The specific regions or subdivisions of the electromagnetic spectrum are named and plotted. Mechanisms for generation of electromagnetic radiation are reviewed. The idea of incoming solar radiation or irradiance and its interaction (relative absorption) with the atmosphere is reviewed, and the notion of multispectral remote sensing over different spectral intervals is illustrated with a set of astronomical examples.
As noted on the previous page, electromagnetic radiation (EMR) extends over a wide range of energies and wavelengths (frequencies). A narrow range of EMR extending from 0.4 to 0.7 µm, the interval detected by the human eye, is known as the visible region (also referred to as light but physicists often use that term to include radiation beyond the visible). White light contains a mix of all wavelengths in the visible region. It was Sir Isaac Newton who first in 1666 carried out an experiment that showed visible light to be a continuous sequence of wavelengths that represented the different color the eye can see. He passed white light through a glass prism and got this result:
The principle supporting this result is that as radiation passes from one medium to another, it is bent according to a number called the index of refraction. This index is dependent on wavelength, so that the angle of bending varies systematically from red (longer wavelength; lower frequency) to blue (shorter wavelength; higher frequency). The process of separating the constituent colors in white light is known as dispersion. These phenomena also apply to radiation of wavelengths outside the visible (e.g., a crystal's atomic lattice serves as a diffraction device that bends x-rays in different directions).
The distribution of the continuum of all radiant energies can be plotted either as a function of wavelength or of frequency in a chart known as the electromagnetic (EM) spectrum. Using spectroscopes and other radiation detection instruments, over the years scientists have arbitrarily divided the EM spectrum into regions or intervals and applied descriptive names to them.The EM spectrum, plotted here in terms of wavelengths, is shown here.
Beneath is a composite illustration taken from the Landsat Tutorial Workbook (credited there to Lintz and Simonett, Remote Sensing of the Environment, who identify it as a modification of an earlier diagram by Robt. Colwell) that shows in its upper diagram the named spectral regions in terms of wavelength and frequency and in the lower diagram the physical phenomena that give rise to these radiation types and the instruments (sensors) used to detect the radiation.
Although it is somewhat redundant, we reproduce here still another plot of the EM Spectrum, with added items that are self-explanatory:
Colors in visible light are familiar to most, but the wavelength limits for each major color are probably not known to most readers. Here is a diagram that specifies these limits (the purple on the far left is in the non-visible ultraviolet; the deep red on the far right is the beginning of the infrared). The human eye is said to be able to distinguish thousands of slightly different colors (one estimate placed this at distinguishable 20000 color tints).
Different names for (wave)length units within intervals (those specified by types) that subdivide the EM spectrum, and based on the metric system, have been adopted by physicists as shown in this table:
(Both in this Tutorial and in other texts, just which units are chosen can be somewhat arbitrary, i.e., the authors may elect to use micrometers or nanometers for a spectral location in the visible. Thus, as an example, 5000 Angstroms, 500 nanometers, and 0.5 micrometers all refer to the same specific wavelength; see next paragraph.)
At the very energetic (high frequency and short wavelength) end are gamma rays and x-rays (whose wavelengths are normally measured in angstroms [Å], which in the metric scale are in units of 10-8 cm). Radiation in the ultraviolet extends from about 300 Å to about 4000 Å. It is convenient to measure the mid-regions of the spectrum in one of two units: micrometers (µm), which are multiples of 10-6 m or nanometers (nm), based on 10-9 m. The visible region occupies the range between 0.4 and 0.7 µm, or its equivalents of 4000 to 7000 Å or 400 to 700 NM The infrared region, spanning between 0.7 and 1000 µm (or 1 mm), has four subintervals of special interest: (1) reflected IR (0.7 - 3.0 µm), and (2) its film responsive subset, the photographic IR (0.7 - 0.9 µm); (3) and (4) thermal bands at (3 - 5 µm) and (8 - 14 µm). We measure longer wavelength intervals in units ranging from mm to cm. to meters. The microwave region spreads across 0.1 to 100 cm, which includes all of the interval used by radar systems. These systems generate their own active radiation and direct it towards targets of interest. The lowest frequency-longest wavelength region beyond 100 cm is the realm of radio bands, from VHF (very high frequency) to ELF (extremely low frequency); units applied to this region is often stated as frequencies in units of Hertz (1 Hz = 1 cycle per second; KHz, MHz and GHz are kilo-, mega-, and giga- Hertz respectively). Within any region, a collection of continuous wavelengths can be partioned into discrete intervals called bands.
