Springer Nature is making Coronavirus research free. View research | View latest news | Sign up for updates

Comments on ‘solar, volcanic and CO2 forcing of recent climatic changes’


The various limitations of a recent analysis of climatic variations in terms of solar volcanic and carbon dioxide forcing have been examined in more detail. In particular, the possibility of additional CO2 release from forest clearing greatly increases the very large statistical uncertainties in the original analysis. While the inclusion of the various forcing terms is a highly desirable approach, the uncertainties in the data used mean that it is not appropriate to suggest that ‘the surface warming due to the greenhouse effect has now been roughly determined’.

This is a preview of subscription content, log in to check access.

References and Notes

  1. [1]

    Gilliland, R. L.: 1982, Climatic Change 4, 111–131.

  2. [2]

    Gilliland, R. L.: 1981, Astrophys. J. 248, 1144–1155.

  3. [3]

    Manabe, S. and Stouffer: 1980, J. Geophys. Res. 85, 5529–5554. (See also ref. [6] below.)

  4. [4]

    Gilliland quotes theoretical models giving values of the ratio of amplitudes of variations in radius to amplitudes of variations in luminosity that cover several orders of magnitude, with phase differences of 0 or 180 °. He also quotes observations of pulsating stars with a luminosity to radius phase difference of 90 °. In addition the parameters of the radius variation (ref. [2]) are period = 76 ± 8 yr, 1/2 amplitude = 0.2 ± 0.1 sec, maximum at 1911 ± 4 yr. Because of these various theoretical and observational uncertainties only the period of the luminosity variation is specified when fitting the climate model.

  5. [5]

    Hammer, C. U.: 1977, Nature 270, 482–484. Actually it is the normalised conductivity that is fitted rather than the acidity.

  6. [6]

    Schneider, S. H. and Thompson, S. L.: 1978, J. Geophys. Res. 86C, 3135–3147.

  7. [7]

    Jones, P. D. and Wigley, T. M. L.: 1980, Climate Monitor 9, 43–47. Jones, P. D., Wigley, T. M. L., and Kelly, P. M.: 1982, Mon. Wea. Rev. 110, 59–69.

  8. [8]

    Borsenkova, I. I., Vinniko, K. Ya., Spirina, L. P., and Stekhnovskii, D. I.: 1976, Meteor. Gidrol 1976, No. 7, 27–35 (Russian).

  9. [9]

    Paltridge, G. and Woodruff, S.: 1981, Mon. Wea. Rev. 109, 2427–2434. Not only does the seasurface temperature lag behind the land-based air temperature record but there are several important differences between the two hemispheres that add to the problems described in note [11] below. Sea-surface temperature data have also been analysed by Polland, C. and Kates, F. (Meteorological Office, Bracknell, U.K.), preprint entitled ‘Changes in decadely averaged sea surface temperatures over the world 1861–1980’. The suggestion that sea-surface temperatures would be more appropriate in Gilliland's analysis has been made by Idso, S. (1980) Carbon Dioxide: Friend or Foe, IBR Press (Tempe, Arizona) 92 pp.

  10. [10]

    Bryan, K., Komro, F. G., Manabe, S., and Spelman, M. J.: 1982, Science 215, 56–58. See in particular their Figure 2.

  11. [11]

    The lag between the original forcing and the response in a particular climatic variable is a different question to that of the lags between two different climatic variables as referred to in note [10] and shown in Figure 1. Calculations of the lag between CO2 forcing and air temperature have been given by Hunt, B. G. and Wells, N. C.: 1979, J. Geophys. Res. 84, 787–791; Gates, W. L. and Cook, K. H.: 1980, Clim. Res. Inst., Pub. 14, Oregon State Univ. This type of lag should already be incorporated in the climate model, but, as discussed in the body of this paper, the 5-yr lag seems to be less than is required.

  12. [12]

    The results are very sensitive to the way in which the fitting is performed. Gilliland found that the size of the CO2 signal changed by a factor of 10 depending on whether the 76-yr solar cyle was included. We have obtained a 50% increase in the CO2 signal simply by dropping the last 5 yr of the data. In view of the many uncertainties in the forcing functions [4, 22] and the apparent mismatch between climate model and data, this extreme sensitivity greatly reduces the credibility of Gilliland's results.

