BALTIMORE, MD.- Small changes in global climate patterns
could produce great increases in the frequency of catastrophic floods like
those seen recently in the Missouri and Mississippi River valleys, according
to researchers at Penn State University.
"If the relative frequency of large floods increases from one year
in 50 to one year in 20, that will radically affect where and how people
will live in the future," said Dr. Ana Paula Barros, assistant professor
of civil engineering, Penn State University, in a presentation at the spring
meeting of the American Geophysical Union.
Dr. Barros and colleagues are evaluating global climate models in an
attempt to predict how climate change could impact flood cycles in local
areas. For example, they have looked at interconnections among temperature,
precipitation, El Nino patterns and the occurrence of large floods in the
Upper Mississippi River Basin.
"We are trying to find measures of the spatial structure of these
variables which could be used as indexes relating climate variation to the
occurrence of large floods," said Barros.
This approach shows promise because the duration and intensity of specific
precipitation events are not the only leading factor of substantial flooding,
but rather, it is the space- time history of precipitation over the landscape
that determines the magnitude and duration of a flood. In other words, flooding
occurs because certain events accumulate and occur over time, not because
there is one massive rain storm, she explained.
An analysis of data from various sources led Barros to develop two monthly
single-number indexes that measure the changes through time of the spatial
textures of precipitation fields for the continental United States. Graphing
both these indexes using historic records reveals a distinct pattern that
always appears to be followed by flooding, said Barros.
The graphs clearly show repeated spikes indicating known times of flooding
and an identifiable pattern that precedes these spikes.
"The historic record seems to show us a pattern to look for to infer
the occurrence of flooding in the global models. This pattern seems to be
related to the occurrence of strong El Nino events," she noted.
Standard climate models can predict precipitation, but not floods because
the models do not have the fine spatial resolution or the complex hydrologic
components needed to simulate runoff and flood routing. However, climate
models do provide all the information needed to generate these kind of indices,
she said:
"If we ran the climate models with expected global warming changes,
perhaps these indices could be calculated and the frequency of flooding
under those conditions determined."
Having reliable long-term predictive models on hand could help civil
planners anticipate disasters before they happen and make contingency plants,
she said. Certain areas could be left unbuilt, and other areas planted with
trees or other flood control methods could be instituted, she reported.
Check out Econet for environmental news and information