I-9: Given that 1 nanometer (NM) = 10-9 m, 1 micrometers = 10-6 m and 1 Angstrom (A) = 10-10 m, how many nanometers in a micrometer; how many Angstrom units in a micrometer? ANSWER
Referring to the Phenomenology diagram (fourth illustration above): That chart indicates many of the atomic or molecular mechanisms for forming these different types of radiation; it also depicts the spectral ranges covered by many of the detector systems in common use. This diagram indicates that electromagnetic radiation is produced in a variety of ways. Most involve actions within the electronic structure of atoms or in movements of atoms within molecular structures (as affected by the type of bonding). One common mechanism is to excite an atom by heating or by electron bombardment which causes electrons in specific orbital shells to momentarily move to higher energy levels; upon dropping back to the original shell the energy gained is emitted as radiation of discrete wavelengths. At high energies even the atom itself can be dissociated, releasing photons of short wavelengths. And photons themselves, in an irradiation mode, are capable of causing atomic or molecular responses in target materials that generate emitted photons (in the reflected light process, the incoming photons that produce the response are not necessarily the same photons that leave the target).
On Earth, the atmosphere is both a boon and a bane in applying remote sensing to carrying out practical applications. Meteorological satellites have as their target the acquisition of data pertinent to weather forecasts and climate studies - this is the "boon" (this is discussed in detail in Section 14). But the atmosphere is simultaneously a hindrance to observing the terrestrial land and sea surfaces because it contributes radiation to the sensed signals and also absorbs radiation both from the Sun and from the target surfaces - this is the "bane". The most obvious negative influence from the atmosphere is the presence of clouds which prevent direct sensing of the surfaces over various wavelength intervals. Fortunately, to varying extents corrections can be made to compensate for the interfering effects of the atmosphere. (As an aside, atmospheres of other Solar System planets may have nil effects - such as Mercury, the Moon, and some of the Giant planet satellites, or have strong masking effects - such as Venus, or may be the actual targets being sensed - such as Jupiter, Saturn, Uranus, and Neptune, which are huge balls of gas.)
The constituents of the Earth's atmosphere that interact with transmitted or reflected radiation include the several gas elements (mainly Nitrogen, Oxygen, and Argon), gaseous compounds (CO, CO2, CH4, SO2, and others), water as vapor, liquid, or ice crystals, and aerosols (tiny particles of dust and various solid compounds and liquid droplets that make up smoke, haze, fog, and air pollutants). These various constituents will absorb or scatter incoming solar irradiance and do the same on refected and emitted radiation from the Earth's surface. Because the atmosphere has its own thermal status, i.e., its constituents are at various temperatures, it is also the source of emitted energy which is particularly a nuisance when thermal remote sensing of the surface is underway.
Most remote sensing is conducted above the Earth either within or above the atmosphere. When the primary objective is to sense the Earth's surface, the strategy is to utilize parts of the EM spectrum that minimalize interactions between radiation and the atmosphere. There are spectral intervals that maximize transmission of radiances. These are called "windows". Here is a generalized diagram showing relative atmospheric radiation transmission of different wavelengths.
Blue zones (absorption bands) mark minimal passage of incoming and/or outgoing radiation, whereas, white areas (transmission peaks) denote atmospheric windows, in which the radiation doesn't interact much with air molecules and hence, isn't absorbed. This next plot, made with the AVIRIS hyperspectral spectrometer (see page page 13-9), gives more a more detailed spectrum, made in the field looking up into the atmosphere, for the interval 0.4 to 2.5 µm (converted in the diagram to 400-2500 nanometers).
Most remote sensing instruments on air or space platforms operate in one or more of these windows by making their measurements with detectors tuned to specific frequencies (wavelengths) that pass through the atmosphere. However, some sensors, especially those on meteorological satellites, directly measure absorption phenomena, such as those associated with carbon dioxide, CO2 and other gaseous molecules. Note in the second diagram above that the atmosphere is nearly opaque to EM radiation in part of the mid-IR and almost all of the far-IR region (20 to 1000 µm). In the microwave region, by contrast, most of this radiation moves through unimpeded, so radar waves reach the surface (although raindrops cause backscattering that allows them to be detected). Fortunately, absorption and other interactions occur over many of the shorter wavelength regions, so that only a fraction of the incoming radiation reaches the surface; thus harmful cosmic rays and ultraviolet (UV) radiation that could inhibit or destroy certain life forms are largely prevented from hitting surface environments. I-10: From the first atmospheric absorption figure, list the four principal windows (by wavelength interval) open to effective remote sensing from above the atmosphere. ANSWER Backscattering (scattering of photons in all directions above the target in the hemisphere that lies on the source side) is a major phenomenon in the atmosphere. Mie scattering refers to reflection and refraction of radiation by atmospheric constituents (e.g., smoke) whose dimensions are of the order of the radiation wavelengths. Rayleigh scattering results from constituents (e.g., molecular gases [O2, N2 {and other nitrogen compounds}, and CO2], and water vapor) that are much smaller than the radiation wavelengths. Rayleigh scattering increases with decreasing (shorter) wavelengths, causing the preferential scattering of blue light (blue sky effect); however, the red sky tones at sunset and sunrise result from significant absorption of shorter wavelength