  13. [13]

    Pittock, A. B.: 1978, Rev. of Geophys. and Space Phys. 16, 400–420. The critical standards proposed by Pittock have been recommended for studies seeking to detect CO2 signals by Kelly, P. M., Wigley, T. M. L., and Jones, P. D. (1982), 246–51 of Carbon Dioxide Review: 1982, ed: W. C. Clark, Clarendon Press (Oxford) 469 pp. Kelly et al. also comment on the limitations of Gilliland's analysis, particularly with regard to the lack of statistical significance and the uncertainties in the forcing functions.

  14. [14]

    See note 35 of [1].

  15. [15]

    It cannot be assumed that additional data would support the results claimed. Gilliland is, in our view, attempting to reverse the onus of proof at this point. See pp. 46–47 of Pittock, A. B.: 1983, Quart. J. R. Met. Soc. 109, 23–55.

  16. [16]

    Wigley, T. M. L.: 1983, Climatic Change 5, 315–320.

  17. [17]

    Broecker, W. S., Takahashi, T., Simpson, H. J., and Peng, T.-H.: 1979, Science 206, 409–418, describes the ‘missing carbon’ problem. The various attempts to explain the problem include modified ocean uptake — Broecker, W. S., Peng., T. -H., and Engh, R.: 1980, Radiocarbon 22, 565–598; current uptake — Oeschger, H., Siegenthaler Scotterer, U., and Gelmann, A.: 1975, Tellus 27, 168–92; or past biospheric release — Bacastow, R. B. and Keeling, C. D., pp. 247–8 of Carbon Cycle Modelling: SCOPE 16, ed. B. Bolin, John Wiley & Sons (Chichester) 390 pp.

  18. [18]

    Neftel, A., Oeschger, H., Schwander, J., Stauffer, B., and Zumbrunn, R.: 1982, Nature 295, 220–225, find concentration of 260–300 ppmv in air trapped 350 to 500 yr before the present. This approach is of course critically dependent on the assumption that there is no selective movement of CO2 between the air and the ice matrix.

  19. [19]

    Brewer, P. G.: 1978, Geophys. Res. Lett. 5, 997; Chen, C. T. and Millero, F. J.: 1979, Nature 277, 205; Chen, C. T.: 1982, Deep Sea Res. 29, 563–580. The technique looks at water masses that were in contact with the atmosphere over a century ago and deduces from chemical analysis that the water was in contact with the atmosphere when the CO2 concentration was 245–305 ppmv. This method has been criticised because of the uncertainties in the corrections required to allow for carbonate sedimentation and biological activity: Shiller, A. M: 1981, J. Geophys. Res. 86, 11083–11088; Chen, C. T.: 1982, J. Geophys. Res. 87, 2083–2085; Shiller, A. M.: 1982, 87, 2086.

  20. [20]

    Studies by Dr R. Francey in our laboratory have made us sceptical about the use of stable isotope trends in tree-rings as a method of deducing the amount of biological carbon released; Francey, R. J.: 1981, Nature 290, 232–235. Doubts arise due to physiological factors that influence fractionation during carbon assimilation in plants; Francey, R. J. and Farquhar, G. D.: 1983, Nature 297, 28–31. However other laboratories feel that by applying appropriate selection criteria these difficulties can be avoided. Peng, T. H., Broecker, W. S., Freyer, H. D., and Trumbore, S.: 1983, J. Geophys. Res. 88, 3609–3620 have reviewed the available data and conclude that the pre-industrial/pre-agricultural CO2 concentration was in the range 245–260 ppmv and that subsequent releases of biospheric carbon totalled 200 Gt.

