Tải bản đầy đủ - 0trang
Chapter 1. Aspects of Precision Agriculture
FRANCIS J. PIERCE AND PETER NOWAK
liming, P, and K)will be more easily managed precisely than those with large temporal
variance (e.g.. mobile insects). The potential for economic, environmental, and social
benefits of precision agriculture is complex and largely unrealized because the spacetime continuum of crop production has not been adequately addressed.
0 1999 Academic Press
It would be a simple matter to describe the earth’s surface if it were the same
everywhere. The environment,however, is not like that: there is almost endless
-Webster and Oliver ( I 990)
The quote by Webster and Oliver (1990) is particularly applicable because precision agriculture is concerned with the management of variability in the dimensions of both space and time. Without variability, the concept of precision agriculture has little meaning (Mulla and Schepers, 1997) and would never have
evolved. It appears that any component of production agriculture-from natural
resources to plants, production inputs, farm machinery, and farm operators-that
is variable in some way is included in the realm of precision agriculture. Aspects
of precision agriculture, therefore, encompass a broad array of topics, including
variability of the soil resource base, weather, plant genetics, crop diversity, machinery performance, and most physical, chemical, and biological inputs used in
the production of a crop, whether natural or synthetic. By necessity, these aspects
are all framed within the context of the socioeconomicaspects of production agriculture because to be successful on the farm, precision agriculture must fit the
needs and capabilities of the farmer (Nowak, 1997) and must be profitable (Lowenberg-DeBoer and Swinton, 1997).
Bell et al. (1995) state correctly that efforts toward precision agricultural management should recognize that the factors affecting crop yields and environmental sensitivity vary in both space and time. Managing soils and crops in space and
time is the sustainable management principle for the twenty-first century, a principle exemplified by farming by soilscapes, managing zones within the field, and
managing the noncrop period (Pierce and Lal, 1991). The unifying theme of this
chapter is that success in precision agriculture is directly related to how well it can
be applied to manage the space-time continuum in crop production. We postulate
that prospects for precision management increase as the degree of spatial dependence increases, but the degree of difficulty in achieving precision management
increases with temporal variance. Thus, for management parameters that vary spatially, those with high temporal correlations (e.g., liming) will be more easily man-
ASPECTS OF PRECISION AGRICULTURE
aged with precision agriculture than those with large temporal variance (e.g., mobile insects). Within a given management parameter, the success to date of precision management is to a large extent determined by the degree to which the spatial variability is temporally stable.
This chapter provides an overview of precision agriculture and an assessment
of its current state and its potential to improve crop performance and environmental quality in production agriculture. In this chapter, we define precision agriculture, explore the technological capabilities that enable it, assess its agronomic
feasibility and environmental efficacy, and evaluate its performance to date relative to economic and social impacts. The chapter concludes with an analysis to
identify needed developments in precision agriculture and we provide some
thoughts for a future research agenda. Given the expansive nature of precision
agriculture coupled with space constraints, we attempt to synthesize the important
aspects of precision agriculture while guiding the reader to the growing volume of
literature on the subject. Readers seeking more detail are referred to the following
major publications related to precision agriculture: Auernhammer ( 1994),American Society of Agricultural Engineers (ASAE) (1991), BIOS (1997), Lake et al.
(1997), National Research Council (NRC) (1997), Pierce and Sadler (1997),
Robert et al. (1993, 1995, 1996), Sawyer (1994, Schueller (1992), Stafford
(1996b), and Sudduth (1998). We are aware of the rapid rate of change in precision agriculture and the inadequacies this causes in an overview of this nature.
Currently, no precision agricultural systems exist; rather, various components
of traditional crop management systems have been addressed separately regarding
their potential for site-specific management, perhaps most notably soil fertility
(Lowenberg-DeBoer and Swinton, 1997).The state of precision agriculture from
a systems perspective is analogous to the early days of no-tillage crop production.
Technology became available in the 1960sto plant seeds in untilled soil, but it was
not until the many aspects of crop production were adequately addressed under
lack of tillage and crop residue management, including the management of fertility and pests, that successful no-tillage systems were developed and implemented
(Blevins et al., 1998). The adoption of no-tillage did not proceed at a significant
rate until the 1980s when the integration of appropriate technologies and public
policies supported its dissemination to farmers (Allmaras et al., 1998; Larson et
al., 1998;Now& and Korsching, 1998).In a similar fashion, although certain technologies in the early days of precision agriculture allowed for the variable application of nutrients and pesticides, there did not exist a thorough understanding of
how soil fertility and pests varied in space and time. Most important, explanations
were lacking on what specifically caused the variability so that appropriate inputs
FRANCIS J. PIERCE AND PETER NOWAK
could be matched to site-specific conditions. Today, farmers are adopting individual components of precision agriculture on the farm but a distinctive precision
farming system has not yet evolved. Technological developments continue to occur and as a result of ongoing research a better understanding of underlying
processes is being developed but a true system has not emerged. Therefore, any
definition of precision agriculture can at best be considered only operational.