visible light owing to greater "depth" of the atmospheric path as the Sun is near the horizon. Particles much larger than the irradiation wavelengths give rise to nonselective (wavelength-independent) scattering. Atmospheric backscatter can, under certain conditions, account for 80 to 90% of the radiant flux observed by a spacecraft sensor. Regardless of the degree of atmospheric interactions as a function of wavelength, there will always be some degradation of the signal, e.g., reflectance, by the atmosphere. The strategy then is to remove as far as feasible the desultory effects of the atmospheric component of the received signal. This is the prime purpose of "atmospheric corrections". There are two practical approaches. The first involves direct determination of the atmospheric contribution. One way is to make measurements on the ground or in the air itself (weather balloons) of the specific properties that make up the radiances added by the atmosphere. Another way is to use sensors on the ground that look upwards into the atmosphere to measure the offending properties. A third way is to transfer data that speak to these properties from meteorological satellites into the data processing programs that operate on the signals received from the ground. All of these require algorithms that intertwine with those data to remove the atmospheric contribution. The most widely used correction programs that rely on direct atmospheric measurements are LOWTRAN 7 and MODTRAN (briefly discussed at this website). The second approach is to make an internal correction to the sensed scene (, e.g., an image) utilizing this concept: Look for deep, dark shadows in the scene. The presumption is that no irradiance has impinged on the shadows; therefore, any none zero radiances (DNs above zero) are just the contributions from the atmosphere. (This presumption can never be exact because some irradiance onto the surface will come from the atmosphere itself, part of which will be sent back upwards.) But, these non-zero DNs have a large component comprised of the atmospheric interference. By subtracting the shadow DNs from all the non-shadow DNs, a partial correction for the atmosphere is achieved - it may indeed be capable of removing most of the atmosphere's contribution. This has another advantage. Over scenes that encompass large spatial areas (such as a Landsat image), the atmosphere will likely vary both horizontally and vertically. In principle, the correction will thus vary itself so that removal can be modified areally by referencing to the closest shadows (in practice, this is seldom done.) This discussion of the role of the atmosphere, and especially the value of atmospheric corrections, can be synergized by the following sequence of illustrations. The first two show how the spectral response curve for a class, or a mixed target, is modified by making an appropriate atmospheric correction (see the captions for details):
Atmosphere has an influence also in the thermal emission region of the EM spectrum. Various gaseous constituents contribute radiant emissions of differing magnitudes over different wavelength intervals. Radiance associated with these constituents can then be subtracted to get revised spectral signatures.
This next pair of images shows how clouds can be removed from a scene by subtracting the radiances they contribute and then resampling the signals from the surface to reproduce an otherwise hidden ground subscene:
The net effect of atmospheric correction is to brighten up a scene, as indicated in these before and after (correction) renditions.
A useful review of the effects of the atmosphere in remotely sensing the Earth's surface is found at this University of California-Berkeley website.
Remote sensing of the Earth traditionally has used reflected energy in the visible and infrared and emitted energy in the thermal infrared and microwave regions to gather radiation that can be analyzed numerically or used to generate images whose tonal variations represent different intensities of photons associated with a range of wavelengths that are received at the sensor. This sampling of a
(continuous or discontinuous) range(s) of wavelengths is the essence of what is usually termed multispectral remote sensing.
Images made from the varying wavelength/intensity signals coming from different parts of a scene will show variations in gray tones in black and white versions or colors (in terms of hue, saturation, and intensity in colored versions). Pictorial (image) representation of target objects and features in different spectral regions, usually using different sensors (commonly with bandpass filters) each tuned to accept and process the wave frequencies (wavelengths) that characterize a given region, will normally show significant differences in the distribution (patterns) of color or gray tones. It is this variation which gives rise to an image or picture. Each spectral band will produce an image which has a range of tones or colors characteristic of the spectral responses of the various objects in the scene; images made from different spectral bands show different tones or colors.
This point - that each spectral band image is unique and characteristic of its spectral makeup - can be dramatically illustrated with views of astronomical bodies viewed through telescopes (some on space platforms) equipped with different multispectral sensing devices. Below are four views of the nearby Crab Nebula, which is now in a state of chaotic expansion after a supernova explosion first sighted in 1054 A.D. by Chinese astronomers (see Section 20 - Cosmology - for other examples). The upper left illustration shows the Nebula as sensed in the high energy x-ray region; the upper right is a visual image; the lower left was acquired from the infrared region; and the lower right is a long wavelength radio telescope image. By sampling the radiation coming from any material or class under observation over a range of continuous (or intermittent, in bands) spectral interval, and measuring the intensity of reflectance or emittance for the different wavelengths involve, a plot of this variation forms what is referred to as a spectral signature, the subject of the next page's discussion.
Primary Author: Nicholas M. Short, Sr.