  21. [21]

    The direct evidence for the widely-held belief that the pre-industrial CO2 concentration was about 290 ppmv is based mainly on the analysis of earlier data by Callender, G. S.: 1958, Tellus 10, 243–248. Wigley [16] has pointed out that Callender excluded some Southern Hemisphere observations that may have been a more accurate indication of the global CO2 concentrations than the largely land-based Northern Hemisphere data used by Callender. Modern measurements have shown that land-based data are greatly influenced by the diurnal exchange of carbon with the surface vegetation. As Wigley suggested, this may explain the unexpectedly large inter-hemi- spheric gradient observed by Muntz, A. and Aubin, E.: 1886, ‘Recherches sur la constitution chemique de l'atmosphere d'apres les experiences de M. le Dr. Hyades’, Mission Scientifique du Cap Horn 1882–1883 Tome III (2) Gauthier-Villan (Paris). However these Southern Hemisphere observations show a correlation with air temperature that would not be expected from modern experience in the Southern Hemisphere and so must be treated with caution until the methodology has been more carefully evaluated.

  22. [22]

    Moore, B., Boome, R. D., Hobbie, J. E., Houghton, R. A., Melillo, J. M., Peterson, B. J., Sharer, G. R., Vorosmarty, G. J. and Woodwell, G. M.: 1981, pp. 365–385 of Carbon Cycle Modelling, SCOPE 16. Ed. B. Bolin, John Wiley & Sons (Chichester) 390 pp. This work quotes a reference scenario with a release of 148 Gt of biological carbon since 1860. Using modelling studies we have found that a consistent description of the carbon cycle including resolution of the missing carbon problem [17] can be obtained using a release of 160 Gt and a pre-industrial concentration of 264 ppmv. Enting, I. G. and Pearman, G. I.: 1982, ‘Description of a One-dimensional Global Carbon Cycle Model’, Division of Atmospheric Physics Technical Paper No. 42 CSIRO (Australia). 95 pp.

  23. [23]

    We use a CO2 concentration (in ppmv) of \(C(t){\text{ }} = {\text{ C}}_{{\text{ff}}} (t) + C_b (t)\) where \(\begin{gathered} {\text{C}}_{{\text{ff}}} (t){\text{ }} = {\text{ }}300,{\text{ t}} < 1925 \hfill \\ {\text{ }} = {\text{ }}300 exp [4.871 x 10^{ - 5} {\text{ }}(t - 1975)^2 ], t > 1925 \hfill \\ \end{gathered} \) and C b(t) \( = {\text{ }}30 [x/2\sqrt {b + x^2 } {\text{ }} - {\text{ }}{\raise0.7ex\hbox{$1$} \!\mathord{\left/ {\vphantom {1 2}}\right.\kern-\nulldelimiterspace}\!\lower0.7ex\hbox{$2$}}]\) \(x{\text{ }} = {\text{ t}} - 1885\) \(b{\text{ }} = {\text{ }}211.\) C ff is the function used by Schneider and Thompson [6] and by Gilliland [1] to represent changes due to industrial CO2. C b , represents an additional change of 30 ppmv due to biospheric release with 90% of the change occurring in the 90-yr period centred on 1885. While the function C b is based on the results of our carbon cycle modelling, it is represented simply as an example of the type of change that must be considered when looking at the climatic response to CO2. The range of possible release functions is exemplified by the tables given by Bacastow and Keeling [17]. As in earlier applications of the climate model [1, 6] we assume that the standard effect of changing CO2 concentration is to change the radiative term in the model by 5.90187 In (C(f)/300) Wm−2.

  24. [24]

    Gilliland points out that a number of astrophysical phenomena have phase shifts of 90° relative to the radius variation but does not mention that the high precision of the results he obtains is not matched by the solar cycle that he fits. The uncertainties listed in note [4] would imply uncertainties of at least ± 30° in any phase shifts deduced and so no particular importance can be attached to the value of 90°.

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Enting, I.G., Pittock, A.B. & Pearman, G.I. Comments on ‘solar, volcanic and CO2 forcing of recent climatic changes’. Climatic Change 6, 397–405 (1984).

Download citation


  • Climatic Change
  • Dioxide
  • Carbon Dioxide
  • Climatic Variation
  • Recent Analysis