But the computer model said the tsunami wouldn't be this high!
There are two ways of evaluating geologic hazards. The first way is to develop sophisticated and complex computer models that will numerically reproduce the geologic process that is being studied. Then, based on the computer model, various scenarios can be tested and an evaluation can be made of just how great various geologic hazards are and what areas are most likely to be affected. Computer models define hazards based on assumptions about the nature of future events.
The second way to evaluate geologic hazards is to study the geology of the area at risk to determine what has actually happened there in the past. The premise of this approach is that if some geologic event has happened in the past, it is possible it will also happen in the future. The geologic studies define a maximum credible event that can occur in the future based on what has actually happened in the past.
Ideally these are complementary approaches---the geologic data showing what has actually happened in the past can provide a test and a calibration of the accuracy of the predictions made by computer models. However, it doesn't always work that way in practice. Consider the recent Fukushima Disaster in Japan, where an earthquake generated a tsunami wave that damaged the Fukushima nuclear plant.
Before the nuclear plant was built the Japanese created sophisticated models of earthquakes and tsunami generation, and based on the computer models they believed that the Fukushim nuclear plant site was too high above sea level to be affected by tsunamis. Based on the predictions of those computer models, the Fukushima nuclear power plant complex was approved and built.
However, a geologic hazards analysis came to a very different conclusion. Deposits of the 869 A.D. Jogon tsunami can be found widely in the Fukushima region, including at the nuclear plant site. Based on the geologic data, the maximum credible event was a tsunami large enough to damage the Fukushima nuclear power plant. And, of course, on March 2012 another tsunami similar in size to the 869 Jogon tsunami occurred and destroyed the Fukushima nuclear power plant. With hindsight it is clear the numerical models significantly underestimated the actual hazards, because the modelers did not completely understand the natural system they were modeling.
Global Climate Models (GCMs) are among the most sophisticated numerical models ever created. They attempt to model Earth's climate by defining interactions for every natural process in the atmosphere, ocean, biosphere, and cryosphere. Many of these processes are not well understood. One lesson of the Fukushima disaster is that predictions from computer models shouldn't be accepted unskeptically, especially if the processes they are modeling are not completely understood. One risk is that serious hazards may be underestimated if the hazards appraisal is solely based on the GCM output. If possible, GCM predictions should be tested and calibrated based on geologic studies of past climate change just as other numerical models of hazards are tested and calibrated.
Perhaps the greatest danger from Global Climate Change is sea level rise, as millions of people live in coastal areas that are at risk. The International Panel on Climate Change used GCMs to estimate that sea level would rise by at most 1 meter by the year 2100. But many glaciologists think sea level rise could be much greater based on recent increases in the amount of glacier melt occurring in Antarctica and Greenland.
What is the maximum credible sea level rise event based on the geologic evidence of past climate change? Data from around the world shows that sea level reached five meters above current levels during the last interglacial period. The geologic evidence is very clear----if Greenhouse Warming continues unabated, we are highly likely to see 5 m (15 feet) of sea level rise in the future.
No comments:
Post a Comment