How reliable are climate predictions?
Professor Lennart Bengtsson
Max Planck Institute for Meteorology, Hamburg
Environmental System Science Centre, University of Reading, UK
Abstract
A critical assessment of recent work on climate change will be done. This will include a discussion on the question of climate predictability, aspects on the role of internal climate variability and the relation between the forcing from greenhouse gases ( and aerosols) and the way the climate system is responding to such forcing.
Climate change is a slow process which requires reliable statistics over a period covering 30-100 years. We are just at the borderline to detect such changes against a background of natural variability.
However, all indications are that the warming trend during the last 50 years is consistent with the increase in greenhouse gases. In view of the apparent long residence time of greenhouse gases in the atmosphere we cannot exclude that we have to face a major aberration of anthropogenic nature in the Earth’s climate during the next hundred years or so, which will be too risky to ignore. Precisely how it will evolve is not possible to outline, but regional drought problems are expected to be among the most serious consequences. And in the longer perspective we may see changes similar to the large changes in the climate of the Earth that have evolved over time scales of tenth- to hundred thousand years.
Unfortunately, it will be significantly more difficult to solve than the successful outcome of the stratospheric ozone issue, although the excessive political focus on carbon dioxide in this respect is surprising. Presently we need more knowledge before drastic measures are taken as these might cause more harm than help. The ethanol production is here a warning example. Not even more energy is needed for its production but the increased emission of nitric oxide will more than offset the reduction of carbon dioxide. Finally, any political decisions must include the repercussions of an energy starving world.