The computer program, called MultiSpec, analyzes data gathered by earth-orbiting satellites, such as Landsat, and creates an image that can be displayed graphically on a computer. Scientists or students then compare the satellite images with field observations taken on the ground, such as types of land cover and species of vegetation.
"Knowing the characteristics of a relatively small area of land, and how those characteristics are represented in the satellite images, scientists can then look at satellite images from a much larger, surrounding area and extrapolate what the conditions in that area are, without having to make extensive, time-consuming field observations," says David Landgrebe. He is a Purdue professor of electrical and computer engineering who develops techniques and instruments for remote sensing.
For example, if a scientist observes trees damaged by acid rain, and then matches those observations with satellite images from the same area, the scientist can look at satellite data from a larger area and determine whether the vegetation there also is damaged.
"MultiSpec does all that for you," Landgrebe says. "That's the whole goal in remote sensing."
Landgrebe and his research associate Larry Biehl originally developed MultiSpec about eight years ago as a convenient, inexpensive and easy-to-use tool for earth scientists and environmental scientists to analyze their own data on Macintosh computers.
"Our system is unique in that many of the current data analysis systems require more expensive computer hardware to run, which is out of reach for many college-level students and all students at the K through 12 level," Landgrebe says.
"Now, we have MultiSpec up on the Web where anyone can download a Macintosh or Windows version for free, though the Windows version does not yet have all the capabilities of the Mac version." The Web address is http://dynamo.ecn.purdue.edu/~biehl/MultiSpec/.
Schoolchildren began using MultiSpec about four years ago when Barry Rock, a professor of forestry at the University of New Hampshire, taught a summer program for high-school students in which they studied the acid rain problem in a forest in their area. Students first made field observations, then Rock brought them back into the lab, where he gave them Landsat data of the area and a copy of MultiSpec.
"The students analyzed the data from the area they just observed, so they had a better understanding of what was going on in their surrounding area," Landgrebe says. "A few weeks after that, we began to get requests from all over New Hampshire for copies of MultiSpec. It was immediately exciting to the kids."
Rock was a chief scientist for an international program called Global Learning and Observations to Benefit the Environment, or GLOBE, that was started in 1995. GLOBE is a worldwide network of elementary and high-school students, teachers and scientists working together to study and understand the global environment. There now are about 3,000 GLOBE schools in more than 30 countries.
GLOBE students make a core set of environmental observations at or near their schools and report their data to a central location via the Internet. Scientists use GLOBE data in their research and provide feedback to the students. Each day, images created from the GLOBE student data are posted on the World Wide Web at http://www.globe.gov/.
"Just going out and collecting data and slipping it through an electronic crack in the wall was ultimately not going to be very exciting for the kids," Landgrebe says. "So Professor Rock wanted to give each school some Landsat data from their area and MultiSpec, plus training for teachers."
Landgrebe says that about 40 percent of the MultiSpec downloads now come from GLOBE schools. He says more than 1,000 copies of MultiSpec have been downloaded so far this year, compared with 529 during all of 1995.
Periodic training sessions on MultiSpec also are held around the country. "I can tell when a training session is held, because I will get a burst of downloads from that area," he says.
"Just looking at what these kids are able to do is amazing. I was at a meeting with some of these teachers from across the country, and one had a three-ring binder with photos of the computer screen of work that fifth-graders had done. They had analyzed the forestry surrounding their area and classified it into different classes of forest cover. That's something I teach in my graduate course!"
Bob Jost, a GLOBE teacher at the Manchester GATE School in Fresno, Calif., has been using MultiSpec in his sixth-grade classroom for the past two years and has conducted nine workshops to train GLOBE teachers on the use of MultiSpec. As an introduction to remote sensing, Jost's students first use MultiSpec to view and manipulate Landsat images of the area around their school, identifying features such as streets and the local university's football stadium. They also use MultiSpec to study images of a 30-meter square plot of land on the school campus that serves as their GLOBE biology study site.
"I especially liked using satellite images," says Jordan Rodriguez, 11, one of Jost's students. "We get to see what we normally don't see, the infrared colors that only a satellite can see. It allows us to see heat, moisture and other things."
Students go out to the study site twice a year to identify species of trees, measure the height and circumference of trees, and calculate the percentage of canopy cover and ground cover. Eventually students will monitor changes in the plot by comparing current Landsat images with those from 1993.
"This is exactly what students in the '90's need, something that challenges them, gives them a new way of looking at their world, technology skills that have real applications, and opens the doors to real science," Jost says. His classroom's Web site is http://www.cybergate.com/~bjost/rm21.html.
Remote sensing researchers also use MultiSpec, but in a more sophisticated way.
"We heard from one geographer who took his laptop computer on a field trip in Africa and he was able to use MultiSpec to bring up a Landsat picture of the area," Landgrebe says. "He was very excited, because he could immediately and directly relate to the ground cover he was looking at."
Instruments on board earth-orbiting satellites such as Landsat detect light that is reflected off the surface of the earth from the sun, as well as thermal infrared energy that is radiated as heat by plants and structures on the ground. As the satellite passes overhead, the instruments detect the wavelength of light coming from a given 30-meter square area on the ground, storing the data as a number. MultiSpec then converts those numbers into an image.
"Vegetation reflects and radiates differently than man-made structures, cultivated areas look different than natural ones, and healthy regions look different than damaged areas," Landgrebe says. "With the current systems, you also can identify different types of crops to some degree. As the instrumentation gets more sophisticated, we should be able to get even more detailed information, such as the health of specific fields."
The current instrument on board the Landsat satellite is called a thematic mapper, which detects seven bands, or wavelengths, of light. The new systems being built, which Landgrebe is working on, are called hyper-spectral sensors because they detect many more bands.
Sources: David Landgrebe, (765) 494-3486; e-mail, firstname.lastname@example.org
Bob Jost, (209) 441-6741;e-mail, email@example.com<
Writer: Amanda Siegfried, (765) 494-4709; e-mail, firstname.lastname@example.org
Purdue News Service: (765) 494-2096; e-mail, email@example.com
NOTE TO JOURNALISTS: A complete list of GLOBE schools and teacher contact information is available from the GLOBE page on the World Wide Web at http://www.globe.gov
To the Purdue News and Photos Page