Viewpoint

Warming Warnings

We believe John Daly's "Climate of Opinion [Sept 1990]" article to be in error in a number of important points:

1. The positive feedback effect of water vapour: General circulation models have long included this water vapour feedback. Recent satellite measurements have shown the greenhouse effect increases more rapidly over warmer water than can be accounted for by the increased surface temperature alone.

Daly exaggerates the extent of the positive feedbacks by using an unreasonably low value for the surface temperature increase due to increased CO₂. He quotes an early estimate of 0.25^oC, but later and more widely accepted values range from 0.5^oC to 1.2^oC.

2. Other feedbacks: It is quite true that a potentially important negative feedback involving cloud reflectivity is not allowed for in most models, and recent experiments that do include this show considerably smaller temperature increases.

However, most of Daly's other comments about feedback are incorrect. He claims the ice-albedo effect is a negative feedback when actually it is positive. He claims the models don't allow for increased evaporation producing more clouds -- the models do, but instead predict more intense convection and slightly less overall cloud cover.

Daly seems to confuse the net cooling of the earth by clouds with negative feedback. If cloudiness decreases as the earth warms, the cooling effect is also decreased and the temperature rises further. This is positive feedback.

There are many cloud properties that can change, and modellers are not yet confident that their climate models can simulate all these properties realistically. So it is not yet clear what the overall cloud feedback will be when all these factors are combined.

3. Chaos and climate prediction: Daly's criticisms that weather prediction models can't predict weather patterns much beyond a week is true, but quite irrelevant. Circulation models aim to simulate the types of weather systems that appear in nature. The predicted flow patterns cannot be ascribed to any particular date and, indeed, we would not want the simulated model to depend on the particular situation that we started with. Our aim is rather to see what differences in the model climate result when the CO₂ concentration is changed and the other factors stay the same.

4. Model verification: The models predict the Earth should be 0.6 to 1.3^oC warmer than 100 years ago, with a most likely estimate of 0.9^oC -- not 1.5^oC as Daly asserts. Estimates of the observed warming averaged over the globe after careful correction for the urban bias (another error by Daly) range from 0.3-0.6^oC, most likely 0.5^oC.

This is only about half of the computer-predicted greenhouse warming but Daly ignores the possible contribution of other external forcing factors, such as solar changes or volcanic eruptions, which caused the earlier changes in global temperatures seen in the first figure of his article.

5. Other comments: (i) Observed regional changes that differ from predicted global mean changes do not invalidate the greenhouse theory and would be expected as a consequence of natural variability and local changes in other forcing mechanisms.

(ii) The NASA satellite study spans 10 years, which is so short that almost any trend detected would not be statistically significant. The important result from that study is that global changes measured by satellite agree very well with the irregularly spaced surface observation network, which lends confidence to the longer-term data set from surface observations.

(iii) The time series of solar forcing based on sunspot number has a poor correspondence with the Northern Hemisphere temperature curve. In fact, temperatures correlate better with volcanic dust forcing.

(iv) The increase in New Zealand temperature occurred since 1950 and there have been measurable changes in annual rainfalls which are regionally distinct.

6: The final paragraphs are an unwarranted criticism of Dr Stephen Schneider, who has written elsewhere of the problems facing a responsible scientist in dealing with the media. He expresses the view that careful, reasoned scientific arguments do not attract much attention from the public, and the media look for dramatic statements to attract readers/viewers.

Where there is reason to believe that severe consequences may result, the responsible scientist must do his/her best to draw these to the attention of the public so that the issues can be debated and further research undertaken to resolve uncertainties and point the way to appropriate policies that may need to be instituted.

We acknowledge that computer models still have large areas of uncertainty in their representation of physical processes, but the possibility of future environmental changes with severe economic consequences needs to be taken seriously until further research is able to estimate with confidence the magnitude of the greenhouse effect.

Brett Mullan is a senior scientist with NIWA in Auckland.
Jim Salinger is a senior scientist with NIWA in Auckland.