Remote Sensing Tutorial Overview

Navigation

THE REMOTE SENSING TUTORIAL

PRIME DEVELOPER AND WRITER: DR. NICHOLAS M. SHORT


——————————————————————

Before entering this Overview, ponder this slogan:

REMOTE SENSING is the BACKBONE of the SPACE PROGRAM

Puzzled by these words? The Overview gives a glimpse into their meaning and significance!

——————————————————————


image1

OVERVIEW AND USE OF THIS TUTORIAL


This Overview serves two main purposes: To provide a brief synopsis of the uses and history of Remote Sensing (especially as carried out from orbiting satellites and deep space probes) and to describe the contents of the entire Tutorial with suggestions on best ways to utilize it (whether accessed as a Web Site or from a CD-ROM). Near the beginning on this page, we will expand on the following thesis which is the prime reason for the importance of Remote Sensing and raison d’etre for this Tutorial: **Remote Sensing is the technology that is now the principal modus operandi (tool) by which (as targets or objects of surveillance) the Earth’s surface and atmosphere, the planets, and the entire Universe are being observed, measured, and interpreted from such vantage points as the terrestrial surface, earth-orbit, and outer space*. The main Overview ends with a quick look at the latest products now being acquired by commercial remote sensing satellites. On the second page are biographies and credits appropriate to the contributors. We strongly recommend that you read through the entire Overview as a proper introduction to both the Tutorial and to the many fields of practical activities involving the principal ways in which Remote Sensing and allied fields contribute to gathering information about the many targets of interest to be examined in the pages that follow.*



NOTE 1: The above is a Summary (and Preview) of the topics and content of this Overview. Most of the pages in the Tutorial will have a Summary, bounded by blue lines, near the top.


NOTE 2: The Tutorial has been prepared for online display and for the CD-ROM using the HomeSite html marker program; it is designed to run on the MS IE browser, and the balance between text and illustrations is best at a monitor screen setting of 600 by 800 pixels. (That setting must be changed if the PIT image processing program is installed onto the user’s desktop; see Appendix B.)


NOTE 3:Those of you who are accessing the Tutorial online using modems that operate at 56 bps or slower should be aware that the Tutorial is designed primarily for CD and broadband (DSL, etc.) users (see What’s New); therefore, those with limited download capability may find the very size of the Tutorial daunting.


NOTE 4: There are many internal links in the Tutorial: These are cross-references that go to other pages and are indicated by color highlighted “page #-#” or “Section #”. They are for the most part intended to go to one specific page, on which (somewhere) is the particular image or text referred to in the starting page. To return to the original page, simply click on your browser BACK button.


SPECIAL NOTE: THE PRINCIPAL AUTHOR OF THIS TUTORIAL, DR. NICHOLAS M. SHORT (hereafter, referred to, in most instance, as NMS), IS NOW RETIRED AND IS NO LONGER AT OR NEAR NASA GODDARD SPACE FLIGHT CENTER. HE CONTINUES TO RECEIVE MANY E-MAIL REQUESTS FOR IMAGERY AND INFORMATION ON WHERE TO GET SPECIFIC PRODUCTS OR REFERENCES. IN MOST CASES, HE CANNOT SATISFY SUCH REQUESTS BUT WHENEVER POSSIBLE WILL TRY TO ANSWER CERTAIN TECHNICAL QUESTIONS OR TO SUGGEST OTHERS TO CONTACT. HE CAN HOWEVER PROVIDE A CD-ROM CONTAINING THE LATEST VERSION (AT A COST OF $20); CONTACT HIM AT HIS EMAIL ADDRESS LISTED AT THE END OF THIS OVERVIEW.



Before beginning, please read through the WHAT’S NEW text accessed by the button at the top right of this long page.


WELCOME to this Tutorial, a training manual for learning the role of space science and technology for using remote sensing to monitor planetary bodies and distant stars and galaxies. The Earth itself will be the main focus. The Remote Sensing Tutorial (occasionally cited as RST) initially was sponsored by the Applied Information Science Branch (Code 935) at NASA’s Goddard Space Flight Center, and for a time was underwritten by the Air Force Academy. Currently without direct continuance funding, it is being improved and updated the the prime writer (Nicholas M. Short [NMS]) and John Bolton of NASA Goddard as the proverbial “labor of love”.

As you work through these pages, you will see how we apply remote sensing (a term defined at the beginning of the Introduction Section) to studying the land, sea, air and biotic communities that comprise our planet’s environments, as well as obtaining a deep understanding of the vital role it plays in exploring the planets and reaching the stars and galaxies well out into the Cosmos. Not only will you gain insight into past uses of aerial photography and space imagery, but you should develop skills in interpreting these visual displays and data sets by direct inspection and by computer processing. You will even be able to apply your newly acquired knowledge to actually doing image interpretation using a processing program called PIT on “raw” image data that together come with this CD-ROM or can be downloaded from the Internet version.

The Tutorial has been developed for certain groups as the primary users: Faculty and students at the college level; Science teachers at the High School level; gifted or interested students mainly from the 8-12 grade levels; professionals in many fields where remote sensing comes into play, who need insights into what this technology can do for them; that segment of the educated general public who is curious about or intrigued with the many accomplishments of the space program that have utilized remote sensing from satellites, space stations, and interplanetary probes to monitor and understand surface features and processes on Earth and other bodies in the solar system and beyond. (Most members of these user groups who access this very long Tutorial through the Internet are likely to be on fast-download lines and hence can retrieve individual pages [which can have 15 or more illustrations] rapidly enough for easy and efficient display.)

The central aim, then, of The Remote Sensing Tutorial is to familiarize, and in so doing instruct, you as to what remote sensing is, what its applications are, and what you need to know in order to interpret and, hopefully, use the data/information being acquired by satellite, air, and ground sensors. We try to accomplish this by presenting a very large number of remote sensing products as images which are described in a running text that explains their characteristics and utility. This Internet/CD-ROM means of delivery of the Tutorial is thus image intensive. The abundance of pictorials becomes the principal learning device rather than the more customary dependence on textual description, supported by photographs, found in most pedagogical textbooks. The old adage that “a picture is worth a 1000 words” holds especially true in remote sensing because it can convey, when accompanied by a brief textual commentary, a great deal about how remote sensing is done and the methodology/rationale by which information is gleaned from a pictorial product. And, with the CD-ROM format, the liberal use of color, which often conveys much more information than black and white, becomes feasible - being not subject to the cost limitations that affect presentation in many books.

The Tutorial may well be the first such (Internet; CD-ROM) “book” in remote sensing to contain a significant part of its illustrations acquired directly from downloading off the Net. (Of course, as those familiar with the Internet well know, commonly Net images are digitized at low resolution [typically 72 dpi] and are thus of limited quality; this accounts for the “fuzziness” of many illustrations in the Tutorial.)

Because of its size and the many illustrations, the Tutorial can be treated almost as a textbook. It is hoped that some teachers, especially at the college level, will use the Tutorial either as a bona fide text or as a supplement.

One singular characteristic of the Remote Sensing Tutorial is the inclusion within the continuing text of each Section (not at the end of a chapter as is the case in most textbooks) of a series of thought or interpretive questions. The answers are included on both the CD-ROM and Internet versions. There will normally be 10 to 40+ questions per Section. This Overview has a get-acquainted short Quiz consisting of only a half dozen questions pertaining to a set of images; its purpose is to help you decide whether you want to “get involved” in the learning experience afforded by the remainder of the Tutorial by showing you what image analysis and interpretation is all about and that your general background knowledge is probably sufficient for you to succeed in this process. There are also two “Exams” (after Section 1 and Section 21) and a scene identification Quiz within Section 6 that challenge you to conduct remote sensing interpretations on images from two adjacent areas in central Pennsylvania. Lets introduce you to the type of questions to expect by asking this one right now.

` <>`__O-1: Most people, even those with a good post high school education, when asked what the term “remote sensing” means to them, don’t have the remotest idea. So, what do you think remote sensing is all about? Try to make up a simple definition. Then, list (mentally, or on paper) five practical applications. `ANSWER <the-answers.html#O-1>`__

With this first insight in mind, consider this: Normally, we experience our world from a more or less horizontal viewpoint while living on its surface. But, under these conditions our view is usually limited to areas of a few square miles at most owing to obstructions such as buildings, trees, and topography. The total area encompassed in our vistas is considerably enlarged if we peer downward from, say, a tall building or a mountain top. This increases even more - to perhaps hundreds of square miles - as we gaze outwards from an airliner cruising above 30000 feet. From a vertical or high oblique perspective, our impression of the surface below is notably different than when we scan our surroundings from a point on that surface. We then see the multitude of surface features as they would appear on a thematic map in their appropriate spatial and contextual relationships. This, in a nutshell, is why remote sensing is most often practiced from platforms such as airplanes and spacecraft with onboard sensors that survey and analyze these features over extended areas from above, unencumbered by the immediate proximity of the neighborhood. It is the practical, orderly, and cost-effective way of maintaining and updating information about the world around us.

