|
Frigid,
battered by hurricane-force winds, dark half of the year,
and littered with steep mountains and deep crevasses that
could swallow you whole, Antarctica isnt a place
you want to visit without a map. Whats more, you
would want that map to be as detailed and up-to-date as
possible. Of course, hostile conditions make thorough
on-the-ground mapping dangerous, if not impossible. But
thanks to NASA satellite data, scientists visiting or
studying Antarctica after October 2005 will have a significantly
better map of the continents surface than they have
had before.
Using data from the Moderate Resolution Imaging Spectroradiometer
flying on NASAs Terra and Aqua satellites, researchers
at the National Snow and Ice Data Center (NSIDC) at the
University of Colorado and the University of New Hampshire
have assembled the most detailed map of Antarcticas
snowy surface yet produced. Called the Mosaic of Antarctica
(MOA), this map shows the continent with unprecedented
detail. Its an almost perfect picture of the
surface of the continent, says Ted Scambos, lead
scientist at NSIDC. The MOA map includes all land areas
south of 60 degrees South that are bigger than a few hundred
metersin other words, all areas covered by the Antarctic
Treaty
To assemble MOA, Scambos and his collaborators collected
260 MODIS images of Antarctica acquired between November
20, 2003, and February 29, 2004. They selected each image
so that the angle of sunlight in the final map would be
the same for all images, giving all of them the same shadows
and contrast. They checked each image for anything that
would obscure the view of the land surfaceblemishes
or noise in the data, glare from the Sun, clouds, cloud
shadows, fog, and blowing snow. We cut those areas
out, and the resulting images looked sort of like Swiss
cheese, Scambos explains.
Next, they stacked the pieces of Swiss cheese
to cover the continent. By stacking the images, the MOA
team accomplished two things. First, the team managed
to fill in all the holes left by the image-cleaning process
so that no land surface was missing. Second, stacking
the images increased the resolution (amount of detail)
beyond what any single image contained.
Adding pictures is a bit like focusing. We can use
the information in a lot of images to make a final picture
thats a little bit better than any single image,
explains Scambos. We stacked at least four images
to make every part of the mosaic. On average, we stacked
15 images. In one area, 38 images overlapped.
To make MOA, the researchers started with MODIS
250-meter-per-pixel resolution data, but the resulting
image has even higher resolution. It doesnt
make sense to stray too far from the resolution of the
satellite data, Scambos says. But to show
all the features in the new image, you must use a finer
scale. MOAs estimated resolution is 150 meters
per pixel.
The MOA imagery has a grayscale color scheme, but in
a land of snow and ice, color isnt as crucial as
it would be in forested or farmed land. Some patches of
color do appear on the ice, however. For example, over
time, windswept snow can eventually transform into pure
ice composed of big crystals. On the ground, this thick,
clear ice appears blue. To identify colored features like
blue ice, scientists can make pseudo-color images by combining
the grayscale MOA images with MODIS data from different
parts of the light spectrum, including red and infrared
light.
Detailed as it is, MOA is not intended to work in isolation;
scientists designed it to integrate well with other images.
For example, NASA and the Byrd Polar Research Center previously
published an image of Antarctica based on radar data from
the Radarsat Antarctic Mapping Project Antarctic Mapping
Mission 1 (RAMP AMM-1). In fact, we used RAMP data
as a reference when we built MOA, says Scambos.
RAMP is great for looking at subsurface features.
MOA focuses on the surface. So depending on what youre
looking for, you can use either mosaic alone, or combine
them [MOA and RAMP maps] for a more complete picture.
-For
the complete article Earth
Observatory -
|