|
The Choctaw Trail of Tears |
 |
What is Remote Sensing?
|
Remote sensing is the science and art of obtaining information about an object, area, or phenomenon through the analysis of data acquired by a device (a sensor) that is not in contact with the object, area, or phenomenon under investigation (Lillesand and Kiefer, 1994). A photograph of the earth taken from a satellite, for example, is remote sensing. Remote sensing, however, is not limited to just the visible light spectrum.
In order to understand remote sensing, it is important to realize that many different data types are usually collected simultaneously. Similar to collecting visual data in a photograph representing many different wavelengths in the visible spectrum, remotely sensed data may also represent many different wavelengths (bands) both inside and outside the visible spectrum. Thus remote sensing is a technology which allows us to extend our physical senses beyond that which we can see.
A collected data set often contains more information than is necessary and therefore is usually analyzed, filtered, and corrected, and bands are selected or rejected to finally produce an image that shows what the remote sensing scientist wants it to show. While a photograph is a replication of an image as it appears to the human eye, a remotely sensed digital image is an image produced by assigning colors or shades of gray to specific data of interest. The final printed or displayed image appears as the user wants it to appear, showing only the features of interest.
Sensors are commonly mounted on satellites and aircraft. Sensors can also be towed behind boats to sound ocean depths; they can be pulled across the surface of the ground to find underground formations and artifacts; they can even be manipulated by hand to produce an image of a developing baby or to extend the senses of a surgeon. For our purposes, we will limit our focus to spaceborne and airborne sensors that collect data about the earth and its changes.
To illustrate the process of remote sensing, consider that electromagnetic radiation from the sun travels through our atmosphere and strikes the earth. Some of that radiation is absorbed, some is scattered, and some is reflected. The electromagnetic radiation that is reflected travels upward through the atmosphere and strikes a sensor mounted on a satellite. The sensor generates data based on the characteristics of the electromagnetic radiation it detects. This data can then be used in digital form or it can be used to construct an image.
Because different objects on the earth’s surface absorb, scatter, and reflect electromagnetic radiation differently, they appear different to the sensor. Thus, electromagnetic radiation reflecting off a corn field generates a different set of data than that reflecting off a wheat field. The unique reflectance characteristics of an object type, as detected by a sensor, determine the object’s spectral signature.
Field verification is very important in remote sensing. Because many different things on the earth’s surface appear to be the same or almost the same when viewed from a distance, it is important to ground reference the data. If a remote sensing scientist wants to map all the cotton fields in Mississippi for example, he or she must first determine the spectral signature of a known cotton field. Basically, the scientist collects remotely sensed data (using many different bands) for the area and analyzes the data on a computer. The scientist determines what data make cotton fields uniquely different from anything else in the data set. The unique combination of reflected wavelengths and the intensity of each make up the spectral signature for the cotton field. The scientist then tells the computer that anything that exhibits the same spectral signature as the known cotton field is to also be considered a cotton field. He might tell the computer to display all matches for that spectral signature in purple overlaid on a map of Mississippi. The resulting image would show the outline of the state of Mississippi and all the cotton fields in purple.
Remote sensing satellites, which are currently in use, include six NASA Landsat satellites (formerly known as Earth Resources Technology Satellites or ERTS), five French SPOT (Systeme Pour l’Observation de la Terre) satellites, meteorological satellites known as metsats launched by numerous countries, and various commercial satellites.
The first two Landsat satellites launched carried both Return Beam Vidicon (RBV) camera systems and Multispectral Scanners (MSS). RBV systems produce very high-resolution photographic type images, but do not have the digital processing advantages of scanners.
The three major types of scanners currently in use on airborne platforms are: multispectral, thermal, and hyperspectral. In multispectral scanning (MSS), data are acquired simultaneously in four spectral bands, two in the visible spectrum (green and red) and two in the near infrared spectrum. In thermal scanning, thermal infrared energy radiating from objects on the earth is sensed and recorded as data. Hyperspectral scanners acquire multispectral images in many, very narrow, contiguous spectral bands throughout the visible, near-infrared, and mid-infrared portions of the spectrum. These systems collect data in hundreds of different channels, which enables the construction of a nearly continuous reflectance spectrum for every pixel in the scene.
A type of highly refined multispectral scanner that incorporates seven instead of four bands is called the thematic mapper (TM). The TM bands were chosen to improve the spectral differentiability of major earth surface features. The satellites launched most recently include the TM.
There are a number of variables that can interfere with the accuracy of remotely sensed data. The angle of the sun, determined by the time of day that data is collected, is a major factor. This variable can be controlled by always collecting data at the same time of day. In fact, the speed and orbits of remote sensing satellites are set so that data is always collected at the same approximate local time, even though the satellite collects data continuously.
Another variable that is not so easy to resolve is atmospheric interference. Stormy or cloudy weather can make remotely sensed data unusable. Haze caused by pollution can absorb reflected electromagnetic radiation. Suspended particulates and aerosols can alter reflectance as well. Even the wind has an effect. In short, the atmosphere around the world is constantly changing and its variability interferes with the accuracy of remotely sensed data. Dark clouds may unavoidably render data useless, but most other forms of atmospheric interference can be corrected or compensated for if they can be identified and measured.
There are many applications for remote sensing technologies, among which include: mapping, earth observations, inventory/assessment, change detection, precision farming, and resource management. New applications are emerging as the field of remote sensing develops. More and more, decision makers at state, regional, national, and global levels are using remote sensing data to support their work. As more remote sensing data becomes available at less cost to the user, the use of data for long range trend assessments and predictions will take on even more importance.
References
Lillesand, T. M., and R. W. Kiefer, Remote Sensing and Image Interpretation, 3rd ed., Wiley & Sons, New York, 1994.
[Home] [History] [Photos & Video] [Choctaw Culture]
[Mapping the Trail] [RS-GIS Data]
[RS Technology] [For Teachers] [Books] [Downloads] [Links]
[Credits]
