Tải bản đầy đủ - 0trang
CHAPTER 7. THE ROLE OF REMOTE SENSING IN DETERMINING THE DISTRIBUTION AND YIELD OF CROPS
MARVIN E. BAUER
much needed information. Remote sensing has the potential to revolutionize the detection and characterization of many agricultural phenomena.
Recent studies indicate that remote sensing techniques can be used in the
visible, infrared, and microwave regions of the electromagnetic spectrum
to collect measurements of reflectance and emittance of plants, soils, water,
and other materials. With a minimum amount of ground sampling, remote
sensing data will permit identification and area measurements of crops,
assessment of crop stress, yield forecasts, range surveys, and mapping of
major soil boundaries, as well as many nonagricultural applications.
This article reviews the physical basis for remote sensing and its historical development and discusses potential agricultural applications of remote
II. Remote Sensing Development
Remote sensing is the acquisition and interpretation of spectral measurements made at a distant location to obtain information about the earth’s
surface. Remote sensing, as it is known today, is an outgrowth of aerial
photography. Although the use of aerial photography has been developing
for more than a hundred years, remote sensing is a relatively new term,
used only since about 1960. And, since 1960, the field has been rapidly
evolving and expanding as new sensors and interpretation techniques become available and new uses for the technology are developed. Fundamental and applied research has consistently led to the effective, sophisticated current capacity to acquire great quantities of spectral data and to
process, analyze, and interpret the data rapidly. The ability to sense and
interpret the radiance of crops and soils has improved as the overall technology advanced. Today, one of the most promising applications of remote
sensing technology is its ability to obtain information about agricultural
Military intelligence photographs, taken from balloons during the American Civil War, were one of the first uses of remote sensing. By World
War I, the photographic processes had been improved, and aerial photos
were used extensively for military reconnaissance. Although the coverage
was poor and the area photographed was limited, the potentials for inventory and cartography were readily recognized.
The first agricultural remote sensing was by Cobb (1922), who experimented with the use of aerial photography for soil mapping as early as
1918. In 1929, Bushnell reported on the use of aerial photographs of an
entire county for the soil survey program of the U.S.Department of Agri-
ROLE OF REMOTE SENSING
culture. He recommended aerial pictures as an aid to all future soil survey
work, claiming the photographs to be practical, economical, and necessary.
Beginning in the 1920s, considerable amounts of aerial photographic
coverage were obtained of the United States. These images were mainly
used by the U.S. Department of Agriculture. By 1938, the use of aerial
photographs as base maps had become standard procedure in the national
cooperative soil survey (Baldwin et al., 1938). And the photographs were
used by the Agricultural Stabilization and Conservation Service in the administration of farm programs. The Forest Service has also used aerial
Many advances in photointerpretation techniques were made during and
after World War 11. Colwell’s (1956) finding that crops stresses could be
identified from infrared imagery was particularly significant.
Since the 1960 launch of the TIROS-1 satellite, which provided television and infrared observations, other sources and uses of remote sensing
data have expanded. TIROS soon evolved into an operational system for
weather observation, with the results attracting much attention, and other
possible uses were quickly recognized. Later pictures taken during U.S.
manned space flights confirmed the potential for obtaining great amounts
of information about earth resources from observational data collected by
Meanwhile, research was being conducted using ground-based and airborne sensors to perfect sensor technology and data analysis, as well as
to develop the uses of remotely sensed data. In 1964, multispectral photography was collected for the first time over agricultural fields, and the potential of the multispectral approach to crop identification was recognized
(Hoffer, 1967). After this approach was further defined, a crop classification of five square miles was made from multispectral scanner data in 1967,
using pattern recognition methods implemented on a digital computer
(Lab. for Agr. Remote Sensing, 1968). During the 1967-1974 period, the
multispectral approach was further developed to encompass increasing land
areas, techniques, and disciplines.
The Corn Blight Watch Experiment, conducted in 1971 by several agencies of NASA and USDA, Purdue University, the University of Michigan,
and the Cooperative Extension Services of seven Corn Belt states, provided
a prototype remote sensing system. The prototype integrated techniques
of sampling, data acquisition, storage, retrieval, processing, analysis, and
information dissemination in a quasi-operational system environment
(MacDonald et al., 1972). The results showed that remote sensing procedures could quantitatively recognize corn leaf blight over broad areas.
Other agricultural crops and land uses were also accurately identified.
The supply of remotely sensed data greatly increased with the launch
MARVIN E. BAUER
of the Earth Resources Technology Satellite (ERTS-1) in 1972. From
an orbit 570 miles above the earth, the satellite can complete a full observation of the earth every 18 days. Its multispectral imagery is collected in
four visible and infrared wavelength bands for 100-mile wide passes
over the earth. This newest source of data has opened a whole new dimension to the capability to obtain information about earth resources, particularly crops.
The remainder of this paper discusses the physical basis for and the
applications of remote sensing of crops.
Physical Basis for Remote Sensing
The physical basis for remote sensing is the distinctive character of electromagnetic radiance from natural and man-made scenes (Holmes and
MacDonald, 1969). In this review, we will define remote sensing as the
acquisition of information about the earth’s surface from measurements
of radiated energy made by aircraft- or spacecraft-borne sensors. This section summarizes the basic systems and concepts involved in the acquisition
and analysis of remotely sensed data. The material will be presented in
the form of an energy-flow profile, consisting of (1) the source of energy;
(2) energy flow through the atmosphere; (3) its interaction with the target;
(4) measurement and recording of energy flow; and ( 5 ) processing and
analysis of recorded energy.
The major objective of remote sensing is to detect, measure, record,
and analyze energy in selected portions of the electromagnetic spectrum.
The variations in electromagnetic fields that can be measured and used to
discriminate among objects are spectral, spatial, and temporal (Landgrebe,
1973). Figure 1 reviews the electromagnetic spectrum. The region from
0.3 pm to 100 cm is used for remote sensing. The visible portion extending
from 0.4 to 0.7 pm is the most familiar because our eyes are sensitive
to radiation at those wavelengths. However, other portions of the spectrum
are equally important for remote sensing because the level of energy reflected or emitted from materials normally varies with wavelength throughout the spectrum. A material can often be identified by its spectral characteristics if the energy that it is reflecting and emitting is broken down
into carefully chosen wavelength bands. Sensors with broad-band sensitivity
may tend to inhibit differentiation of crops and other materials; however,
discrimination capabilities are generally improved by selectively measuring
and analyzing the energy from several discrete wavelength bands. The latter
approach is known as the multispectral approach and will be described
further in the discussion of data analysis.
ROLE OF REMOTE SENSING
tmic ron s)
FIG.1. The electromagnetic spectrum. The lower part emphasizes the regions of
primary importance in remote multispectral sensing. (From Hoffer, 1967.)
Sensors that measure the energy naturally radiated by objects are passive
remote sensors. Active remote sensors, such as radar, transmit energy to
the object and measure the portion which is reflected back. Solar radiation
is the ultimate source of energy for passive systems. A portion of the incident solar radiation (at wavelengths from 0.4 to 3 pm) is immediately