U.S. Energy Secretary Steven Chu says that climate models that don't include the impact of "tipping points," aren't measuring all the risks posed by climate change.
What is a tipping point? In climate change it is the point that will lead to cascading events, positive feedback loops, such as rising temperatures that result in the melting of the Greenland ice, leading to higher sea levels and who knows what else.
Chu spoke Thursday night at the American Association for the Advancement of Science. It was a talk, by this Nobel Prize winner in physics, that had some parallels to the problem of measuring risk in IT.
There will be certain risks and damages that might occur if the world temperature goes up three degrees centigrade, said Chu. "The question you should ask yourself if it goes up six degrees centigrade would it be four times worse, would it be two times worse, or will it be a whole lot worse," he said.
Most climate scientists don't want to put these tipping points in their models because of the huge uncertainties, said Chu, but that means the models don't show the full risk.
"To be sure, if you start to model the tipping points you put in much larger uncertainties, but there is a difference between uncertainty and inaccuracy," said Chu.
The "long tail of the damage tail is out there," said Chu, who urged climate researchers to include tipping points in their models.
If scientists are averse to pushing their models, the reasons are understandable. Creating models for anything that include worse-case scenarios and positive feedback loops are potential targets of ridicule.
Chu is arguing that if you avoid the larger and more uncertain questions, the model will have problems. But most people, IT managers included, will limit their assessments of risk. How can we know this? A recent survey by the data center association AFCOM found that 15% of the data centers have no plans for backup and recovery and about 30% don't have backup sites, a situation the group called "shocking."
Japan may offer a good illustration of just how bad things can get when risks are underestimated, according to Kerry Emanuel, a professor of atmospheric science at MIT, who testified Thursday on climate change before the House Science Committee.
Emanuel cited Japan's recent 9 magnitude earthquake as an example. Seismologist estimated that the largest equarthquake that one could reasonably expect was about an 8.2 magnitude. "For this reason, the Fukushima-Daiichi nuclear power plant was not designed to withstand the magnitude of earthquake and tsunami that disabled it," he said.
"In our own country, the levees that protect New Orleans were designed for storm surge events somewhat less severe than we now believe are likely there. And, in the climate arena, summertime arctic sea ice has been declining somewhat more rapidly than had been projected," said Emanuel.
"Far from being alarmist, scientists have historically erred on the side of underestimating risk," said Emanuel.