HR: 16:00h
AN: U54A-01 INVITED [Abstracts]
TI: Modeling the Climatic Consequences of Geoengineering
AU: * Somerville, R C
EM: rsomerville@ucsd.edu
AF: University of California, San Diego, 9500 Gilman Drive, Dept. 0224, La Jolla, CA 92093-0224
United States
AB:
The last half-century has seen the development of physically comprehensive computer models of the climate system. These
models are the primary tool for making predictions of climate change due to human activities, such as emitting greenhouse
gases into the atmosphere. Because scientific understanding of the climate system is incomplete, however, any climate model
will necessarily have imperfections. The inevitable uncertainties associated with these models have sometimes been cited as
reasons for not taking action to reduce such emissions.
Climate models could certainly be employed to predict the results of various attempts at geoengineering, but many questions
would arise. For example, in considering proposals to increase the planetary reflectivity by brightening parts of the land
surface or by orbiting mirrors, can models be used to bound the results and to warm of unintended consequences? How could
confidence limits be placed on such model results? How can climate changes due to proposed geoengineering be distinguished
from natural variability?
There are historical parallels on smaller scales, in which models have been employed to predict the results of attempts to
alter the weather, such as the use of cloud seeding for precipitation enhancement, hail suppression and hurricane
modification. However, there are also many lessons to be learned from the recent record of using models to simulate the
effects of the great unintended geoengineering experiment involving greenhouse gases, now in progress. In this major
research effort, the same types of questions have been studied at length.
The best modern models have demonstrated an impressive ability to predict some aspects of climate change. A large body of
evidence has already accumulated through comparing model predictions to many observed aspects of recent climate change,
ranging from increases in ocean heat content to changes in atmospheric water vapor to reductions in glacier extent. The
preponderance of expert opinion is that this evidence is now sufficient to establish the human cause of much recent climate
change.
Nevertheless, no model can provide detailed and fully trustworthy answers to every possible question of interest. As an
example, how will the climatology of Atlantic hurricanes change as the greenhouse effect becomes stronger? Can models
reliably forecast changes in the length of the hurricane season or changes in the geographical regions affected by
hurricanes? The answer is no, or at least, not yet.
Additionally, climate models are not based entirely on first principles, such as Newtonian physics. Instead, they have been
developed primarily to simulate the present climate and relatively small departures from it. To achieve this goal, a certain
amount of empiricism has been built into the models. The result has sometimes been to increase the apparent realism of
models at the cost of limiting their generality. Thus, the available climate models may well be less capable of simulating
a geoengineering experiment that might lead to a radically different climate. New model development may be required for this
new application.
The challenge is to distinguish between what models can and cannot do well. It would be irresponsible and unethical, either
to undertake geoengineering projects without modeling their consequences, or to place blind faith in the models. To decide
how best to model a proposed geoengineering technique requires a deep understanding of the strengths and weaknesses of
climate models. The history of modeling successes and failures is a valuable guide to the wise interpretation of model
results.
DE: 1615 Biogeochemical cycles, processes, and modeling (0412, 0414, 0793, 4805, 4912)
DE: 1620 Climate dynamics (0429, 3309)
DE: 1622 Earth system modeling (1225)
DE: 1626 Global climate models (3337, 4928)
DE: 6309 Decision making under uncertainty
SC: Union [U]
MN: Fall Meeting 2005