` <>`__O-2: State an advantage and a disadvantage in conducting a remote sensing viewing from progressively higher altitudes. `ANSWER <the-answers.html#O-2>`__

Remote sensing began on the ground, then moved into the air in the second half of the 19th Century, next on to airplanes in the first part of the 20th Century and by the 1960s entered space as cameras and electronic sensors were mounted on spacecraft to open the era of satellite remote sensing. Various sensors can process radiation not only in the visible but in shorter or longer wavelengths within the electromagnetic spectrum (see below). This chart summarizes the main benefits of satellite remote sensing:

Satellite remote sensing.

Now consider this very important precept or thesis spelled out in bold red letters to accentuate the importance of remote sensing:

Most remote sensing systems (built around cameras, scanners, radiometers, CCD-based detectors, radar, etc.) of various kinds are the most widely used tools (instruments) for acquiring information about Earth, the planets, the stars, and ultimately the whole Cosmos. These normally look at their targets from a distance. One can argue that geophysical instruments operating on the Earth’s surface or in boreholes are also remote sensing devices. And the instruments on the Moon’s and Mars’ surfaces likewise fall broadly into this category. In other words, remote sensing lies at the heart of the majority of unmanned (and as important tasks during some manned) missions flown by NASA and the Russian space agency, as well as programs by other nations (mainly, Canada, France, Germany, Italy, India, China, Japan, and Brazil) to explore space, from our terrestrial surface to the farthest galaxies. NASA and other space agencies *have spent more money* (the principal writer [NMS] estimates this sum to be in excess of a $trillion dollars) on activities that - directly or indirectly - utilize remote sensors as their primary data-gathering instruments than on those other systems operating in space (such as Shuttle/MIR/ISS and communications satellites), in which remote sensing usually plays only a subordinate role. Add to this the idea that ground-based telescopes, photo cameras, and our eyes used in everyday life are also remote sensors, then one can rightly conclude that remote sensing is a dominant component of the scientific and technical aspects of human activity - a subtle realization since most of us do not use the term “*remote sensing*” in our normal vocabulary.

Having made this “sales pitch” let us turn the now convinced to a brief look at the history of Remote Sensing (covered in further detail on page I-7). Close-up photography (Promimal Remote Sensing) began in 1839 with the primitive but amazing images by by the Frenchmen Daguerre and Neipce. Distal Remote Sensing from above ground began in the 1860s as balloonists took pictures of the Earth’s surface using the newly invented photo-camera. Most photos were made from tethered balloons but later free-flying balloons provided the platform. The earliest balloon photo was made of Paris in 1858 by Honore Daumier but this historic first has been lost. The photo below from a balloon anchored in Boston, made in 1860, is the first aerial photo surviving in the U.S. Balloons were used for reconnaissance during the Civil War; legend has it that General McClelland had a battlefield photo made from such an aerial post but it has disappeared.

Boston from a tethered balloon; photograph by James Wallace Black.

NOTE: Each image throughout this Tutorial will have a caption that is accessed simply by placing your mouse on the lower right portion of the image.

Perhaps the most novel platform at the beginning of the 20th century was the famed Bavarian pigeon fleet that operated in Europe. Pigeons at the ready are shown here, with a famed 1903 picture taken of a Bavarian castle beneath (the irregular objects on either side are the flapping wings.

|Pigeon fleet used in the early 9800s in Europe to carry cameras above the terrain that were timed to automatically expose a series of film shots |

Pigeon with camera in flight.

Photograph of a castle taken automatically by a camera strapped on a pigeon in flight

` <>`__O-3: What is an obvious disadvantage in using this primitive pigeon system? `ANSWER <the-answers.html#O-3>`__

Aerial photography became a valuable reconnaissance tool during the First World War and came fully into its own during the Second World War.

The possibility of conducting “aerial” photography from space hinges on the ability to use rockets to launch the equipment, either up some distance to then fall back to Earth or into Earth orbit. Page 7 in the Introduction describes the earliest successes. Rocketry can be traced back to ancient times when the Chinese used solid materials, similar to their firecracker powders, to provide the thrust. In the 19th Century, the famed French science fiction writer, Jules Verne, conceived of launching a manned projectile to the Moon (in his book “From the Earth to the Moon”, which formed the inspiration for this writer’s [NMS] first presented science paper on rocketry to his high school Science Club). In the first half of the 20th Century, a leader in rocketry was Robert Goddard (1889-1945) after whom Goddard Space Flight Center (where this Tutorial is based) was named. Below is a 1926 photo of Dr. Goddard with one of his first liquid fuel rockets (the motor is on the top of this 10 foot vehicle [it would break free from the frame holding it up]).

Robert Goddard aside one of his successful liquid fuel rockets.

The logical entry of remote sensors into space on a routine basis began with automated photo-camera systems mounted on captured German V-2 rockets, launched out of White Sands, NM. These rockets also carried geophysical instruments in their nose cones, which were returned to Earth by parachute. (The writer [NMS] during his Army service at Fort Bliss, El Paso, TX in 1946-47 was doubly privileged. First he was part of a group of GIs assigned to search for a missing instrument package in its nose cone. Then, in Spring 1947, as a Post newspaper reporter, he interviewed Dr. Wernher von Braun - the guru of post WWII rocketry - and was present during a V-2 launch. Little did I realize then that Space would become my career.) Below is an example of one of the first photo pictures returned from a V-2 firing, along with a list of specific localities recognizable in this view covering 800000 square miles of the western U.S. and showing the Earth’s curvature:

V-2 photo of the western U. looking west from White Sands, NM.

Key to Major land features, corresponding to numbers in the above picture.

The modern Space program is held by many historians to truly have begun with the launch of Sputnik I by the Soviets on October 4, 1957 (like many noted events that stick in one’s memory, the writer recalls vividly exactly where he was as the news was read over a radio: eating breakfast in a cafeteria in Casper, Wyoming at the start of a day of geological field work). Here is a full scale model of the first Sputnik (about the size of a basketball, weighing 83 kg [182 lb]), with radio and one scientific instrument), on display at the National Air and Space Museum in Washington, D.C.:

Model of Sputnik I at the NASM.

This tiny satellite was hurled into space by the Semiorka rocket, seen below. The Soviet program was led by Sergei Korolev. An interesting perspective on the world-stunning effects of this pioneering launch can be read at this Web site.

The Semiorka rocket enroute to the launch site of the first Sputnik.

Several larger Sputniks soon followed, each with scientific payloads. Sputnik II carried a live dog, Troika, instrumented to determine its reaction to being in space. The U.S. launched its first orbiting satellite, Explorer 1 in January, 1958, followed shortly by the Vanguard series (see page Intro-1a) for more details. Much more about the history of Man in Space is reviewed in Appendix 2.

After the launch of Sputnik in 1957, putting film cameras on orbiting spacecraft became possible. The first cosmonauts and astronauts used hand-held cameras to document selected regions and targets of opportunity as they orbited the globe. Sensors tuned to obtain black and white TV-like images of Earth flew on meteorological satellites in the 1960s. Other sensors on those satellites made soundings or measurements of atmospheric properties at various heights.

` <>`__O-4: On TV, you are most likely to encounter a satellite remote sensing product of what kind (hint: think local news)? `ANSWER <the-answers.html#O-4>`__

As an operational system for collecting information about Earth on a repetitive schedule, remote sensing matured in the 1970s, when instruments flew on Skylab (and later, the Space Shuttle) and on Landsat (early on, called ERTS), the first satellite dedicated to mapping natural and cultural resources on land and ocean surfaces. A radar imaging system was the main sensor on Seasat, launched in June, 1978. In the 1980s, a variety of specialized sensors - Coastal Zone Color Scanner (CZCS), Heat Capacity Mapping Mission (HCMM), and Advanced Very High Resolution Radiometer (AVHRR) among others - orbited primarily as research or feasibility programs. The first non-military radar system was JPL’s Shuttle Imaging Radar (SIR-A) on the Space Shuttle in 1982. Other nations soon followed with remote sensors that provided similar or distinctly different capabilities. By the 1980s, Landsat had been privatized and a widespread commercial use of remote sensing had taken root in the U.S., France, Russia, Japan and other nations. Much of this growth was, and is still being, driven by the increasing awareness that Earth’s environments are in peril from man’s activities and misuses.

` <>`__O-5: Where might you have seen a Landsat image before? `ANSWER <the-answers.html#O-5>`__

It is generally agreed that Landsat set the stage for the advent of these other satellite systems in that it demonstrated the power and versatility of multispectral imagery for observing the Earth for purposes of monitoring its natural and manmade features over time, from which the many applications of remote sensing have now become important in managing our planet’s “health” and the utilization of its resources. Since 1972, six Landsats have been orbited successfully. Here is the history of this highly successful program:

History of the Landsat program.

So, how is remote sensing actually done from such satellites as Landsat, or for that matter, from airplanes or on the ground? Remote sensing uses instruments that house sensors to view the spectral and spatial relations of observable objects and materials at a distance, typically from above them, or in astronomy, by looking out. Geophysics (mainly gravity, magnetic, and seismic surveys; also external fields) is considered by many to be a form of remote sensing. But, except for three pages in the Introduction that summarize doing geophysics measurements from space, we will confine our study in this Tutorial mainly to methods and applications of spaceborne sensors that produce images and thematic maps. Most methods are based on sensing of photons (quantum particles that have a wide range of energies; a specific photon will have some energy value that has its own unique corresponding frequency [number of cycles of a sine waveform per unit time]) in the electromagnetic (EM) spectrum.. Here is a simple EM Spectrum Chart, with different wavelength intervals named according to common usage in remote sensing (the wavelength units are in micrometers (µm); a micrometer is 1/1,000,000 of a meter.

