Keller uses Earth System Analysis to improve Climate Decisions

Klaus Keller (photo)


Imagine you have to decide whether and how much to raise a levee to protect a coastal community from the projected rise in sea levels associated with climate change.

You know that sea levels have risen during the last century and are projected to continue to rise-leading to an increased risk of flooding absent of adaptations. You also know that raising the levee will reduce the flooding risk.

The question is how much would you raise the levee?

One option is to heighten the levee by a small amount, hoping that the future sea level rise will be at the low end of the projections. A more expensive but arguably safer option is to build a higher levee that can handle the high-end of the sea-level projections.

"A workable risk management strategy has to be scientifically sound, economically efficient, and ethically defensible—which requires an interdisciplinary approach," says Klaus Keller, director of the Center for Climate Risk Management (CLIMA) and associate professor of geosciences. "Analyzing those strategies requires that natural and social scientists and ethicists collaborate effectively—a non-trivial challenge.

"However, this is a stimulating challenge because the collaborations between decision makers, stakeholders and natural and social scientists can lead to exciting, mission-oriented basic research," Keller says.

CLIMA helps this research by catalyzing the interdisciplinary interactions that improve the analysis and design of climate risk management strategies.

To return to the example of how high to raise the levee, the design of flood risk management strategies is often driven by low-probability, high-impact events such as one-in-100- or one-in-10,000-year flooding. Characterizing those low-probability, high-impact events requires mechanistically sound models that are carefully tested and calibrated with instrumental and paleo observations.

Keller acknowledges that all models are imperfect, but argues that some can still be useful to improve decisions.

"The basic choice is often: Do we wait for better data and models and in the meantime neglect low-probability, high-impact events or do we analyze them with imperfect but still useful models?" Keller says. "Neglecting low-probability, high-impact events can lead to downward biases in risk estimates."

While climate science is a well-established and active science, considerable uncertainties—such as about future sea levels—still remain. But uncertainty does not have to result in paralysis.

We make many decisions under deep uncertainty: which career path to pick, whether to marry or whether to purchase life insurance. Hedging against a low-probability but high-impact event—by purchasing life insurance, for example—can be a rational decision, a strategy that Keller says also can make sense for dealing with climate change.

"Investments to reduce the vulnerability of current ecological and economic systems to future climate change can be interpreted as insurance strategies," Keller says.

Keller earned an Engineer's degree from the Technische Universität Berlin, an M.S. in Civil and Environmental Engineering from M.I.T., an M.A. in Civil engineering and Operations Research from Princeton University, and a Ph.D. in Civil and Environmental Engineering from Princeton. He was a contributing author to the Intergovernmental Panel on Climate Change’s (IPCC) Fourth Assessment Report. The IPCC shared half of the 2007 Nobel Peace Prize.