American Journal of Agricultural and Biological Sciences 5 (1): 50-55, 2010

ISSN 1557-4989
© 2010 Science Publications

A Review: The Role of Remote Sensing in Precision Agriculture

S. Liaghat and S.K. Balasundram

Department of Agriculture Technology, Faculty of Agriculture,
University Putra Malaysia, 43400 Serdang, Selangor, Malaysia

Abstract: Precision agriculture is an emerging farm management strategy that is changing the way
people farm. At present, there is an increasing commitment to reduce reliance on excessive chemical
inputs in agriculture. Numerous technologies have been applied to make agricultural products safer and
to lower their adverse impacts on the environment, a goal that is consistent with sustainable
agriculture. Precision agriculture has emerged as a valuable component of the framework to achieve
this goal. This review highlights on remote sensing technology and describes how it can be used as
an effective tool in precision agriculture.

Key words: Sustainable agriculture, environment, remotely sensed data, spectral reflectance

INTRODUCTION impacts on the environment, a goal that is consistent

with sustainable agriculture. PA has emerged as a
Since the Industrial Revolution, agriculture valuable component of the framework to achieve this
practices in developed countries have tended to support goal (Kropff et al., 1997; Tellaeche et al., 2008;
greater energy inputs using large machineries and Zhang et al., 2002).
increased applications of chemicals and fertilizers. PA is an integrated, information- and production-
While these practices have negative societal and based farming system that is designed to increase long
environmental implications such as soil erosion and term, site-specific and whole farm production
salinization, soil fertility, compaction of subsoils and efficiency, productivity and profitability while avoiding
soil/water pollution, they have generally supported the the undesirable effects of excess chemical loading to the
food and fiber needs of a rapidly growing human environment or productivity loss due to insufficient
population. A paradigm shift toward a new production input application. The inference is that better decision
method that ensures safe and sustainable agriculture is making will provide a wide range of benefits in
needed. Across the globe, Precision Agriculture (PA) is economic, environmental and social aspects that may or
changing the way people farm as it offers a myriad of may not be known or measurable at present
potential benefits in profitability, productivity, (Auernhammer, 2001). Worldwide, investments in
sustainability, crop quality, environmental protection, research and technology development on PA have
on-farm quality of life, food safety and rural economic considerably increased during the past decade
development. PA is an innovative, integrated and (Schellberg et al., 2008).
internationally standardized approach aiming to The importance of PA has been widely recognized
increase the efficiency of resource use and to reduce the as a key contributor in crop production technology
uncertainty of decisions required to manage variability around the globe, but so far, this technology is only
on farms. PA has been hailed as one of the most becoming practicable on large farms. PA is based on
scientific and modern approaches to production innovative systems approach and these new systems
agriculture in the 21st century, as it epitomizes a better approach depends on a combination of fundamental
balance between reliance on traditional knowledge and technologies such as Geographic Information System
information and management-intensive technologies. (GIS), Global Positioning System (GPS), computer
At present, there is an increasing commitment to reduce modeling, ground based/airborne/satellite remote
reliance on excessive chemical inputs in agriculture. sensing, variable rate technology and advanced
Numerous technologies have been applied to make information processing for timely in-season and
agricultural products safer and to lower their adverse between season crop management. Protocols for PA

Corresponding Author: Siva K. Balasundram, Department of Agriculture Technology, Faculty of Agriculture,

University Putra Malaysia, 43400 Serdang, Selangor, Malaysia
Tel: +603-8946 4186 Fax: +603-8946 4146
50
 Am. J. Agri. & Biol. Sci., 5 (1): 50-55, 2010