The Electromagnetic Spectrum

This term EM Spectrum refers to the distribution of radiant energy as a function of wavelengths (distance in metric units between successive wave crests in an oscillating sine wave, which for radiation is the trace of a forward moving photon as it revolves 360° through one cycle) or their inverse, frequencies (number of cycles per second) presented usually as a chart or diagram with highest frequencies (shortest wavelengths) at one end and lowest frequencies (longest wavelengths) at the other. Radiation may be continuous (no break in the range of wavelengths), its plot consisting of a sequence of all wavelengths over a spectral range whose low and frequencies are at some beginning and end values. It can also be discrete, i.e., photon energies are associated with specific, generally narrow wavelength intervals, with radiation outside these intervals being absent (these discontinuous intervals are representative of energies released when atomic or molecular species are excited in specific ways [determined by quantum physics]). Thus, chemical elements, when excited by thermal or electrical energy, give off EM radiation at discrete (particular) wavelength values unique to each element species; these may appear as lines in a spectrogram made by dispersing the radiation using a prism or diffraction grating..

One type of a continuous spectrum is the blackbody radiation (BBR) emitted by all bodies whose temperature is above absolute zero. A given BBR spectral plot, characterized by a total spectral interval fixed on end points of specific wavelengths, is determined by the thermal state of the object sensed and varies in “area under the curve” (the plot) and peak intensity, both determined by the surface temperature of the body. BBR curves for three stars of differing surface temperatures illustrate this type of radiation; note that as temperatures increase the radiation intensity also increases and the peak wavelength decreases.

Planck Black Body Radiation curves.

To synposize these last ideas about electromagnetic radiation, consider this diagram:

Three modes of spectral radiation.

Photons are emitted from a hot source (the Sun, an electric light, etc). The spectral curve for this condition or mode is like the above BBR curves. Now this light passes through a target, in this instance a cloud containing atoms and molecules. On the right is an absorption spectrum in which the black lines are at wavelengths characteristic of elements or molecules that absorb some of the photons of specific energies (proxied by their characteristic wavelengths). At the same time, some of these photons cause atoms and molecules in the cloud to be excited such that they give off (emit) radiation at particular wavelengths, as shown in the bottom spectrum.

In actual practice in remote sensing, classes of features (leaves, soil, rock, buildings, etc), upon excitation reflect or emit radiation that produces plots of photon energy variations as a function of wavelength or frequency that comprise characteristic spectral signatures (curves) (see below). The ideas in this paragraph are treated in detail in the Introduction that follows this Overview.

Most remote sensing data consists of receiving and measuring reflected and/or emitted radiation from different parts of the electromagnetic spectrum. Those parts of the spectrum most commonly sampled are the ultraviolet, visible, reflected infrared, thermal infrared, and microwave segments. Multispectral (or the closely related multiband data consist of radiation collected over sets of electromagnetic radiation that individually extend over (usually narrow) intervals of continuous wavelengths within some part of the spectrum. Each interval makes up a band or channel identified by a color (if in the visible), a descriptive label (e.g., Near IR), or a specified range of wavelengths. The data are utilized by computer-based processing to produce images of scenes (Earth’s surface and atmosphere; planets; cosmological features) or to serve as digital inputs to analytical programs (see Section 1 for a thorough examination of imaging techniques and categories of analysis).

Multiband images collected by one sensor will usually show notable differences from one band to the next. This is because the radiation from point to point in an array of sampling areas making up a scene will vary depending on the reflectance or emittance response of the various features/materials are different within an interval, and different again when other bands are examined. The band to band response (in terms of magnitude or intensity of radiation) of any such point can be connected to become the spectral signature for a given feature or class of materials. Different features/classes have differing and normally distinctive signatures.

Much of the ideas given in the preceding three paragraphs can be summarized in this diagram:

Sensing a spectral signature.

In this case, the target is a field of actively growing crops - the main components are thus vegetation, soil, and moisture. The detailed spectral signature for this composite of materials is shown in the lower right. Some fraction of the incoming solar radiation is reflected towards a sensor above (on an aircraft or spacecraft). While it is now possible for a sensor system to almost duplicate the signature using the mode called hyperspectral remote sensing, in this example the broadband mode, initially the normal configuration for obtaining reflectance measurements and still in common use, is illustrated here. Thus the sensor employs bandpass filters to break the reflected radiation into discrete intervals (bands)of continuous wavelengths, each consisting of a segment of the EM spectrum (red, green, infrared, etc.). The radiation consists of photons that impign upon a plate that converts the photon energy to a voltage (photoelectric effect). At the instant of sampling this radiation, each band will have some voltage value (indicated on the dials). Assuming proper calibration of each band (channel), this voltage is a measure of the reflectance from the target composited for each spectral interval. The resulting values represent a crude approximation of the spectral signature. However, even these few values may be sufficiently distinct to establish the identity of the target. Obviously, the more bands (and narrower bandwidths), the better is the discrimination.

To whet your appetite for remote sensing and to familiarize you with some of the principal types of image products that are used to monitor and document the Earth’s surface, we will now present an example of multispectral images and then a sequence of space images of an area of the United States that occupied centerstage during February of 2002:

We will illustrate these ideas by showing images representing 4 of the 7 bands acquired by the Thematic Mapper (TM), the main sensor on Landsats 4 through 7. Each image was constructed from numerical values called Digital Numbers (DNs) which correlate with the intensity of reflected or emitted radiation averaged for the spectral interval (Band) displayed; the DNs in this case range from 0 to 255 in whole number increments. Levels of gray in the resulting image range from black (DN = 0) to white (DN = 255) with shades of dark gray to very light gray associated with increasing DN values. The scene, a subset of a full Landsat TM image, shows the western shore of the Keweenaw Peninsula of northern Michigan (for this and other related images, link onto the Michigan Technological University Web site). Wavelength intervals (in micrometers) are shown; check the captions (cursor on lower right) for more information.

|TM band 1 covering the blue visible region of the spectrum |

Band 4 extending into the Near infrared just beyond visible red in the spectrum.

The thermal band 6, light tones are warmer than the darker cooler areas; resolution is 120 meters, causing the 'fuzzier' appearance of the scene.

Band 7, covering part of the mid-infrared interval of the spectrum.

For bands 1, 4, and 7, the darker (gray scale) tones in these black and white renditions represent low (intensity) reflectances whereas light tones are high reflectances. In band 6 what is measured is emitted radiation which becomes more intense (leading to lighter to white tones) with higher temperatures. Starting with Band 1, pick out certain features (a pattern of usually uniform gray tones), without concern about their identities, and find the gray tones at equivalent points in the other three images - this will give you a feel for how reflectances (or emittances in Band 6) vary as a function of wavelengths used to monitor features/classes.

Combinations of any 3 of the 7 bands on TM can be registered spatially and then each assigned to one of the three primary colors: blue, green, red to yield what is called a color composite. This can be done photographically using color filters or in a computer display in which the colors are determined by the assignment (using an image processing program) of a given band to one of three color guns in the monitor (and the remaining two bands each to the remaining colors). For the TM, the most frequently used combination is Band 2 = blue; Band 3 = blue; Band 4 = red, giving the standard false color version in which most of the reds and off-reds are the color signatures of vegetation. This is present here as a larger subset showing nearly all of the Keweenaw Peninsula:

Band 4(red), 3(green), and 2(blue) TM false color composite of the Keweenaw Peninsula

Below is another combination applied to the smaller using 3-red, 2-green, and 1-blue. In this image, for the “fun of it” locate where this subset is in the image above and try to identify (give them names, like water, town) features you recognize.

Band 3,2,1 = R, G, B color composite of the Keweenaw subscene imaged by TM; the large town is Calumet.

This brief primer on the appearance of individual multispectral bands and on making color composites from combinations of three bands (or other variables) from one (or perhaps two or more) sensors designed to scan the target (Earth’s surface; a galaxy, etc.) as an array of spatially distinct points (in many cases, the sampled point can be called a pixel (picture element) should help you to appreciate some of the images from various sensors and sources to follow in this Overview. Again, more details on this subject are given in the Introduction and in Section 1.

The Internet is a prime source for information on almost every aspect of remote sensing. Many sites offer good overviews of satellite remote sensing. For a general listing of these sites, consult Remote Sensing Tutorials and Training Courses. One that has recently appeared, and provides an excellent synopsis of the main principles and applications, has been constructed by the Canadian Center for Remote Sensing. Click here if you want to view it now, or at your leisure. Another of merit is the Remote Sensing Core Curriculum, project which highlights a new educational approach now under development. A thorough treatment of the basics of remote sensing, prepared by the Japanese Association of Remote Sensing, is online at this mirror site. A somewhat briefer review has been prepared by Harrison and Jupp. A recent NASA-supported initiative in curriculum development is described at the Geospatial Information Technology website of the University of Mississippi. For a broad perspective on how remote sensing has flourished in the last 30 years, one needs only to check out the still growing number of U.S. and International organizations - government, university, and private - that are largely concerned with various facets of remote sensing. A listing of most of these is found at The Remote Sensing Organizations site. Another source of information on various remote sensing tutorials and related topics is located as a link on the Home Page of the Remote Sensing Tutorial; for those accessing the RST through the CD, we provide this link here as the Carstad site. Lastly, for those who might wish to build or expand their knowledge and background in several sciences that are relevant to remote sensing, we strongly recommend exploring the PSIGate site maintained by the University of Manchester (England) that has many useful links in Astronomy, Earth Science, and Physics.

