Climate Studies Point to More Floods in This Century

Hillary Mayell
for National Geographic News
January 30, 2002
Two separate teams of scientists are predicting more extreme rainfall
and greater flooding in this century.

According to their
projections, it will be particularly striking at northern
latitudes—across Canada, Alaska, northern Europe, and northern
Asia, regions that already receive the most precipitation. But the
equatorial tropics and Southeast Asia are also likely to have increased
rainfall and flooding.

Both teams, one from the United States and the other from Europe, attribute the expected pattern to global warming accelerated by human activities.

Although people may adapt to gradual climate change, the effects of extreme rain and flooding are often broad, devastating, and costly to society. Landslides, avalanches, and flooding damage infrastructure such as roads, bridges, and buildings, and hurt agricultural productivity because of lost crops and soil erosion. Disaster relief often requires enormous funding, and the loss of human life may also be high.

The two new studies, published in the January 31 issue of the journal Nature, point to trends and probabilities, not specific events. But they emphasize the need for more sophisticated climate models to help decision makers.

Seeking Trends

Christopher Milly, an atmospheric scientist at the U.S. Geological Survey, and his colleagues reviewed data on 100-year floods that occurred in the last century for 29 major river basins around the world.

"By definition, a 100-year flood is really extreme and rare," said Milly. "What we can observe when we look at those records is that the number of these extreme flooding events occurred disproportionately in the last decades of the 20th century. The difference is large enough to make you raise your eyebrows. It's hard to believe it could happen by chance, enough that it's worth looking for other reasons why there were so many floods in the last few decades."

Scientists use computer models to predict possible changes in climate. The models uses a variety of data on various conditions, such as atmospheric concentrations of greenhouse gases and ocean circulation patterns. By changing these variable conditions, scientists can simulate what might happen in the real world when climate conditions change.

Milly and his colleagues combined a climate model with a river model, tied them together, and examined different scenarios to identify trends. The trends for the 20th century, particularly at high altitudes and in equatorial regions, closely mirrored what had actually happened, giving the scientists a fair degree of confidence in the model.

Based on the model's projections, said Milly, it's reasonable to assume that increased flooding over the past century was related to global warming, and the increase is likely to continue.

"The models suggest that instead of the chances of a 100-year flood occurring once every 100 years, which is what you would expect, the risk will increase in the 21st century to somewhere between 3 to 6 chances in 100, which is a manifold increase," he said.

In the other study, Tim Palmer of the European Centre for Medium Range Weather Forecasts in the United Kingdom and Jouni Räisänen of the Rossby Centre in Sweden studied increased rainfall rather than flooding, but arrived at similar conclusions.

Analyzing data from 19 climate models, they concluded that much of central and northern Europe will be five times more likely to suffer very wet winters during this century. They also predict the Asian monsoon region will experience a five-fold increase of very wet summers.

Building Better Models

In both studies the scientists tied the increase in extreme climatic conditions to a rise in global temperatures driven largely by an increase in atmospheric greenhouse gases resulting from human activities.

These gases, primarily carbon dioxide, methane, and nitrous oxide, trap heat in the atmosphere, causing an increase in temperature. The increase in temperature sparks an increase in evaporation. "Warm air can hold more water," said Milly. "[Clouds] tend to suck more water up out of the ocean, and it has to drop sometime."

But climate modeling has limitations, and the authors warn that several factors could skew the results of the trend analyses. Carbon dioxide emissions are overestimated in the models, for example, because other gases such as methane are not included. Sulphate aerosols, which tend to have a cooling effect, are also not included in the computations. In addition, changes in global vegetation and land use could affect the accuracy of the models.

In a related article, Reiner Schnur of the Max Planck Institute for Meteorology in Germany argues that studies such as these make a strong case for improving computational resources in climate research so the results will be useful to decision makers.

"Until computational power increases significantly, climate scientists will have to patch models together, taking the results of ensemble climate projections, for example, and inserting their output into a high-resolution hydrological model for a specific river basin," he writes.

© 1996-2008 National Geographic Society. All rights reserved.