Early warnings about global warming included massive flooding of coastal areas as glaciers melted and sea levels rose. Concern focused on the ice sheet in Antarctica, which was expected at some point to succumb to rising surface temperatures as heat-trapping gases built up in the atmosphere.
The dire predictions came from computer models that have made erroneous predictions about temperature increases in the past 100 years, about relative warming speed in the Northern and Southern Hemispheres, and about extent of warming in high vs. low latitudes in the Northern Hemisphere (OGJ, Apr. 27, p. 13). Computer model forecasts about glacial melting appear to be no better. Based on computer models, scientists in 1980 projected a possible 25 ft sea level increase associated with greenhouse warming in the next 150 years. After further analysis, they have moderated their expectations about a sea level rise to 1 ft or less.
FAMILIAR PATTERN
It's a familiar pattern: As computer models take increasing account of natural processes, the future consequences of global warming look less grim. The Antarctic ice sheet now seems safe. If temperatures do rise, snowfall will increase, adding more water to the ice sheet than it loses through melting. What's more, temperatures might not rise much. When computer models account for ocean currents, warming predictions drop. Before adjustments for ocean currents, one model projects a global temperature increase during the next 50 years of 2-3 C. After the adjustments, the model predicts temperatures rise by about the same amount in the Northern Hemisphere but by only 1 C. in the Southern Hemisphere; the Antarctic Ocean warms only slightly or even cools.
Another natural process that won't hold still to accommodate computer models is cloud formation. Clouds have both heating and cooling effects: They trap reflected heat, and they block incoming solar radiation. Their effects on long term temperature trends depend on how they change in response to initial warming. Those changes are extremely difficult to predict, but sensitivity of warming forecasts to them is great. A study by the U.K. Meteorological Office reduced computed greenhouse warming to 1.9 C. from 5.2 C. when it assumed simply that initial warming replaced ice-crystal clouds with those containing water droplets and that cloud reflectivity increased.
NATURE'S ADAPTABILITY
Nature's wondrous adaptability, of which cloud changes and ocean currents are just two examples, plays havoc with computer models. Natural variability creates even more problems. Greenhouse gas concentrations fluctuated before humans existed. Temperatures rose and fell. Observed warming may result more from changes in solar activity and the earth's position relative to the sun than from a greenhouse gas buildup. Human warming thus may be inconsequential-or nonexistent. Last week, George C. Marshall Institute scientists said satellite measurements show little change in the planet's temperature during the past 12 years, contrary to popular assumptions and computer model predictions.
The earth has not shrunk, and nature has become no less complex because human beings possess computers and have begun to assess their ecological imprint. Humans still cannot measure and predict global climate change with precision. The very idea of an average global temperature presumes much. Global warming looks less threatening as knowledge about it grows. Until scientists understand the phenomenon better than they do now, politicians have no basis upon which to impose huge, possibly unnecessary, remedial costs-about which, more later.
Copyright 1992 Oil & Gas Journal. All Rights Reserved.