Having surveyed some basic principles and examples of remote sensing and its products, we now move on to the aforementioned sequence of various types of imagery that relate to a major event in Utah during February, 2002. We turn to The Salt Lake City, Utah region, site of the 2002 Winter Olympics. This should help you appreciate the advantages of both multiplatform, multisensor, and multitemporal data and imagery.

To set the Salt Lake City area into a large, i.e., regional context, look first at this Daytime Thermal image made by the Heat Capacity Mapping Mission (HCMM):

HCMM image of part of the western interior of the U.S. showing the Great Salt Lake as a focal point.

The Great Salt Lake is the dark, elongate feature in the upper right quadrant. It is dark because, thermally, it is cool and in conventional thermal images cold features tend to be dark gray to black and warm in light gray to white. The mountain chains show up moderately dark because they are cooler - at higher altitudes - than lowlands and basins. Next to the Great Salt Lake just to its lower right is Salt lake City. The dark vertical area to SLC’s right is the Wasatch Range - home of many Olympic events. The east-west chain of mountains to its east is the Uinta Mountains. Many elongate dark features, running mostly up-down in the image, are individual mountains that make up the Basin and Range tectonic and geomorphic provinces.

But, at the outset it is instructive for comparative reasons to look at the most common type of earth-surface image available prior to the Space Age: a black and white aerial photo. Here is a 1:62500 scale (see Section 10) photo of part of North Salt Lake City:

Part of Salt Lake City shown in an aerial photo.

To put the “venue” of this great sports event into context with its surroundings, at a regional scale, we’ll start with one of the typical aerial oblique photos taken by the astronauts on a Space shuttle mission; read the caption (click on picture) for a general description.

Space Shuttle photo looking northwest and showing the Great Salt Lake, the surrounding deserts and block fault mountains, the Wasatch Range in the foreground; Salt Lake City is not well defined in this image, but when you become more familiar with its location later on this page, come back to pinpoint it.

Now, we introduce you to a characteristic unmanned satellite image: Shown first are the four individual band images made by the Landsat 1 Multispectral Scanner (MSS; 79 m resolution) presenting a view of north-central Utah taken just 15 days after launch of ERTS-1 (the name given this satellite before it and its successors were renamed the Landsat series), on August 7, 1972. The bands are identified in the caption (note: these are somewhat degraded in quality because they were scanned from a 35 mm slide; the two IR band images are also not well-balanced in gray tones owing to rather poor tonal stretching as those in the image-processing lab were still learning how to generate good quality photo prints).

The first ERTS-1 multiband images of the Salt Lake City area of Utah; the individual bands are: Upper Left - MSS Band 4 (green); Lower Left - Band 5 (red); Upper Right - Band 6 (Near IR); Lower Right - Band 7 - IR.

The scene below is a an early (Summer of 1972) Landsat (ERTS-1) false color image (185 km [110 miles] on a side) that helped to generate widespread interest in using satellites to monitor the Earth’s surface. Images of this type are made by sensors that receive reflective light which is split into several Bands (made by subdividing both the visible and the near infrared spectrum into narrower wavelength intervals), each with different tonal intensities (gray levels) in its image. Three of the band images are then recombined (registered) photographically (or by a computer program) using red, green, and blue filters (this idea, treated very briefly here but in detail in the Introduction. The combination of bands and filters can vary, giving rise to color composites that differ in colors associated with different features depending on the band/filter pairing.

|A Landsat-1 (ERTS-1) false color image showing much of north-central Utah, and Great Salt Lake being the dominant feature; Salt Lake City is a bluish area near the center and south-east of the lake; the Wasatch Mountains occupy the eastern third of the image and appear reddish because that is the color |

The version shown here is a composite made by projecting the MSS Band 4 (green wavelength interval) through a blue filter, 5 through green, and 7 (IR) through red. The right side of the image is bright red, which is the normal color for thick forests and grasslands as rendered in a standard false color image in which we associate red with healthy vegetation that is usually very bright (high reflectance, appearing in light tones) in the near-infrared (see page I-13in the Introduction Section for the explanation of color response and assignment). This widespread red area coincides with the high Wasatch Mountains that run east of the block-fault mountains and deserts (gray-tan tones) of western Utah. Other reds in small patches mark the farmlands of the desert plains whose potential inspired Brigham Young to settle his group in this “promised land”. The Great Salt Lake occupies part of the upper scene. Lake Utah (bluer because of silt) is to its south. We challenge you to find the metropolitan area of Salt Lake City in this image.

` <>`__O-6: This is a good moment to begin to associate locations and features within a space image such as Landsat with their counterparts on a map. Using a U.S. Atlas or a state map, fit the Landsat image to its equivalent map area. In addition to places mentioned above, also find these features: The small cities of Ogden, Orem, and Provo; Park City; Utah Lake; the Bingham Open Pit Copper mine in the Oquirhh Mountains; large areas devoted to agriculture; heavily forested lands; desert flats. Also, in your atlas, if it is nearly new, the shape of the Great Salt Lake may differ from that in the image; why? Finally, why is the central part of Salt Lake City (which appears as a long darker blue strip) so narrow, when the greater area of the city and suburbs seems to appear reddish? `ANSWER <the-answers.html#O-6>`__

Below this image, we place a subscene (part of the total area covered; the image was made using a subset of data points sampled by the MSS) image of the same area made from a Landsat-7 image acquired in the late 1990s. For the moment, just look it over and try to note any conspicuous differences between it and the corresponding area in the Landsat-1 image. We will take this comparison up again a few paragraphs later.

Landsat-7 subset image enlarged from a full scene.

If you look intently at the Landsat-1 image, you may see a slight tonal difference along a straight, sharp boundary; this is due to a cutoff of water circulation by the Union Pacific railroad causeway. This is much easier to see in this next near-true color image made by the MODIS sensor onboard Terra (see Section 16). The tonal discontinuity is almost “invisible” in the Landsat-1 image because the multispectral sensor (MSS) onboard does not have a blue band that would have picked out the presence of silt (which increases the reflectivity of the water, producing a ligher tone in the blue region of the visible spectrum) in the Upper Great Salt Lake.

MODIS subscene in approximately true color, showing the sharp divide within the Great Salt Lake; the upper part contains much more silt in contrast to the clearer water in the lower part.

Notice in the Landsat-7 image shown above that this straight border is now altered to a bend in which the two segments appear to meet at a high angle. The left (western) segment is still a straight line but the right (eastern) segment has a blurred boundary. The area to its north was mostly saline silt deposits in the 1972 image but water has since spilled into the land above the train causeway. Close to the tracks, the water to the north is relatively clear but the blue silt tones start to show up a short distance further up.

In case you had difficulty in pinpointing the city, this next view should help. It is a Landsat-5 Thematic Mapper (TM; 30 m resolution) natural color image of the immediate urban area. It also demonstrates the improvement in detail that has transpired in the later Landsats owing to this new sensor.

Landsat-5 TM subscene of Salt Lake City.

These images, of course, are vertical (straight down) views. To acquaint you with looking at Earth this way, we draw upon a more familiar viewing vantage by showing this near-horizontal aerial view of the city and the Wasatch Front to its east. (The famed Mormon Tabernacle is the multi-spired building near the center).

Near-horizontal view of Salt Lake City and the Wasatch Front mountains east of the city.

Compare the downtown area shown above in a photo taken in the mid-1960s with a photo taken in 2001. Try to determine which major buildings (some are high-rise) have been added to the skyline since the ’60s.

Color photo of central Salt Lake City, taken in 2001.

As will be repeatedly demonstrated throughout the Tutorial, space imagery can be combined digitally through specialized computer processing which uses a digital elevation data set to produce what is known as a perspective view (as though you were approaching the scene in a low flying aircraft and looked ahead; much like the above aerial photo). Here is a Landsat-5 perspective of the Wasatch Front with much of Salt Lake City in the foreground.

Landsat Perspective View of Salt Lake City.

Another Landsat perspective view from a different direction shows the location of the principal Olympics venue sites both in Salt Lake City and the mountains to its east:

Perspective view of the Salt Lake City area and the main Olympic sports sites.

To get a more intimate feel for the downtown part of Salt Lake City, here is two high resolution images made by the IKONOS satellite (see below). The first, in color, shows much of the downtown (at 4 m resolution), including part of the University of Utah. The second depicts, at 1 m resolution how city blocks in this town tend to be square; the two large buildings in it can be located near the left center edge of the first image.

IKONOS 4 m color image of Downtown Salt Lake City.

Downtown Salt Lake City, imaged by IKONOS at 1 meter.

