August
2007
Before the flood
NCAR wades into forecasting dangerous
waters
The Ganges and Brahmaputra rivers
pass through Bangladesh, converging and forming the
largest intertidal delta in the world. The rivers
and their tributaries deposit massive amounts of
silt and clay that create a maze of waterways and
islands in the Bay of Bengal, as seen in this satellite
image. (Image courtesy of the Image Science & Analysis
Laboratory, NASA Johnson Space Center.) |
Epic floods around the world this summer have forced thousands
from their homes in Britain, spurred emergency evacuations
in Texas, and killed hundreds in China. But perhaps no place
in the world is as vulnerable to floods as Bangladesh, where
more than 40% of the land area is within about 30 feet (10
meters) of sea level. As of early August, millions of Bangladeshis
had been marooned or displaced by floodwaters, with the death
toll at more than 1,000.
What’s more, such catastrophes are a recurring event
in this densely populated, low-lying nation. Two great rivers
that cut through the country—the Ganges and Brahmaputra—can
adversely affect millions of people when they overflow their
banks every few years. Farmers and fishers can easily lose
a year’s worth of income during a single flood. And
with climate change expected to cause a greater number
of heavy rainfall events in the future, scientists are studying
the likelihood of more severe flood scenarios in Bangladesh
and other low-lying places.
But thanks to a collaboration between RAL, Georgia Institute
of Technology, and the Asian Disaster Preparedness Center,
now Bangladeshis in flood-prone areas can at least expect
more advance warning of rising waters. The team has developed
a high-tech system for predicting floods in the country.
This summer, its forecasts began reaching the people who
need them most.
“Our goal is that long-range flood forecasts, for the
first time, will consistently reach many rural individuals
in Bangladesh who are in jeopardy of losing their homes,
businesses, and possibly their lives,” says Tom Hopson,
a postdoctoral fellow in RAL who is NCAR’s leader on
the project. “It’s hugely satisfying to see these
forecasts being used to help people.”
The system uses a combination of weather forecasts, satellite
observations, river gauges, and hydrologic modeling techniques
to predict when major rivers will crest in selected regions
of the country during the monsoon season, which runs approximately
June–October. The 1- to 10-day forecasts are then distributed
to more than 100,000 residents of districts adjacent to the
Ganges and Brahmaputra. Because few households in this extremely
poor region have radios or even electricity, the nonprofit
Asian Disaster Preparedness Center disseminates forecasts
by activating a network of governmental and nongovernmental
organizations, as well as volunteers, to spread the word
in person.
The forecasts will provide Bangladeshis with more advance
notice of floods than they’ve had in the past, giving
farmers a chance to either postpone planting or harvest at
least a portion of their crops, move livestock to safety,
encircle fish ponds with nets to prevent fish from escaping,
and stock food and other supplies. Over the next year or
two, the system will begin providing 20-day forecasts, to
be eventually followed by seasonal forecasts.
Tom Hopson. |
“The goal here is to help very local, grassroots economies,” Tom
says. “The forecasts can also alert relief agencies
to prepare to bring in drinking water, cholera tablets, and
other essentials in case of a major flood.”
The system is based on an ensemble of weather predictions
for South Asia generated by the European Centre for Medium-Range
Weather Forecasts. It incorporates estimates of precipitation
from satellite observations provided by NASA and NOAA, along
with discharge measurements from Bangladesh’s rivers.
Updated daily
with new information, the system emphasizes modeling and
satellite data
to compensate for a lack of river gauge and radar data upstream
of Bangladesh.
“This is cutting-edge technology, with which we analyze
information from a number of sources to generate forecasts
of the probability of major flooding,” Tom says.
The team, which includes Georgia Tech’s Peter Webster,
who is the principal investigator, and Ramasamy Selvaraju
of the Asian Disaster Preparedness Center, hopes to expand
the system to encompass more of Bangladesh over the next
few years, since past floods have submerged as much as two-thirds
of the nation. The team may also study the feasibility of
applying the forecasting system technology in other countries
that are vulnerable to floods, such as Cambodia and Ghana.
Front
Range flash floods |
Closer
to home, RAL’s Dave Gochis and David Yates
are developing a flash flood prediction system for
the Front Range’s South Platte River Basin,
spanning the area from Cheyenne to the Palmer Divide
north of Colorado Springs.
The system will provide forecasters, emergency
managers, and the public with flash flood forecasts
that are more targeted than those currently available,
giving up to a few hours’ notice of potential
flash floods in specific drainages.
The system combines data from meteorology and hydrology,
disciplines that scientists often struggle to link.
On the weather side, it utilizes radar data, short-term
forecasts and a high-resolution version of the Weather
Research and Forecasting model (WRF), while on the
hydrological side it incorporates rain gauge and
stream flow data and physically-based hydrology models.
“This project is about integrating science
and technology from two scientific disciplines that
have evolved faster individually than at their interface,” Dave
Gochis says.
The two plan to perform case studies this summer
by simulating past Front Range flash floods, including
the deadly Fort Collins flood of 1997. The system
will undergo operational testing on a trial basis
next summer when forecasts are shared with the Denver
Urban Drainage and Flood Control District and the
Denver/Boulder National Weather Service. |
In this issue...
Before
the flood
Storm
World author comes to Center Green
Another
successful year for leadership programs
Mary
Marlino to head NCAR Library, e-Science
Random
profile: Jennifer Boehnert
CISL
cultivates the next generation
Double
trouble on the storm front
Delphi
Question
Just One Look
Staff Notes home page | News Center
|