implementation can be encapsulated in three general

steps: (1) Gathering information about variability, (2)
Processing and analyzing information to assess the
significance of variability and (3) Implementing change
in the management of inputs. PA subscribes to a cyclic
process that is typified by a system that gets smarter
every year a farm operator uses it. Taken together, this
implies that the three general steps may each require a
timeframe ranging from months to years so as to
collectively generate a stable and workable structure,
but yet be subjected to further amendments and
refinements. The importance of PA evolved during the
era of agricultural mechanization in the 20th century
where there was strong economic pressure to treat large Fig. 1: The visible region of the spectrum ranges from
fields with uniform agronomic practices. PA provides a about 0.4-0.7 µm (Nowatzki et al., 2004)
means to automate Site-Specific Management (SSM)
using information technology, thereby making SSM Remote sensing technology: Remote Sensing (RS) is
practical in commercial agriculture (Auernhammer, the science of obtaining and interpreting information
2001). from a distance, using sensors that are not in physical
The PA database generally includes contact with the object being observed (Jensen, 1996).
(Venkataratnam, 2001): The science of remote sensing includes aerial, satellite
and spacecraft observations of the surfaces and
• Crop information such as growth stage, health, atmospheres of the planets in our solar system, while
nutrient requirement the Earth is noticeably the most frequent target of study.
• Soil physical and chemical properties, depth, RS is usually restricted to methods that detect and
texture, nutrient status, salinity and toxicity, soil measure electromagnetic energy including visible and
temperature, productivity potential non-visible radiation that interact with surface materials
• Microclimatic data (seasonal and daily) such as and the atmosphere (Fig. 1).
canopy temperature, wind direction and speed, RS and GIS technologies have been of great use to
humidity planners in planning for efficient use of natural
• Surface and sub surface drainage conditions resources at national, state and district levels.
Application of these technologies in the management of
• Irrigation facilities, water availability and planning
natural resources are increasing rapidly due to great
of other inputs
strides made in space-borne RS satellites in terms of
spatial, temporal, spectral and radiometric resolutions
A PA system that harnesses recent advances in
(Venkataratnam, 2001).
sensor technology can play a crucial role toward an
Remote sensing has several unique advantages
intelligent crop production system. Specifically, remote (Jensen, 1996):
sensing technology that allows non-destructive
acquisition of information about the Earth’s surface can • RS technology is well-known as a non destructive
facilitate the implementation of PA. For example method to collect information about earth features
current crop status (including maturity period) and crop • RS data may be obtained systematically over very
stresses such as nutrient and water stress, disease, pest large geographical areas rather than just single
and weed infestations can be discerned by means of point observations
remote sensing instruments such as cameras, laser • RS data can reveal information about places that
scanners, linear arrays and area arrays, without actually are inaccessible to human exploration
being in contact with them. Information gathered via • The systematic (raster) data collection in RS can
different sensors and referenced using a GPS can be remove sampling bias
integrated to create field management strategies for • RS can provide fundamental biophysical
chemical application, cultivation and harvest. This information that can be used in other sciences
review highlights on remote sensing technology and • RS is independent from the data produced
describes how it can be used as an effective tool in elsewhere, in comparison with the other mapping
Precision Agriculture. sciences such as cartography or GIS
51
 Am. J. Agri. & Biol. Sci., 5 (1): 50-55, 2010

The history of modern RS began when black and Table 1: QuickBird satellite imagery specification (Pan et al., 2009)
white photographs of the landscape were first taken Spatial resolution 0.61 m panchromatic resolution and 2.44 m
multispectral resolution at nadir
from the air. The first utilization of aerial photographs Spectral resolution Four multispectral bands (three in the visible
appeared for military purposes. Eventually, the and one in the near infrared)
Agriculture Stabilization and Conservation Service Radiometric resolution 11 bit dynamic range (2048 levels of grey)
Temporal resolution A reviewing rate that depends on the off nadir
adopted the technology and began collecting indexed angle. In any case, applications that require
photographs of the landscape for agricultural purposes. multitemporal observation with a seasonal
RS, today, incorporates new technologies that provide frequency can be easily carried out with
QuickBird data
increasingly efficient, complete, accurate and timely Swath In the range of 16.5-18.0 km
information. Today, RS is potentially a practical Frame 270 up to more than 300 km2
management tool for site-specific crop management
(Casady and Palm, 2002).