We can zero in on the Olympics infrastructure that has been home to more than 2500 international athletes. Again, two high resolution IKONOS color images, one taken in the summer of 2001 and below it part of the same area taken during the Winter Games in February, 2002.

Part of Salt Lake City that includes Olympic Games facilities, and the Olympic Village.

Salt Lake City; Winter 2002.

The part of the Olympics that includes downslope sking lies in the Wasatch Range near Park City, just east of Salt Lake City. Here is an IKONOS view which shows the ski trails, and housing to the west:

The Ski area near Park City, part of the 2002 Winter Olympics.

Leaning on your new found familiarity with Salt Lake City, try to find some of the features shown in the above images in this very different-appearing image. This scene was obtained during the SIR-C radar mission carried out by astronauts in 1994. Each of the three radar bands (C, L, X) were assigned color used to generate this “false color” composite (see page 8-7). The image is oriented with the top boundary running NE-SW; the Great Salt Lake is the black area below the top.

SIR-C radar image of Salt Lake City.

So far in our excursion in and around Salt Lake City, we have treated you to what might be called “pretty pictures”. But, now is a good time to stress the practical use or applications of space imagery. One such use comes under the term “change detection” - determining what features or conditions in a scene have been introduced, modified, or expanded over short to long time periods. Scroll upwards now to the two Landsat scenes (full Ls-1 and subset Ls-7). There are at least three major features or categories that are different (changed) in the lower Ls-7 scene representing a time span of 27 years. Do this before advancing to the next four scenes.

Urban population is one change that you would expect of this lengthy time period. The United States has increased its citizens considerably since 1972. The West, in particular, is experiencing a population boom, both from increased childbirth and from the influx of people from the eastern U.S as well as Mexicans who have emigrated from their native country. Salt Lake City shares this trend, as is evident from this pair of Landsat images. To estimate the extent of the growth, look for street patterns in each image - the major clue is the spread of buildings as the suburbs expand away from the mountains.

1972 Landsat-1 subset image of the Salt Lake City area.

2001 Landsat-7 subset image of the Salt Lake City area.“>

The most noticeable area of growth occurs in the middle of these images (the 2001 image shows urban/suburban sections of the city in a grayish tone; this is probably due to that image being taken at a different time of the year). Note the large, irregular “scar” in light brown in the lower left quadrant of each image. This is the Bingham Canyon copper mine, located in the Oquirhh Mountains. This is the largest open pit mine in the world (note the increase in peripheral size in the 2001 image). You will see this mine again in an enlarged image subset at the bottom of Page 5-4.

The second pair of Change Detection images focuses on the southern end of the Great Salt Lake. Significant differences between the 1972 and 2001 Landsat images occur at several places, In the 1972 scene, the peninsula of land near the bottom center is tied to the shore with all land exposed. By 2001, that peninsula has been cut off from the mainland. This is the result

We return for the moment to what has happened in and around the Great Salt Lake since 1972. Two Landsat images, the top taken in 1972 and the bottom in 2001, will allow easy comparison that facilitates picking out the changes in 29 years.

Subscene from a 1972 Landsat-1 MSS color composite, showing the land and water at the southern edge of the Great Salt Lake.

|Same as above, except that here the spacecraft was Landsat-7, date of image is Summer 2001. |

The most obvious modification noted in the 2001 image is that the peninsula at the southern end of the Lake has become isolated (into Antelope Island) owing to the lake surface level’s rise since 1972. At first, this seems counterintuitive since the ultimate fate of lakes is for them to dry up (some as rapidly as a few thousand to 20000 years). But this is not a uni-directional process. Changes in climate from dry to wet and reverse can have measurable effects over spans of decades. In 1963, the Great Salt lake had shrunk from the hundred year average of 4200 square miles to a value of ~950 square miles. This shrinkage was the consequence of a continuing drought that began in the 1950s. By 1972, the area covered by the lake had extended to about 2500 miles2. At that time there was still a land bridge to Antelope Island. By 2001, that bridge was inundated, restoring the peninsula to island status. Elsewhere in the subscenes being compared, a tongue of sandy land on the southeast corner of the lake was resubmerged by (actually before) 2001 and the lowlands adjacent to a mountain outlier in the southwest corner have become partially covered with shallow water that supports vegetation.

Now, let’s leave the specificity of a single scene, which we have used to introduce you to some of ways in which satellite imagery can depict the Earth’s surface, and return to the more general overview of what Remote sensing is all about and can do in practical ways. In addition to regional and local scale coverage, sensing from satellites allows images to be created that can envisage the full Earth or entire continents. Here, for example, is the quasi-natural color view of the 48 continental U.S landmass (Courtesy Earthsat Corp, Rockville, MD) made from summer AVHRR (page 14-2) imagery. Notice the regionally variable distribution of vegetative cover (green).

The continental United States in natural color, constructed as a mosaic from AVHRR imagery.

This Tutorial draws principally on the Landsat satellites for the images of the Earth’s surface you will see, in part because there are so many outstanding scenes acquired since 1972 but also in part because the writer (NMS) spent most of his career at NASA Goddard working on data from these satellites. The RST also utilizes imagery from a variety of sensors operating from land and sea satellites launched by U.S. government and private U.S. industry and by governments and commercial firms in other countries. Most of these observe in the visible, near infrared and thermal infrared spectral intervals, but images from several radar systems are also included as examples of common space data sets.

Listed here are the principal (non-commercial) remote sensing spacecraft flown by the U.S. and other nations (identified in parentheses) along with the launch date (if more than one in a series, this date refers to the first one put successfully into orbit. These fall naturally into three Groups based on their principal applications: Land, Meteorology, Oceanography. However, many of the satellites provide useful information for more than one Group:

Group 1 - Primarily Land Observers: Landsat (1-7) (1973); Seasat (1978); HCMM (1978); RESURS (Russia) (1985); IRS(1A-1D) (India) (1986); ERS (1-2) (1991); JERS (1-2) (Japan) (1992); Radarsat (Canada) (1995); ADEOS (Japan) (1996); Terra (1999); Proba/Chris (2001)


(Note 1: SIR-A (1981), SIR-B (1984), and SIR-C (1994) are radar systems flown on Space Shuttles; a Laser Altimeter also flew on Shuttle)

Group 2 - Primarily Meteorological Observers: TIROS (1-9) (1960); Nimbus (1-7) (1964); ESSA (1-9) (1966); ATS(g) (1-3) (1966); DMSP series I (1966); the Russian Kosmos (1968) and Meteor series (1969); ITOS series (1970); SMS(g) (1975); GOES(g) series (1975); NOAA (1-5) (1976); DMSP series 2 (1976); GMS (Himawari)(g) series (Japan) (1977); Meteosat(g) series (Europe) (1978); TIROS-N series (1978); Bhaskara(g) (India) (1979); NOAA (6-14) (1982); Insat (1983); ERBS (1984); MOS (Japan) (1987); UARS (1991); TRMM (U.S./Japan) (1997); Envisat (European Space Agency) (2002); Aqua (2002)


(Note 2: g = geostationary) (Note 3: Nimbus also observed general land features; e.g., Nimbus 6 carried SCMR, an experimental sensor designed to obtain information on surface composition)

Group 3 - Major use in Oceanography: Seasat (1978); Nimbus 7 (1978) included the CZCS, the Coastal Zone Color Scanner that measures chlorophyll concentration in seawater; Topex-Poseidon (1992); SeaWiFS (1997)


(Note 4: NSCAT, the NASA Scatterometer, developed at JPL and launched in 1996 by a Japanese rocket, was designed mainly for oceanographic studies but has provided valuable information applicable to meteorology and land observations.)

Commercial Satellites designed to produce imagery useful to the above Groups started to operate by the mid 1980s. Among the growing number of these privately owned satellites are: SPOT (France) (1986); Resurs-01 series (Russia) (1989; became commercial in the 1990s); Orbview-2 (U.S.)(1997) SPIN-2 (Russia)(1998); IKONOS (U.S) (1999); Quickbird (U.S) (2001); Resource21 (first 4 satellites yet to be launched); EROS A (ImageSat International; Israel) (2000).

A very good review of most of the major satellites dedicated to earth observations and their characteristics, with links (some of which no longer work [404 Not Found]) to parent Web sites, can be called up from these two sites:National Air and Space Museum and University of Wisconsin.

Another site that emphasizes remote sensing and imagery is the Eduspace program sponsored by the European Space Agency (ESA). The site can be accessed by clicking on Eduspace links. Be advised that to get into some of the features at this site, you must be able to register as a member of a teaching institution - primary through college.

Another Website dealing with most aspects of remote sensing is The WWW Virtual Library of Remote Sensing, out of Finland. It has an abundance and variety of links, many of which are worth exploring at some stage in your use of this Tutorial. However, it is not maintained for currency, so that some enticing titles are no longer active.

It helps to picture the dazzling array of operational satellites by looking graphically at launch dates and lifetime of some of those in the above list (primarily land observing satellites) through the year 1996; others since 1997 are listed on a bar chart found on the second page of the Overview.:

A bar chart history of land observing satellites between 1972 and 1996.