Since the mid-l980s, a host of terms have been used to describe the concept of
precision agriculture, including farming by the foot (Reichenberger and Russnogle, 1989), farming by soil (Carr et al., 1991; Larson and Robert, 1991). variable rate technology (VRT) (Sawyer, 1994), spatially variable, precision, prescription, or site-specific crop production (Schueller, 1991), and site-specific
management (Pierce and Sadler, 1997). All these terms, however, have in common the concept of managing variability at scales that are within fields. As
Stafford (1996b; p. 595) states, precision agriculture involves “the targeting of inputs to arable crop production according to crop requirements on a localized basis.” Thus, the intent of precision agriculture is to match agricultural inputs and
practices to localized conditions within a field to do the right thing, in the right
place, at the right time, and in the right way (Pierce ef al., 1994). A recent report
of a National Research Council, Board on Agriculture Committee defined precision agriculture as “a management strategy that uses information technologies to
bring data from multiple sources to bear on decisions associated with crop production” (NRC, 1997; p. 17). While the NRC definition raises important informational dimensions of precision agriculture, it fails to emphasize the basic
premise of precision agriculture-the management of spatial and temporal variability. In this chapter, we use the following definition of precision agriculture as
the basis of our discussions: Precision agriculture is the application of technologies and principles to manage spatial and temporal variability associated with all
aspects of agricultural production for the purpose of improving crop performance
and environmental quality.
We provide a final note on the word precision because there is sure to be confusion regarding its meaning in precision agriculture versus its use in statistics. The
term precision refers to the quality or state of being precise, where precise means
minutely exact, a term synonymous with correct. Precision agriculturerefers to exactness and implies correctness or accuracy in any aspect of production. In statistics, however, precision is the closeness of repeated measurements of the same
quantity to each other, whereas accuracy is the closeness of a measured or computed value to its true value (Sokal and Rohlf, 1995). In measurements, accuracy
is synonymous with correctness (i.e., validity), whereas precision refers to reproducibility (i.e., reliability). Thus, something can be precise but not accurate. Another matter is measurement precision implied by number of digits reported for a
given measurement. The nature of computers makes it easy to imply more precision than was possible in various aspects of data collection, analysis, and compu-
ASPECTS OF PRECISION AGRICULTURE
tation in precision agriculture. Precision here refers to the limits on the measurement scale between which the true measurement is believed to lie, implied by the
number of digits reported for a measurement (Sokal and Rohlf, 1995). The more
digits reported for a measurement, the higher the precision implied. A pH of 5.44
implies more precision than a pH of 5.4. The appropriate precision with which to
report a number is to include one additional digit beyond the last significant one
measured by the observer. Statistics plays an important role in the application of
precision agriculture and care should be taken in dealing with accuracy, precision,
and implied precision in the reporting data.
Precision agriculture is intuitively appealing because it is closely aligned with
the scientific principles of management of soils, crops, and pests. Few would argue against a management philosophy that espouses matching inputs to the exact
needs everywhere. Precision agriculture is intuitively appealing because it offers
a means to improve crop performance and environmental quality in production
agriculture (Wolf and Nowak, 1995). While the intuitive appeal creates high expectations for precision agriculture, the physical evidence supporting the agronomic (Lowenberg-DeBoer and Swinton, 1997; Sawyer, 1994) and environmental (Larson et al., 1997) benefits of precision agriculture is limited in part because
it is still in its infancy.
As we will demonstrate in our discussion, successful implementation of precision agriculture depends on numerous factors, including (i) the extent to which
conditions within a field are known and manageable, (ii) the adequacy of input recommendations, (iii) the degree of application control, and (iv) the degree of support through private and public infrastructures. Individual success also depends on
the expectationsplaced on precision agriculture which represent the difference between promotional and educational efforts versus the actual experience of farmers.
II. OVERVIEW OF THE BASIC COMPONENTS
OF PRECISION FARMING
The main componentsof any precision agriculture system that may emerge must
first address the measurement and understanding of variability. Next, this system
must use information to manage this variability by matching inputs to conditions
within fields using site-specific management recommendations and mechanisms
to control the accuracy of site-specific inputs. Finally, and most important, this system must provide for the measurement and recording of the efficiency and effica-
FRANCIS J. PIERCE AND PETER NOWAK
cy of these site-specific practices in order to assess value on and off the farm. Thus,
precision agriculture is technology enabled, information based, and decision focused (Pierce, 1997a).
While the concept of matching inputs to site-specific conditions is not new, as
just discussed, there is little doubt that important advances in technology continue to enable precision agriculture. The enabling technologies of precision agriculture can be grouped into five major categories: computers, global position system
(GPS), geographic information systems (GIS), sensors, and application control.
Few of the enabling technologies were developed specifically for agriculture and
their origins date back more than 20 years, as illustrated in the time chart in Figure 1. It is the integration of these technologies that has enabled farmers and their
service providers to do things not previously possible, at levels of detail never
before obtainable, and, when done correctly, at levels of quality never before
achieved (Fortin and Pierce, 1998).
Many technologies support precision agriculture, but none is more important
than computers in making precision agriculture possible. Also, it is not computers
alone that are important but their ability to communicate that is so powerful for
agriculture.As Taylor and Wacker (1997) suggest, it is the fusion of computers and
communication that gave birth to connectivity, and it is connectivity that is driving
the access of everyone to everyone, everything to everything, and everything to
everyone. This electronic linkage and communication define the age of access
(Taylor and Wacker, 1997). It is this notion that may have prompted the NRC
(1997) to define precision agriculture in terms of a management strategy that uses
information technologies for decision making.
Precision agriculture requires the acquisition, management, analysis, and output of large amounts of spatial and temporal data. Mobile computing systems were
needed to function on the go in farming operations because desktop systems in the
farm office were not sufficient. These mobile systems needed microprocessorsthat
could operate at speeds of millions of instructions per second (MIPS), had expansive memory, and could store massive amounts of data. The first microchip created by Intel in 1971 (Intel 4044 processor) contained a mere 2300 transistors and
performed about 60,000instructions per second. Since 1971, the number of transistors per chip has doubled every 18 months (Fig. 2) affirming Gordon Moore’s
observation in 1965 that a doubling of transistor density on a manufactured die
was occurring every year, a concept referred to as “Moore’s law” (Moore, 1997).