How can remotely-sensed data be used in

agriculture? RS technology is a key component of PA
and is being used by an increasing number of scientists,
engineers and large-scale crop growers. During the last
two decades, development in RS data acquisition
capabilities, data processing and interpretation of
ground based, airborne and satellite observations have
made it possible to couple RS technologies and
precision crop management systems (Waheed et al.,
2006).
Currently, there is a wide range of satellite data
that varies in (i) technique (active/passive,
radiometer/scatterometer), (ii) spatial resolution from
submeter to kilometers (iii) spectral range, and (iv)
viewing geometry (Oza et al., 2008). The full
commercial availability of very high resolution satellite
data has opened up a number of new opportunities for
the use of Earth Observation (EO) data. Today, we can
perform many applications with EO data that in only Fig. 2: A QuickBird satellite imagery taken from an oil
the recent past were exclusive to manpower palm plantation in Merlimau, Melaka, Malaysia
investigation and in situ surveys, which was time-
consuming and hard-sledding, despite the geographic parameters (such as crop cover, crop health and soil
limitations of such data and techniques. Satellite moisture) and are useful for operations such as stress
imagery can be acquired over any area globally, in a mapping, fertilization and pesticide application and
time frame and at a given price. At present, higher irrigation management (Barnes and Baker, 2000; Barroso
resolution satellite imagery overcomes previous et al., 2008; Hinzman et al., 1986; Lelong et al., 1998;
constraints and permits the use of such data as a quick Pal and Mather, 2003; Singh et al., 2007; Tilling et al.,
and easy tool for territorial management, including 2007; Yang et al., 2003). Nutrient contents of different
agricultural analysis, statistics and subsidy control. crops such as wheat (Lelong et al., 1998; Silva and
QuickBird is currently the satellite with the highest Beyl, 2005; Tilling et al., 2007), paddy rice
resolution which is available for agricultural and (Stroppiana et al., 2008), sorghum (Zhao et al., 2005),
civilian uses (Fig. 2). The main requirements of corn (Samson et al., 2000), broccoli (Shikha et al.,
QuickBird as a source of information for agricultural
2007), citrus (Min, 2008), grape (Smart et al., 2007),
and forestry applications is identified in Table 1.
apple (Perry and Davenport, 2007) have also been
New RS multispectral and hyperspectral sensors
are swiftly generating vast amounts of data in a cost- assessed using hyperspectral and multispectral RS data.
effective manner and at higher spatial and spectral Interpretation of RS data is often aided by specialized
resolutions. Hyperspectral and multispectral images, techniques such as geostatistics, image analysis and
consisting of reflectance from the visible, near infrared classification, and artificial intelligence.
and mid-infrared regions of the electromagnetic RS techniques also play an important role in
spectrum, can be interpreted in terms of physical assessing crop condition and yield forecasting, acreage
52
 Am. J. Agri. & Biol. Sci., 5 (1): 50-55, 2010

estimates of specific crops, detection of crop pests and (Aqua and Terra) and EUMETSAT’S MSG-SEVIRI
diseases, disaster location and mapping, wild life satellite sensors were used to characterize fire disaster
management, water supply information and in Swaziland. Combination of RS and GIS techniques
management, weather forecasting, rangeland were used to characterize the geographic and temporal
management, and livestock surveys. (including diurnal) evolution of the July, 2007 fire
Recently, QuickBird imagery and a Production disaster. Significant fire activity was observed during a
Efficiency Model (PEM) were used to estimate crop three-day period beginning July 27, 2007. A total of
yields in Zhonglianchuan, a hilly area on Loess Plateau, 1358 and 4365 active fire hotpots were detected by
China. In the PEM model, crop yields were a function MODIS and MSG SEVIRI, respectively. Results
of the Photosynthetic Active Radiation (PAR), fraction revealed the potential use of RS and GIS for fire
of Absorbed Photosynthetically Active Radiation disaster and risk assessment in a developing country,
(fAPAR) and Light Use Efficiency (LUE). Results where fire monitoring resources are limited (Dlamini,
showed QuickBird imagery can improve the yield 2009). Conversely, Yang et al. (2007) demonstrated
estimation accuracy. The information extracted from that satellite-based RS is a very useful method of
the image was highly correlated to estimated yields forecasting heavy rainfall.
from ground data collection (r2 = 0.86) (Pan et al., RS applications in agriculture have progressed to a
2009). stage where information from RS imagery is being used
Some disease and insect pests of crops may be for a number of policy level decisions related to food
monitored by remote sensing. Riedell et al. (2004) security, poverty alleviation and sustainable
introduced remote sensing technology as an effective development. Decision on buffer stock of food grains
and inexpensive method to identify pest-infested and could be based on pre-harvest crop acreage and
diseased plants. They used remote sensing techniques to production estimates while the ground water potential
detect specific insect pests and to distinguish between maps serve as a major source of information in ensuring
insect and disease damage on oat. Results suggested drinking water and other needs in rain-fed and less
that canopy characteristics and spectral reflectance favorable areas. Nationwide land use, land cover, soil
differences between insect infestation damage and and wasteland mapping have helped in expansion and
disease infection damage can be measured in oat crop intensification of agricultural activities and also in
canopies by remote sensing but that these differences identification of land capability classes and crop
may not be consistent from one growing season to the suitability indices (Venkataratnam, 2001).
next.
Kurtz et al. (2009) used multi-temporal Landsat CONCLUSION
imagery in order to classify land cover types and
grazing intensity. Grazing intensity categories were With increasing population pressure throughout the
defined based on percentage of bare soil, sward height world and the need for increased agricultural
and standing dead material. Correlation analysis production, there is a definite need for improved
between spectral ratio, i.e. Normalized Difference management of the world's agricultural resources. To
Vegetation Index (NDVI), and above ground biomass, make this happen, it is first necessary to obtain reliable
was significant. Meanwhile, Moreau and Toan (2003) data on not only the types of resources, but also the
utilized Synthetic Aperture Radar (SAR) data to quantify quality, quantity and location of these resources.
biomass in an Andean wetland for the purpose of Satellite-or aerial-based RS technologies will become
optimizing livestock management. The signal sensitivity important tools in improving the present system(s) of
corresponding to biomass variation was high enough to acquiring and generating agricultural and natural
facilitate high accuracy biomass mapping. resource data.
In semi-arid Northeast of Brazil, Folhes et al. Agriculture surveys are presently conducted
(2009) employed Landsat imagery in conjunction with throughout the world in order to gather empirical
an evapotranspiration model to measure water use information on crops, rangeland, livestock and other
levels in an irrigated area. Results showed that the agricultural resources. Such information is critical for
combination approach of RS and process modeling effective management of depleting and scarce
produced better predictability of water consumption in resources. Surveys that are based on the PA concept can
irrigated agriculture, and hence improved water facilitate planning and allocation of limited resources to
resource management in irrigated areas. different sectors of the economy. RS technology has the
RS also provide useful information to detect and potential of revolutionizing the detection and
map disaster location. Data from the NASA’S MODIS characterization of agricultural productivity based on
53
 Am. J. Agri. & Biol. Sci., 5 (1): 50-55, 2010

