Читать книгу Introduction to the Physics and Techniques of Remote Sensing - Jakob J. van Zyl - Страница 14

The early development of remote sensing as a scientific field was closely tied to developments in photography. The first photographs were reportedly taken by Daguerre and Niepce in 1839. The following year, Arago, Director of the Paris Observatory, advocated the use of photography for topographic purposes. In 1849, Colonel Aimé Laussedat, an officer in the French Corps of Engineers, embarked on an exhaustive program to use photography in topographic mapping. By 1858, balloons were being used to acquire photography of large areas. This was followed by the use of kites in the 1880s and pigeons in the early 1900s to carry cameras to many hundred meters of altitude. The advent of the airplane made aerial photography a very useful tool because acquisition of data over specific areas and under controlled conditions became possible. The first recorded photographs were taken from an airplane piloted by Wilbur Wright in 1909 over Centocelli, Italy.

Figure 1.7 Absorption spectrum of H₂O for two pressures (100 and 1000 mbars), at a constant temperature of 273 ° K.

Source: Chahine et al. (1983). © 1983, American Society of Photogrammetry.

Color photography became available in the mid‐1930s. At the same time, work was continuing on the development of films that were sensitive to near‐infrared radiation. Near‐infrared photography was particularly useful for haze penetration. During World War II, research was conducted on the spectral reflectance properties of natural terrain and the availability of photographic emulsions for aerial color infrared photography. The main incentive was to develop techniques for camouflage detection.

In 1956, Colwell performed some of the early experiments on the use of special‐purpose aerial photography for the classification and recognition of vegetation types and the detection of diseased and damaged vegetation. Beginning in the mid‐1960s, a large number of studies of the application of color infrared and multispectral photography were undertaken under the sponsorship of NASA, leading to the launch of multispectral imagers on the Landsat satellites in the 1970s. A major breakthrough was the advent of electronic detectors particularly in large arrays and across many spectral regions.

Figure 1.8 Spectral signature of some vegetation types.

Source: From Brooks (1972).

At the long wavelength end of the spectrum, active microwave systems have been used since early this century and particularly after World War II to detect and track moving objects such as ships and, later, planes. More recently, active microwave sensors have been developed providing two‐dimensional images that look very similar to regular photography, except the image brightness is a reflection of the scattering properties of the surface in the microwave region. Passive microwave sensors were also developed to provide “photographs” of the microwave emission of natural objects.

The tracking and ranging capabilities of radio systems were known as early as 1889, when Heinrich Hertz showed that solid objects reflected radio waves. In the first quarter of this century, a number of investigations were conducted in the use of radar systems for the detection and tracking of ships and planes and for the study of the ionosphere.

Radar work expanded dramatically during World War II. Today, the diversity of applications for radar is truly startling. It is being used to study ocean surface features, lower and upper atmospheric phenomena, subsurface and surface land structures, and surface cover. Radar sensors exist in many different configurations. These include altimeters to provide topographic measurements, scatterometers to measure surface roughness, and polarimetric and interferometric imagers.

Figure 1.9 Landsat TM images of Death Valley acquired at 0.48 μm (a), 0.56 μm (b), 0.66 μm (c), 0.83 μm (d), 1.65 μm (e), and 11.5 μm (f).

In the mid‐1950s, extensive work took place in the development of real aperture airborne imaging radars. At about the same time, work was ongoing in developing synthetic aperture imaging radars (SAR), which use coherent signals to achieve high‐resolution capability from high‐flying aircraft. These systems became available to the scientific community in the mid‐1960s. Since then, work has continued at a number of institutions to develop the capability of radar sensors to study natural surfaces. This work led to the orbital flight around the Earth of the Seasat SAR (1978) and the Shuttle Imaging Radar (1981, 1984). Since then, several countries have flown orbital SAR systems.

The most recently introduced remote sensing instrument is the laser, which was first developed in 1960. It is mainly being used for atmospheric studies, topographic mapping, and surface studies by fluorescence.

There has been great progress in spaceborne remote sensing over the past three decades. Most of the early remote sensing satellites were developed exclusively by government agencies in a small number of countries. Now, nearly 20 countries are either developing or flying remote sensing satellites. And many of these satellites are developed, launched, and operated by commercial firms. In some cases, these commercial firms have completely replaced government developments, and the original developers in the governments now are simply the customers of the commercial firms.

Figure 1.10 Images of an area near Cuprite, Nevada, acquired with an airborne imaging spectrometer. The image is shown to the left. The spectral curves derived from the image data are compared to the spectral curves measured in the laboratory using samples from the same area.

Source: Courtesy of JPL. See color section.

The capabilities of remote sensing satellites have also dramatically increased over the past three decades. The number of spectral channels available has grown from a few to more than 200 in the case of the Hyperion instrument. Resolutions of a few meters or less are now available from commercial vendors. Synthetic aperture radars are now capable of collecting images on demand in many different modes. Satellites are now acquiring images of other planets in more spectral channels and with better resolutions than what was available for the Earth two decades ago. And as the remote sensing data have become more available, the number of applications has grown. In many cases, the limitation now has shifted from the technology that acquires the data to the techniques and training to optimally exploit the information embedded in the remote sensing data.

Figure 1.11 Sea surface temperature derived from ship observations (a) and from the Seasat Multispectral Microwave Radiometer (b). (c) shows the difference.

Source: From Liu (1983). © 1983, John Wiley & Sons.

Figure 1.12 Backscatter data acquired over the Amazon region (insert). The different curves correspond to different incidence angles. Data were acquired by the Seasat Scatterometer at 14.6 GHz and at VV polarization.

Source: Bracalante et al. (1980). © 1980, IEEE.

Figure 1.13 Profile of Tharsis region (Mars) acquired with Earth‐based radar.

Figure 1.14 Profiles of an unnamed impact basin on Mars using Earth‐based radar. The set of profiles shown correspond to the box overlay on the figure.