New way to measure
atmospheric moisture may improve weather forecasting
BY LISA TREI
More accurate weather
forecasting soon may become a reality thanks to a new way
of pinpointing the amount of moisture in the atmosphere.
In a paper published in Science
on Feb. 26 called "High-Resolution Water Vapor
Mapping from Interferometric Radar Measurements,"
researchers show that radar signals can be used to study
spatial atmospheric structure.
"The major advantage
of studying the weather with this approach is that it
allows you to measure water vapor distributions in the
atmosphere at very, very high spatial resolutions,"
said Howard A. Zebker, associate professor of electrical
engineering and geophysics and one of the authors of the
paper.
The technique allows
measurements to be made every 10 to 100 meters over wide
areas. "That's a very much finer density than you
usually see on weather maps," he said.
The technique, called
synthetic aperture radar (SAR) interferometry, is an
application using satellites orbiting 500 miles above
earth; it can produce images of very small surface
movements. Zebker has used radar interferometry to look
at earth movements connected to earthquakes and
volcanoes, but he says that researcher Ramon F. Hanssen,
the paper's lead author, has demonstrated that it also
can be used to observe changes in water vapor.
While Hanssen was a
visiting scholar at Stanford last year, he used SAR
interferometry to produce data on water vapor over a part
of Holland. Zebker explained that the satellite sends a
wave that reflects off the surface of the ground and back
up to the satellite. With SAR, the scientists can measure
the propagation delay for a particular ray and use that
to measure water vapor. After collecting data, Hanssen
obtained corroborating weather records to show that
different weather phenomena have signatures that can be
identified in the radar.
Zebker said the technique
measures "columnar abundance," which is the
total amount of water vapor from the ground level to the
top of the atmosphere. "We can't tell if that water
vapor is in a fog sitting down on the ground or if it's a
cloud at the top of the troposphere, or even if it's
invisible water vapor in the air," he said.
Despite the technique's
inability to distinguish the height of the water vapor
being measured, it can illustrate water vapor
distributions associated with precipitating clouds and
partly precipitating cold fronts. It also can measure
water vapor in horizontal convective rolls, where the
atmosphere circulates, similar to the way waves propagate
in the ocean.
Zebker said that the
meteorological community is intrigued by the new
technique although at first it will be difficult to
assimilate into existing meteorological models.
"The main problem is
that these data are produced every 10 meters on the
ground and most weather models break the earth up into
much larger squares, perhaps 110 kilometers [in length]
on one side," he said. "So there's a remaining
step that needs to be done in the meteorological
community in looking at micro-weather phenomenon."
Zebker said that weather forecasting is increasingly
moving toward microclimate observations, especially in
densely populated regions such as the Bay Area where the
climate changes dramatically within a small area. "I
think what you're going to see is people developing
regional weather models and using high-resolution data
such as what we're beginning to produce," he said.
In addition to the gap
between the old and new weather forecasting technology,
practical limitations exist, namely that the ground being
measured must remain essentially unchanged over time.
Vegetation is a problem because it grows constantly,
Zebker said. The European satellite being used to make
measurements, the ERS-1, operates a 6-centimeter
wavelength radar system that is quite sensitive to
vegetation growth. But, Zebker said, there are plans for
NASA to design a new radar system with a 25-centimeter
wavelength that will be much less sensitive to plant
growth.
Proposals for this
program, called LightSAR, will be completed soon.
Construction of the satellite, expected to cost between
$150 million and $200 million, should start this fall and
be ready for launching in 2002 or 2003. "LightSAR is
only one satellite," Zebker said, and "it only
repeats every eight days. That's really not often enough
for detailed weather measurements."
If meteorologists and
researchers decide to embrace this technology, up to five
satellites will be needed to allow measurements of more
parts of the earth, Zebker said. Each of these would cost
a fraction of the original LightSAR satellite, he added.
SR
|