biophysical attributes of crops and/or soils. Essentially, Nowatzki, J., R. Andres and K. Kyllo, 2004. Agricultural
like other PA components, the information gained from remote sensing basics.
RS data is more meaningful when used in combination http://www.ag.ndsu.edu/pubs/ageng/gis/ae1262.pdf
with ground data. Lelong, C.C.D., P.C. Pinet and H. Poilve, 1998.
Although RS cannot capture all types of Hyperspectral imaging and stress mapping in
agricultural information, it can reliably provide accurate
agriculture: A case study on wheat in Beauce
and timely information to guide agronomic and
(France). Remote Sens. Environ., 66: 179-191. DOI:
economic decision-making.
10.1016/S0034-4257(98)00049-2
REFERENCES Min, M., W.S. Lee, T.F. Burks, J.D. Jordan and A.W. Schumann,
2008. Design of a hyperspectral nitrogen sensing
Auernhammer, H., 2001. Precision farming-the system for orange leaves. Comput. Elect. Agric.,
environmental challenge. Comput. Elect. Agric., 30: 63: 215-226. DOI: 10.1016/j.compag.2008.03.004
41-33. DOI: 10.1016/S0168-1699(00)00153-8 Moreau, S. and T.L. Toan, 2003. Biomass quantification
Barnes, E.M. and M.G. Baker, 2000. Multispectral data for of Andean wetland forages using ERS satellite SAR
mapping soil texture: Possibilities and limitations.
data for optimizing livestock management. Remote
Applied Eng. Agric., 16: 731-741.
http://asae.frymulti.com/abstract.asp?aid=5370&t=1 Sens. Environ., 84: 477-492. DOI: 10.1016/S0034-
Barroso, L.A.M., J.G. Payan and E.R. Vivoni, 2008. 4257(02)00111-6
Quantifying water stress on wheat using remote Oza, S.R., S. Panigrahy and J.S. Parihar, 2008. Concurrent
sensing in the Yaqui Valley, Sonora, Mexico. Agric. use of active and passive microwave remote sensing
Water Mgmt., 95: 725-736. DOI: data for monitoring of rice crop. Int. J. Applied Earth
10.1016/j.agwat.2008.01.016 Obst. Geoinform., 10: 296-304. DOI:
Casady, W.W. and H. L. Palm, 2002. Precision
10.1016/j.jag.2007.12.002
agriculture: Remote sensing and ground truthing. MU
Extension, University of Missouri-Columbia. Pal, M. and P.M. Mather, 2003. An assessment of the
http://extension.missouri.edu/explorepdf/envqual/eq0 effectiveness of decision tree methods for land cover
453.pdf classification. Remote Sens. Environ., 86: 554-565.
Dlamini, W.M., 2009. Characterization of the July 2007 DOI: 10.1016/S0034-4257(03)00132-9
Swaziland fire disaster using satellite remote sensing Pan, G., G.J. Sun and F.M. Li, 2009. Using QuickBird
and GIS. Applied Geogr., 29: 299-307. DOI: imagery and a production efficiency model to
10.1016/j.apgeog.2008.10.007
improve. Environ. Model. Software, 24: 510-516.
Folhes, M.T., C.D. Renno and J.V. Soares, 2009. Remote
sensing for irrigation water management in the semi- DOI: 10.1016/j.envsoft.2008.09.014
arid Northeast of Brazil. Agric. Water Mgmt., Perry, E.M. and J.R. Davenport, 2007. Spectral and spatial
96: 1398-1408. DOI: 10.1016/j.agwat.2009.04.021 differences in response of vegetation indices to
Hinzman, L.D., M.E. Bauer and C.S.T. Daughtry, 1986. nitrogen treatments on apple. Comput. Elect. Agric.,
Effects of nitrogen fertilization on growth and 59: 56-65. DOI: 10.1016/j.compag.2007.05.002
reflectance characteristics of winter wheat. Remote Riedell, W.E., S.L. Osborne and L.S. Hesler, 2004. Insect
Sens. Environ., 19: 47-61. DOI: 10.1016/0034- pest and disease detection using remote sensing
4257(86)90040-4
Jensen, J. R., 1996. Remote sensing of the environment: techniques. Proceedings of 7th International
An Earth Resource Perspective. 3th Edn., Prentice Conference on Precision Agriculture. Minneapolis,
Hall, USA, pp: 1-28. MN USA.
Kropff, M.J., J. Wallinga and L.A.P. Lotz, 1997. http://www.ars.usda.gov/research/publications/Public
Modelling for Precision Weed Management. In: ations.htm?seq_no_115=166154
Precision Agriculture: Spatial and Temporal Samson, G., N. Tremblay, A.E. Dudelzak, S.M. Babichenko,
Variability of Environmental Quality, Lake, J.V., L. Dextraze and J. Wollring, 2000. Nutrient stress of
G.R. Bock and J.A. Goode (Eds.). Chichester, Wiley, corn plants: Early detection and discrimination using
pp: 182-204.
a compact multiwavelength fluorescent lidar.
Kurtz, D.B., J. Schellberg and M. Braun, 2009. Ground
and satellite based assessment of rangeland Proceedings of EARSeL-SIG-Workshop LIDAR,
management in sub-tropical Argentina. Applied June 16-17, Dresden/FRG, pp: 214-223.
Geogr. (In Press). DOI: http://www.eproceedings.org/static/vol01_1/01_1_sa
10.1016/j.apgeog.2009.01.006 mson1.pdf
54
 Am. J. Agri. & Biol. Sci., 5 (1): 50-55, 2010