This impressive list convinces us that remote sensing has become a major technological and scientific tool for monitoring planetary surfaces and atmospheres. In fact, the budgetary expenditures on observing Earth and other planets, since the the space program began, now exceed $150 billion. Much of this money has been directed towards practical applications, largely focused on environmental and natural resource management. The Table below, put together in 1981 by the writer, summarizes the principal uses in six disciplines.

Table summarizing the principle uses of remote sensing in various professional disciplines.

All of these applications are valid today, and many others have been devised and tested, some of which we introduce in other Sections of this Tutorial. The literature on remote sensing theory, instrumentation, and applications is now vast, including a number of journals and reports of numerous conferences and meetings. The great improvements in computer-based image processing, especially personal computers that handle large amounts of remote sensing data, have made robotic and manned platform observations accessible to universities, resource-responsible agencies, small environmental companies, and even individuals. Geographic Information Systems (GIS) provide an exceptional means for integrating timely remote sensing data with other spatial types of data. The GIS approach (explained in Section 15) stores, integrates, and analyzes information that has a practical value in many fields concerned with decision-making in resource management, environmental control, and site development.

The need for monitoring terrestrial systems that observe, quantify and map changing land use, search for and protect natural resources, and track interactions within the biosphere, atmosphere, hydrosphere, and geosphere has become a paramount concern to managers, politicians, and the general citizenry in developed and developing nations. This need has led to a mammoth international program to use a variety of technologies, centered on observation systems from space, to improve our ability to oversee and regulate the systems that govern Earth’s effective operations. Among names associated with this concept are the International Geosphere and Biosphere Programme (IGBP) (a synopsis of which is found at this United Nations Environomental Program site) and the International Global Change Program (IGCP). These programs cover a range of research and applications that embrace primarily climate studies, oceanography, and terrestrial environment monitoring. National programs include organizations that mainly make ground measurements but the current availability of suitable satellites flown by several countries leads to a symbiotic integration of space observations and ground measurements. This diagram depicts some of the primary topical activities, as described by their acronyms.

Some of the main components of the IGBP multinational effort.

The United States has been the kingpin in these efforts. Its chief role has been in providing many of the versatile satellites that make the critical land, sea, and air measurements on a global scale. The ESE Logo The program began in the early 1990s under the name Mission to Planet Earth; that program was renamed Earth Science Enterprise. ESE involves many federal agencies as well as some private organizations. NASA’s role, located primarily at Goddard Space Flight Center, is to operate the Earth Observing System (EOS) program which will plan, build, and launch a number of satellites, a list of these being found at this NASA Headquarters site.

The EOS Logo.

Closely allied to these and other programs is a new field of the geosciences called Earth System Science. Many Universities are now offering courses and even majors in this new field of natural science.

When these various programs are examined closely, as will be done throughout Section 16 of the Tutorial, two principal areas of emphasis underlie the goals and means of IGBP and ESE: 1) the concept of Global Change, which recognizes that the Earth’s natural systems are constantly modifying, with various diverse aspects such as atmospheric temperatures, air and water pollutants, and land cover interacting in often complex ways to alter environments; and 2) Global Climate, which is often the most important single component of the Earth System in controlling the changes over time and in different regions of the Earth. These modifications may be cyclical or unidirectional but generally take place slowly (almost imperceptibly over short time spans) and thus require extended, repeated coverage over years to decades using a variety of observational means (of which satellites are proving the most facile). These two Logos give URLs (which you must access separately) for these specific U.S. programs .

The U.S. Global Research Program

The U.S. Climate Science Program.

These programs will last well into the first decade of the 21st Century. Starting in 1998, several major platforms launched with broad complements of sensors supported by continuing operation of current sensor systems. The programs will have far-reaching impact on all nations and at least an indirect effect on all people on our planet, as they address problems and concerns tied to the environment and to resources. When coupled and integrated with other major data management and decision making approaches, GIS, ESE, and EOS should evolve into highly efficient implements for continuous gathering and processing of key elements of knowledge required to administer the complex interactions between nature and human endeavors.

If you want a preview of how some scientists apply remote sensing to monitor mankind’s influence on the environment, then go to the Home Page recently added to the Internet byThe Consortium for International Earth Science Information Network. Offline sources of basic information about remote sensing are the writer’s (NMS) still relevant 1982 NASA Publication RP 1078: The LANDSAT TUTORIAL WORKBOOK; MISSION TO PLANET EARTH: LANDSAT VIEWS THE WORLD (co-authored with Paul D. Lowman, Jr, Stanley C. Freden, and William C. Finch, Jr); THE HCMM ANTHOLOGY, NASA SP-465; and (co-authored with Robert Blair, Jr) GEOMORPHOLOGY FROM SPACE, NASA SP-486.

Here is a list of nine well-known textbooks that detail most of the fundamentals and applications of Earth Remote Sensing:

  • Avery, T.E. and Berlin, G.L., Fundamentals of Remote Sensing and Airphoto Interpretation, 5th Ed., 1992, MacMillan Publ. Co., 472 pp. (Note: 6th Ed forthcoming in 2001)

  • Campbell, J.B., Introduction to Remote Sensing, 2nd Ed., 1996, The Guilford Press

  • Drury, S.A., Image Interpretation in Geology, 2nd Ed., 1993, Chapman & Hall, 243 pp.

  • Drury, S.A., Images of the Earth: A Guide to Remote Sensing, 2nd Ed., 2nd Ed., 1998, Oxford University Press, 212 pp.

  • Kuehn, F. (Editor), Introductory Remote Sensing Principles and Concepts, 2000, Routledge, 215 pp.

  • Lillesand, T.M. and Kiefer, R.W., Remote Sensing and Image Interpretation, 4th Ed., 2000, J. Wiley & Sons, 720 pp.

  • Sabins, Jr., F.F., Remote Sensing: Principles and Interpretation. 3rd Ed., 1996, W.H. Freeman & Co., 496 pp.

  • Siegal, B.S. and Gillespie, A.R., Remote Sensing in Geology, 1980, J. Wiley& Sons (especially Chapters 1 through 11)

  • Swain, P.H. and Davis, S.M., Remote Sensing - the Quantitative Approach, 1978, McGraw-Hill Book Co.

An excellent blending of remote sensing imagery, ground photos, maps, and other types of geographic information is found in the Atlas of North America: A Space Portrait of a Continent, published by the National Geographic Society (1986).

Also of value are these Periodicals devoted largely to remote sensing methods and applications:

  • Canadian Journal of Remote Sensing

  • IEEE Transactions on Geoscience and Remote Sensing.

  • International Journal of Remote Sensing.

  • Photogrammetric Engineering and Remote Sensing.

  • Remote Sensing of the Environment

To expand upon the remarks at the beginning of the Overview, the primary purpose of this Tutorial is to be a learning resource for college students, as well as for individuals now in the work force who require indoctrination in the basics of space-centered remote sensing. In both instances the objective is to offer a background that will actually be useful in current or eventual job performance to those who may need to provide input information obtainable from remote sensing into day-to-day operations. We also think the Tutorial can be an invaluable resource for pre-college (mostly Secondary School) teachers who want to build a background in the essential contributions of the space program to society so as to better teach their students (many of whom should also be capable of working through the main ideas in the Tutorial). Our hope is that this survey of Satellite Remote Sensing will attract and inspire a few individuals from the world community who might consider a specialized career in this field or in the broader fields allied with Earth System Science (ESS) and the Environment (see below). An additional goal is to interest and inform the general public about the principles and achievements of remote sensing, with emphasis on demonstrated applications.

Here is our list of topics chosen to accomplish these objectives, by providing a comprehensive survey of remote sensing and its many ramifications.

TABLE OF CONTENTS

Foreword

Overview of this Remote Sensing Tutorial; “Getting Acquainted” Quiz

Introduction to Remote Sensing: Technical and Historical Perspectives; Special Applications such as Geophysical Satellites, Military Surveillance, and Medical Imaging

Section:

1. Image Processing and Interpretation: Morro Bay, California; First Exam

  1. Geologic Applications: Stratigraphy; Structure; Landforms

  2. Vegetation Applications: Agriculture; Forestry; Ecology

  3. Urban and Land Use Applications

  4. Mineral and Oil Resource Exploration:

6. Flight Across the United States: Boston to San Francisco; Quiz; World Tour

  1. Regional Studies: Use of Mosaics from Landsat

  2. Radar and Microwave Remote Sensing

  3. The Warm Earth: Thermal Remote Sensing

  4. Aerial Photography as Primary and Ancillary Data Sources

11. The Earth’s Surface in 3-Dimensions: Stereo Systems and Topographic Mapping

  1. The Human Remote Senser in Space: Astronaut Photography

13. Collecting Data at the Surface: Ground Truth; the “Multi” Concept; Hyperspectral Remote Sensing

14. The Water Planet: Meteorological, Oceanographic and Hydrologic Remote Sensing

  1. Geographic Information Systems: The GIS Approach to Decision Making

  2. Earth Systems Science; Earth Science Enterprise; and the EOS Program

  3. Use of Remote Sensing in Basic Science Studies I: Mega-Geomorphology

  4. Basic Science Studies II: Impact Cratering

  5. Planetary Remote Sensing: The Exploration of Extraterrestrial Bodies

20. Cosmology: Remote Sensing Systems that provide observations on the Content, Origin, and Development of the Universe

21. Remote Sensing into the 21st Century; Outlook for the Future; Final Exam

Appendix A: Modern History of Space

Appendix B: Interactive Image Processing

Appendix C: Principal Components Analysis

Appendix D: Glossary

Unlike a formal course in the subject, with chapters covering principles, techniques and applications in a pedagogic and systematic way, we lead you through a series of Sections focused on one to several relevant themes and topics. Because we can represent most remote sensing data as visuals, we will our organize our instructional treatment around illustrations, such as space images, classifications, maps, and plots, rather than numerical data sets. These data sets are the real knowledge base for application scientists in putting this information to practical use. (Much of this material has been acquired by direct downloading off the Internet. We are grateful to the source organizations and individuals but, for the most part, we do not acknowledge each contribution per se.) Descriptions and discussions accompany these illustrations to aid in interpreting the visual concepts. “Standard” space images, particularly those from Landsat sensors, are usually the focal points of a Section, but we frequently add special computer processed renditions with ground photos that depict features in a scene and descriptive maps where appropriate.

