that solar irradiance did vary by about 0.1% over the two sunspot cycles since 1978 [6]. We do not know observationally, however, whether solar irradiance before 1978 varied by more than it did after 1978. Tett et al. [1] used putative solar-irradiance variations constructed using 'surrogates' by Hoyt and Schatten [7] (the fraction of penumbral spots, solar-cycle length, equatorial rotation rate, decay rate of the solar cycle, and mean level of solar activity) and Lean et al. [8] (sunspot areas and locations, He 1083 nm emission, group sunspot numbers, and Ca emissions from the sun and sunlike stars). The resulting constructions of solar-irradiance variations are not in agreement with each other. Thus both cannot be completely correct, while both could be wrong. In contrast, there is a growing body of evidence that the warming during the first half of the twentieth century was the result, at least in part, of an oscillation in the surface temperature of the North Atlantic Ocean with a period of about 65-70 years, a possibility not excluded by Tett et al. [1]. This oscillation was found by Schlesinger and Ramankutty [9] in their analysis of the instrumental near-surface temperature record, first from its manifestation in the global-mean temperature, and second from the geographical distribution of temperature, the latter showing the oscillation to be located only over the North Atlantic and its surrounding continental margins. Subsequent analyses of much longer paleoclimate proxy data (tree rings, ice cores, ice melt, lake varves, coral, historical data) have shown the existence of this oscillation during the pre-instrumental period, going back as far as the paleodata allow, to the year 500 [10-12]. Moreover, simulations by the coupled atmosphere-ocean general circulation model of the Geophysical Fluid Dynamics Laboratory have shown such an oscillation in the surface temperature of the North Atlantic, initially with a period of 40-50 years [13], but as the length of the simulation was increased from 300 to 2000 years, with a longer period of about 40-80 years [14]. A version of this model with doubled horizontal resolution and 50% greater vertical resolution simulates the oscillation with a period of about 60-80 years [Delworth, personal communication]. It is important to determine which of these two explanations for the early twentieth century warming is correct, if either. Our maximum-likelihood estimation of the climate sensitivity - the change in the earth's equilibrium near-surface temperature in response to a radiative forcing equivalent to that for a doubling of the pre-industrial CO2 concentration, =90T2x - indicates that if the solar irradiance varied as suggested by Hoyt and Schatten [7] or Lean et al. [8], then the value of =90T2x required to best represent the instrumental near-surface temperature changes is about 40% smaller than if there was no variation in solar irradiance [15]. Clearly, whether or not this is so can have a significant impact on decisions about climate-change policy. Conversely, if the more correct answer is the 65-70 year oscillation in North Atlantic near-surface temperature, then any coupled atmosphere-ocean model that either fails to simulate this oscillation or simulates it too weakly would underestimate the internal variability of the climate system [16]. This in turn would overestimate the signal-to-noise ratio for the observed near-surface temperature changes during the second half of the twentieth century, and thereby contribute to a possibly spurious claim of the detection of climate change and its attribution to anthropogenic causes. It is therefore of paramount importance to learn whether the observed warming during the first half of the twentieth century was caused by variations in the sun's irradiance, the 65-70 year oscillation in North Atlantic surface temperature, their combined effect, or something as yet undiscerned. Michael E. Schlesinger and Natalia G. Andronova Department of Atmospheric Sciences University of Illinois @ Urbana-Champaign 105 S. Gregory Avenue Urbana, IL 61801 USA e-mail: Times0000,0000,00FFschlesin@atmos.uiuc.edu Times1.= Tett, S.F.B., Stott, P.A., Allen, M.R., Ingram, W.J. & Mitchell, J.F.B. Nature (1999). 2. Kelly, P.M. & Wigley, T.M.L. Nature 360, 328-330 (1992). 3. Schlesinger, M.E. & Ramankutty, N. Nature 360, 330-333 (1992). 4. Fr=F6hlich, C. in The Solar Output and its Variation (ed. White, O.R.) 93-109 (Colorado Associated University Press, Boulder, 1977). 5. Willson, R.C. in The Sun as a Variable Star, International Astronomical Union Colloquium 143 Proceedings (eds. Pap, J., =46r=F6hlich, X., Hudson, H. & Solanki, K.) 54-62 (Cambridge Univ. Press, New York, 1994). 6. Fr=F6hlich, C. & Lean, J. Geophys. Res. Lettr. 25, 4377-4380 (1998). 7. Hoyt, D.V. & Schatten, K.H. J. Geophys. Res. 98, 18,895-18,906 (1993). 8. Lean, J., Beer, J. & Bradley, R. Geophys. Res. Lettr. 22, 3195-3198 (1995). 9. Schlesinger, M.E. & Ramankutty, N. Nature 367, 723-726 (1994). 10. Mann, M.E., Park, J. & Bradley, R.S. Nature 378, 266-270 (1995). 11. Mahasenan, N., Watts, R. & Dowlatabadi, H. Geophys. Res. Lettr. 24, 563-566 (1997). 12. Shabalova, M.V. & Weber, S.L. J. Geophys. Res. , In press (1999). 13. Delworth, T., Manabe, S. & Stouffer, R.J. J. Climate 6, 1993-2011 (1993). 14. Delworth, T.L., Manabe, S. & Stouffer, R.J. Geophys. Res. Lettr. 24, 257-260 (1997). 15. Andronova, N.G. & Schlesinger, M.E. J. Geophys. Res. , (in preparation) (1999). 16. Barnett, T.P. J. Climate 12, 511-518 (1999). - --Boundary_(ID_nWE4dxCiUjFpUfoOD6FgRg)-- ------- End of forwarded message ------- -- ============================================================ + Spinning in the wind at the UKMO + ============================================================ Tel : +[44]-1344-856886 Fax: +[44]-1344-854898