Comparisons - Understanding Change
he images on this page are excerpts from full satellite “scenes” acquired by the Thematic Mapper (TM) sensor on board the Landsat 5 and Landsat 7 satellites.1 Those satellites are the most recent in chronological sequence that began with the launching of Landsat 1 in 1972. The project, which began under the National Aeronautics and Space Administration and is now administered by the United States Geological Survey, was conceived and implimented for the express purposes of monitoring the health of the Earth and facilitating a better understanding of the complex interactions that drive global change. Satellite images such as these have been used in a wide variety of ways—measuring the ebb and flow of glaciers, tracking population changes in and around metropolitan areas, measuring droughts and floods, the effects of global warming, and many other applications in geology, agriculture, forestry, education, etc.
Landsats 5 and 7 currently (2005) are in geosynchronous2 orbit approximately 435 miles above the Earth. Even at that distance, their sensors can “see” features on the planet as small as 30 meters (98.4 feet) square. Each 30-meter square is represented by one pixel (one seventy-secondth of an inch) of a digital image. The space vehicles, called satellite platforms, are designed to cover the entire Earth in a synchronized path every sixteen days. Landsat 5 is coordinated with Landsat 7 so as to enable the sensors on board to revisit the same area once every eight days, allowing scientists to quantify changes that take place between image dates, and to develop appropriate land management strategies.
Figure 3Lolo Pass from Landsat
aptured in October of 1989, this image and its companion, taken in the same month in 2000, each cover approximately 60 square miles. For clarification, the main route of K'useyneiskit, plus highways and a few landmarks, may be seen in the accompanying popup window .
Consistent with the false-color principle, the dark-red areas are heavily forested; the lighter patterns represent forests that have been logged—the whiter they appear, the more recent they were logged.
Nature, of course, does not suffer straight lines for long, so the surveyors' lines defining the boundaries of square-mile sections of land stand out boldly among the chaotic shapes of mountains and valleys.
The two examples shown above are of an identical area, centering on Lolo Pass and Packer Meadows, taken eleven years apart. The colors are not those we would normally associate with the features depicted. Instead, they represent what is known as a false-color composite, in which colors are defined in terms of the bands3 of data are displayed—in this instance band 3, which is visible red; band 4, near-infrared; and band 5, mid-infrared. The false-color convention is superior to natural colors for purposes of scientific analysis because it is not subject to the atmospheric interferences that we take for granted at eye level. For example, we all know that trees are green, yet we're also thrilled by those "purple mountains' majesty," and we've seen hillsides of trees that look almost black from a certain distance. But by sensing ground cover digitally as a complex set of reflections of electromagnetic waves, and then converting them into arbitrarily assigned colors in the visual spectrum, there can be no doubt as to which waves come from trees.
In the above image, land devoid of any vegetation appears as white, gray, or shades of blue due to the high reflectivity of those surfaces, whereas water, which absorbs almost all radiation, is black. Healthy living vegetation, however, is represented by shades of orange and red. The deeper the red, the greener and more dense the vegetation.
Thus the intact coniferous forests of the Lolo Pass area appear dark red, while the sparse vegetation in the clear-cuts appears as light orange. Those differences in color are really differences in reflectance values. In other words, each pixel in the image, representing 30 square meters, has a number associated with the amount of reflectance in each of the eight bands of radiation sensed by the ETM+. Those numeric values allow scientists not only to study the images by looking at them, but also to perform complex statistical calculations that may reveal information that is invisible to the eye.
n this image the electromagnetic qualities of each pixel are recorded in different numeric values than were applied in the previous images. Scientists can translate those values into information that the human eye interprets as colors. In this vivid composite image, black represents areas in which the ground cover has not changed substantially since 1989. Blue identifies cleared and bare, or partially bare, ground that existed at the time of the earlier imaging. Green areas were cleared of trees sometime between 1989 and 2000.
he simple comparisons shown in these three images are merely the surface of this new realm of scientific investigation. Special software programs are capable of identifying the dominant types of vegetation that existed at the time the first image was captured in 1989 and what remains in 2000, or has been added or removed in the intervening eleven years. Observations made this way, in this place, enable scientists and land managers to assess the health of the ground cover in and around the Lolo Pass area, and to make decisions that will help to ensure the preservation of a site that is important to our historical and ecological heritage.
The on-the-ground, hands-on modes of study and analysis that scientists used at the opening of the 19th century are still necessary. However, we may suppose that Lewis and Clark would be fascinated and inspired by this giant, space-age leap in our ability to understand the world around us today.
--Gary Gooch and Joseph Mussulman. 5/05; reviewed by Marjorie Lubinski and Robert Bergantino
1. Landsat 6, launched in October of 1993, failed to reach orbit.
2. Geosynchronous means that each satellite remains stationary relative to the Earth and the sun throughout the earth's daily orbit, while the earth revolves and tilts beneath it. Therefore the earth always presents a daylight image to Landsat's sensor.
3. Bands are wavelengths of radiation. Landsat 5 is capable of sensing seven different wavelengths; Landsat 7 records eight. The Thematic Mapper is the satellite's sensor that reads bands of solar radiation reflected by the Earth.
Funded in part by a grant from the Idaho Governor's Lewis and Clark Trail Committee.