We also call out numerous links to other remote sensing sources and to various continuing or planned programs. Some of these programs are federal or international programs such as ESE, whereas, others are programs from educational or commercial organizations that provide training and services. These links, in turn, have their own sets of links, which, as you explore them, will broaden your acquaintance with the many facets of remote sensing and its popular applications.

The Tutorial begins with an Introduction, which covers the principles of physics (especially electromagnetic radiation) underlying remote sensing, then considers the main kinds of observing platforms, and includes the history of satellite systems, with a focus on Landsat. Many of the subsequent Sections and topics center on Landsat because it continues to be a kingpin among the current remote sensing systems. This Introduction also delves into three special topics: Use of satellites for geophysical measurements of Earth’s force fields; a survey of satellite programs (military and security agencies) employed in monitor activities detrimental to a country’s safety (these are often called “spy satellites), and the applications of intruments and techniques within the purview of remote sensing that are used in medical diagnosis.

This last topic may seem a bit strange as part of this Tutorial, which deals almost entirely with remote sensing data from satellites and spacecraft that look inwardly at Earth and outward at the heavens. But, medical remote sensing (or “medical imaging”) has been around for 100 years. For most people, use of medical instruments that examine the bodies of humans and their pets by means of electromagnetic radiation or force fields is the application of remote sensing of greatest personal familiarity and value in their lives. We treat this subject in three review pages in the Introduction. For now, let’s just look at two examples of the sensing of the human body using X-rays. The first image shows an x-ray radiograph of a diseased lung; the second is a CAT Scan (CAT = Computer Assisted Tomography) slice through the midsection of a torso showing the labelled organs:

X-ray radiograph of a human's chest showing a ancerous area in the left lung

CAT Scan slice through a human mid-torso.

Perusal through the Introduction and Sections 1, 8 and 9 is the minimum effort we suggest if you want to master the basics. The first Section (1) is one of the key chapters in this Tutorial because we try to introduce most of the major concepts of image analysis and interpretation by walking you through the product types and processing outputs in common use, using a single subscene as the focus. That subscene is a Landsat Thematic Mapper image of Morro Bay, California. This is what it looks like in a false color rendition:

Landsat TM image of Morro Bay, California (west of San Luis Obispo), in the standard false color rendition using Bands 2, 3, 4.

One ultimate goal in image processing is to produce a classification map of the identifiable features or classes of land cover in a scene. In Section 1 we examine various ways of enhancing a scene’s appearance and end with a supervised classification of the surface features we choose as meaningful to our intended use. Here is the classification of Morro Bay:

Supervised classification (16 classes) of the above Morro Bay TM subscene.

Section 8 is concerned with another mode of remote sensing, the use of radar and passive microwave. Seasat was the first civilian spacecraft that was dedicated to radar imaging. Radar has been flown several times on the U.S. Space Shuttle. This X-band Synthetic Aperture Radar (SAR) image (SIR-C mission) of Hong Kong is typical of this type of imagery:

SIR-C image of Hong Kong; note the characteristic mark of most radar images, namely, the distortion of hilly topography (one slope bright; foreshortening of that slope)

A later Shuttle flight - the Shuttle Radar Topography Mission (SRTM) - acquired both C-band and X-band images; these were utilized in calculating topographic altitudes. This SRTM image of Patagonia, Chile is assigned colors that correspond to ranges in altitude:

|SRTM C-band image of a mountainous region in Patagonia, South America; colors represent altitude intervals. |

Section 9 focuses on the increasing use of thermal infrared imagery obtained both from aircraft- and spacecraft-mounted sensors that operate mainly in two spectral regions: 3-5µm and 8-12µm. Here we will look at two modes of operation. Both images were acquired by the ASTER instrument on NASA’s Terra. The top is a 3.8µm image taken at night, showing the coastline of Eritrea in eastern Africa. At night, water is normally warmer than much of the land, so it show as a light tone. The bottom image is a multispectral color composite of three bands in the 8-10µm range. The area shown is the Saline Valley of eastern California (near Death Valley); most of the colors in this image can be related to rock types (silicates, carbonates, etc.).

Nighttime thermal IR image (ASTER) of part of Eritrea, in eastern Africa.

Color composite made with ASTER thermal bands showing the Saline Valley of California and surrounding mountains.

In Section 13, after a review of the methods of and necessity of “Ground Truth”, you are introduced to the concepts of spectroscopy and, more particularly, hyperspectral remote sensors (this is also discussed on page Intro-24. Hyperspectral sensors are revolutionizing the ability of remote sensing to make accurate and precise measurements of individual materials (e.g., rock types; plant species) using “spectrometers” operating on the ground, from the air, and now from space. Such a sensor is capable of imaging in narrow spectral width bands (typically 0.01 to 0.02 micrometers) over a broad, continuous range of the visible-Near Infrared spectrum. The resulting data set produces a detailed spectral signature (a plot of wavelengths versus some intensity function such as reflectance) for various features or classes within a scene which can be used to better identify these classes, often, in the case of composition of rocks or varieties of vegetation, leading to much higher accuracy in separating and discriminating the classes. To appreciate this grand “leap forward”, compare the two spectral plots in this figure - the upper one a spectral signature of a specific substance made with the 4 MSS bands on Landsat; the lower the hyperspectral equivalent signature:

Comparison of crude spectral signature made from MSS data with the signature made by hyperspectral sensing of the same materia.

Many examples of hyperspectral signatures, images, and applications are starting to appear in the literature and on the Internet. A much favored demonstration target is the Cuprite mining district in Nevada. Here is a natural color hyperspectral image (made with bands in the blue, green, and red part of the visible spectrum) of the hills around Cuprite that show distinctive alteration:

Cuprite, Nevada natural color image.

One of the first systems used for airborne hyperspectral surveying is JPL’s AVIRIS. Narrow spectral bands (equivalent to individual absorption bands in the detailed spectral signature) between 1.0 µm and 2.5µm are particularly sensitive to key diagnostic inflections of the spectral curve obtained. Sulphides, oxides, carbonates, etc. among ore minerals and alteration products can be pinpointed by their characteristic wavelengths such that individual mineral species can be identified. Here is an AVIRIS image of part of Cuprite, NV made from 3 longer wavelength bands:

AVIRIS IR image of mineralization at Cuprite, NV

Using appropriate analytical techniques, these different minerals can be highlighted after identification at specific locations, with other materials blacked out, thus producing a mineral distribution map of minerals that are specific ore guides. Thus:

Locations of specific minerals, using selected AVIRIS hyperspectral bands

A NASA satellite - EO-1, the first in the New Millenium series - hosts the first hyperspectral sensor (Hyperion) flown on a satellite, which subdivides the spectrum between 0.4 and 2.5 µm into 220 channels. To gain a feel for how well a scene can be classified using a large number of bands (channels), we show on the left a Landsat TM scene of a forest in which only the broad differentiation of tree types can be made and on the right a Hyperion-based classification that convincingly demonstrates the degree to which individual tree species can be identified:

image65image66


The emergence of hyperspectral sensors flown on both aircraft and spacecraft greatly increases the analysis capability in remote sensing, owing to the ability to generate a detailed spectral curve by dispersing the sensed electromagnetic radiation onto a large number of CCDs (charge-coupled detectors) which are resampled in microseconds, that may well be the most important new tool in earth-observing systems in the last ten years.

Although satellites concerned with meteorological and oceanographic/hydrologic phenomena constitute the largest number of Earth-observing platforms by far, we consider them primarily in Section 14. Readers of this Tutorial are certainly familiar with the Visible, Infrared, and Radar images of local, regional, continental, and hemispheric images of realtime weather systems moving in their vicinity because today’s area-specific Newscasts use and show relevant images during the Weather segment of the programs. These Metsat (a general term for meteorological satellites) images are for most people the most commonly encountered satellite data presented to the general public. For instance, this is an Accuweather image (Near Infrared) of the cloud distribution in the United States on June 30, 2002 (this will be updated hour by hour and can be easily accessed on the Internet):

Accuweather cloud map over the U.S. for June 30, 2002.

