Detecting Landscape Change: The View from Above
Porter, Jess, Journal of College Science Teaching
This article will demonstrate an approach for discovering and assessing local landscape change through the use of remotely sensed images. A brief introduction to remotely sensed imagery is followed by a discussion of relevant ways to introduce this technology into the college science classroom. The Map Detective activity demonstrates the integration of remotely sensed imagery into introductory-level geography courses by illuminating elements of human and physical landscape change via analysis of a series of historical aerial photographs.
University campuses and their host communities are rarely static places. Be it an expanded football stadium, a relocated highway, or a new housing addition, changes to the human landscape abound. Physical landscapes in and around these communities evolve as well, albeit at a usually slower pace. An expanding stream meander undercuts a roadway, an invasive tree species fills an abandoned field, or a hillside incrementally succumbs to gravity. Landscape changes like these are a common area of study for geographers. Discovery and analysis of change processes occurring in the everyday environs of students present an ideal gateway for the study of geography by nonscience majors.
It can be difficult, however, to study or even recognize landscape-scale changes from a ground-level perspective. The view from above, in the form of remotely sensed imagery, provides a perspective that can bring the "big picture" into focus. Remotely sensed images can include a variety of satellite-image types, as well as aerial photographs. These bird's-eye views allow students to identify the forms and patterns of landscape that escape earthbound observers.
Remotely sensed imagery
Remote sensing refers to the acquisition of data about an object from a recording device that is not in intimate contact with that object. Remote sensors target a variety of frequencies of electromagnetic radiation depending on the object being examined. For example, infrared frequencies are often used to ascertain the health of vegetation. This information can be particularly valuable to specialized audiences such as agronomists or plant pathologists. The approach discussed here will focus on less-specialized applications of remote sensing that are applicable to undergraduate geography courses. Remote sensing devices that capture visible light images of the Earth are most relevant to this work.
An additional consideration in viewing remotely sensed imagery is resolution, or the amount of detail an image contains. Higher-resolution images have more detail. Resolution varies based on the type of instrument used to capture the image and the altitude of the instrument above the object. Some types of satellite imagery have resolutions of one square kilometer, indicating that a single pixel of the image covers one square kilometer. Therefore, no individual features less than one square kilometer can be identified from the image. This is not particularly useful for studying community-level features. Fortunately, satellite imagery supplemented by aerial photographs is available at resolutions ranging from 1 to 10 square meters for most cities in the United States. Features such as sidewalks, individual trees, and very small streams are readily discernable in an image with one-square-meter resolution.
Remotely sensed imagery is increasingly available to the public at no cost via the internet. Specific websites are provided in the resources section at the end of this article. Perhaps the best-known source is Google Earth. This virtual globe program has popularized remotely sensed imagery as users get driving directions, find businesses, and explore the world with seamless global imagery. NASA's World Wind provides the broadest range of imagery at any one internet destination. Landsat 7, MODIS, GLOBE, SRTM, and Blue Marble satellite imagery are available along with United States Geological Survey (USGS) topographic maps and a number of additional add-on data packs. …