Less well known, but of great importance in understanding and predicting weather and climate on a global basis, are on-going measurements of oceanographic physical states. These, too, are investigated in Section 14 as satellites that obtain marine data are described. Here is a map of global ocean temperatures for a 3-day period in early June, 2002 as determined by Aqua, a mainstay of the Earth Observing System (EOS) program:

Sea surface temperatures in early June, 2000 determined by the AMSR-E (Advanced Microwave Scanning Radiometer) on Aqua.

Section 16 also deserves your careful reading. It treats an on-going program (EOS) started in the 1980’s that involves not only NASA but nearly all of the space agencies worldwide as well as environmental organizations from most of the nations now in the UN. Specifically treated are the status and results of several very sophisticated satellites - especially Terra and Aqua - that are part of the U.S.’s Earth Science Enterprise. This program, including satellites being launched by other countries, will peak during the first decade of the 21st Century but long range missions extend well into the new Millenium. The fleet of satellites is dedicated to supporting a new field of science, known as Earth System Science. That is a multidisciplinary approach to study of Earth at a global as well as regional scales. Particularly involved are oceanographers, meteorologists/climatologists, biologist/botanists, geologists/volcanologists, environmentalists/ ecologist,physicists, chemists, and even sociologists, economists, and members of the legal profession. To learn more about these programs prior to working through Section 16, check these links: Earth Science Enterprise and USRA Earth System Science

We shall see that observations made by Landsat, SPOT, the Metsats and oceanographic satellites, Terra, Aqua, and many other satellites described both below and elsewhere in the Tutorial have a very valuable functional asset: There are now enough of these in active orbits to cover almost the entire globe a number of times each day. (Good views are limited mainly by cloud cover.) Thus, timely (updated) information about continuing events can often be monitored successfully. One example that illustrates this is the great wildfire (more than 470,000 acres burned) that raged for weeks from June into July of 2002 in eastern Arizona. Because the multiplatform and multitemporal aspects of observations from space orbit are two of the most powerful ways to keep track of dynamic, changing events, we will demonstrate this now, in the Overview, with a series of images.

First are two NOAA-15 (meteorological satellite) images. The top covers a wide area of the desert Southwest. Acquired on June 20, 2002, this image has been processed to highlight the fire areas in red. The Arizona fire is sending smoke northeastward towards a second fire near Mesaverde, Colorado. In the bottom image, another set of bands on NOAA-15 were combined to show a false color composite that displays the two Arizona fires, Cheldiski (west) and Rodeo (east) before they had coalesced. Proximity to the town of Show Low (population 8000) suggests its citizenry had to be evacuated; exceptional efforts by the 2000+ firefighters saved it.

NOAA-15 image of parts of northeast Arizona, southeast Utah, and Colorado, showing two areas experiencing major fires in June, 2002.

Fires in the pinewoods of Arizona, near Show Low, as seen in a subset taken from a NOAA-15 image, acquired around June 22, 2002.

This next image was taken from the Internet but had no identification as to the spacecraft or sensor (it looks like an ASTER image from Terra). The burned areas stand out in this version. At this stage, the Cheldiski fire was still small and might have been contained if weather conditions favored the fire fighters who were preoccupied with the then larger Rodeo fire (which had been started through the carelessness of a Forest Ranger).

Image showing the burn scars of the Cheldiski-Rodeo fires.

Here is an image made by Terra’s MODIS sensor around June 25 still before two separate adjacent fires had merged into the largest wildfire in Arizona’s history.

MODIS image of northeast Arizona showing the wildfires around Show Low, AZ.

Landsat 7 also obtained timely images of the fires. Here it is shown at an early stage; the red outlines areas eventually burned up to June 30.

Landsat-7 image of the Arizona wildfires.

The final image in this set was taken by MODIS in early July, after the fire was 85% contained. This shows clearly how the two merged fires now produce a burn scar that looks like the end result of a single blaze

MODIS view of burn scar from merged Cheldiski-Rodeo fires.

Returning to our preview of especially germane chapters in the Tutorial, we especially call your attention to Sections 19 and 20 - Planetary Remote Sensing, and Cosmology. It is likely self-evident that the study of outer space - the Planets and the Cosmos - using remote sensors as the prime tool has a direct and vital bearing on how Man needs to understand the Universe beyond but will necessarily relate all that to life on Earth. One of the most famous of all pictures taken from Space - the view of Earth from above the Moon as Apollo 8 passed overhead at Christmastime in 1968 is reproduced here as a reminder that humankind’s quest for knowledge now links between our planet and those beyond it in the Solar System and by inference most probably in other planetary systems in faraway galaxies.

The rising Earth as seen from the Moon during the Apollo 8 circumlunar mission.

The reviews in Section 19 and 20 of these accomplishments elucidate what Science has learned about these fascinating other worlds (planets, satellites, and asteroids) and about the stars and galaxies and their origins. While these topics seemingly stray from the main Tutorial theme focusing on the Remote Sensing of Earth, they offer an in-depth summary of the main achievements in the exploration of our Solar System and the Universe beyond. This exploration was the centerpiece of the U.S. and Russian space programs and relied heavily on remote sensing techniques. In fact, as much or perhaps even more money has been spent on extraterrestrial remote sensing (consider the costs of Magellan and Voyager and the Hubble telescope) than on the study of the Earth (although, the dollar balance may be shifting with the new commericalization of terrestrial observations.)

In the first full decade of America’s Space Program, the Kennedy commitment to land astronauts on the Moon captured this country’s, and the world’s, imagination as no other space adventure has matched. Exploration of the Moon is symbolic of NASA’s greatest achievement. Even after the last Apollo crewmen left the lunar surface, its features have continued to be measured and analyzed. To commemorate this ongoing study of our satellite, shown here are two images of the Moon’s front side, one just before Apollo, the other in the last decade of the 20th Century. On the left is a full view of the Moon obtained through an Earth-based telescope. On the right is a false color composite of much the same area made by sensors aboard the Galileo spacecraft as it sat in an earth-parking orbit prior to being sent on its main mission to Jupiter.

Telescope picture of Earth's full MoonGalileo false color rendition of variations in the Moon's reflectance measured at different wavelengths.

Space probes with a variety of imaging sensors have allowed planetary scientists to look closely at the Outer Planets - Jupiter, Saturn, Uranus, and Neptune - and have revealed the great variety and complexity of the many moons (satellites) around these Giant planets. To introduce the wondrous information gathered by spacecraft such as Mariner, Voyager, and Galileo, we show this full hemisphere view of Io, the innermost jovian moon. Io can be nominated as the most active, dynamic planetary body in the Solar System, if as the prime criterion volcanism is selected as the indicator of this status.

Galileo image of Jupiter's volcanic moon Io; this has been superposed on a blue background (not the real color of outer space) to afford a nice contrast.

Section 20 considers most of the basic ideas of Astronomy and Cosmology (which, based on a Web Search, may well be the most comprehensive treatment of those two fundamental sciences now on the Internet). As a preview of the many truly beautiful, fascinating, and scientifically informative images spread throughout Section 20, we show here a montage of what has been called planetary nebulae (a misnomer based on an earlier misconception, since these great blobs of glowing gas and dust are not the precursors of eventual planet formation but are remnants of stars that have exploded as supernovae).

Montage of images taken through the Hubble Space Telescope of various planetary nebulae.

In the Cosmology Section (20) (specifically, page 20-4) images of stars and galaxies made by instruments on telescopes using different intervals of the spectrum are discussed in some detail. Here we give one specific example: the Andromeda Galaxy as seen in a visible light image and an infrared image (wavelength of 175 mm), which is also then reoriented by a computer program to show it face on. The differences in information displayed and revealed are striking.

The Andromeda Galaxy (M31) imaged in the visible and in the infrared by the International Space Observatory.

Section 21 is a brief (and somewhat out-of-date) review of some aspects of future satellites and programs, as well as a further look at products from several recently launched satellites.

Modern History of Space, Appendix A, was prepared by staff at the Air Force Academy as part of their contribution to the Internet version. It is an exceptional review and well worth a full read. For those who would like a quick (thumbnail) sketch of the role of remote sensing in space utilization, check this outline version prepared by Dr. John Estes of Univ. of California-Santa Barbara.

Appendix B can be a very important part of your learning efforts. It contains a downloadable image processing program called PIT, several sets of images, and detailed instructions on getting PIT to run and on carrying out specific processing functions, duplicating much of what is demonstrated in Section 1 about ways to enhance imagery and extract information. (CAUTION: While PIT has been reconfigured to download and install easily for most who choose to load this program onto their computers from the CD-ROM they have purchased, Internet users of this Tutorial will experience problems in downloading off the Internet - efforts to correct this are being made and should be done by April 1, 2003. )

Appendix C is a rather technical review of the concepts and underlying theory of Principal Components Analysis (PCA).

Appendix D is a fairly comprehensive Glossary. If you encounter a term or idea as you proceed through the Sections that may not be defined to your satisfaction, the Glossary is likely to have a concise definition to clarify the meaning.

NOTE: The Overview continues by clicking on the Next button in the banner below. Previous button returns you to the Foreword.


Primary Author: Nicholas M. Short, Sr. email: nmshort@nationi.net
Collaborators: Code 935 NASA GSFC, GST, USAF Academy