Schellberg, J., M.J. Hill, R. Gerhards, M. Rothmund and Venkataratnam, L., 2001. Remote sensing and GIS in
M. Braun, 2008. Precision agriculture on grassland: agricultural resources management. Proceedings of
Applications, perspectives and constraints. Eur. J. the 1st National Conference on Agro-Informatics,
June 3-4, Dharwad, India, pp: 20-29.
Agron., 29: 59-71. DOI: 10.1016/j.eja.2008.05.005
http://www.insait.org/abstracts.pdf
Shikha, D.M.E., P. Waller, D. Hunsaker, T. Clarke and E. Waheed, T., R.B. Bonnell, S.O. Prasher and E. Paulet,
Barnes, 2007. Ground based remote sensing for 2006. Measuring performance in precision
assessing water and nitrogen status of broccoli. agriculture: CART-A decision tree approach. Agric.
Agric. Water Mgmt., 92: 183-193. DOI: Water Mgmt., 84: 173-185. DOI:
10.1016/j.agwat.2007.05.020 10.1016/j.agwat.2005.12.003
Silva, T.A. and C.A. Beyl, 2005. Changes in spectral Yang, C.C., S.O. Prasher, P. Enright, C. Madramootoo and
reflectance of wheat leaves in response to specific M. Burgess, 2003. Application of decision tree
macronutrient deficiency. Adv. Space Res., technology for image classification using remote
35: 305-317. DOI: 10.1016/j.asr.2004.09.008 sensing data. Agric. Syst., 76: 1101-1117. DOI:
Singh, D., Sao, R. and Singh, K.P. 2007. A remote 10.1016/S0308-521X(02)00051-3
sensing assessment of pest infestation on sorghum. Yang, Y., H. Lin, Z. Guo and J. Jiang, 2007. A data
Adv. Space Res., 39: 155-163. DOI: mining approach for heavy rainfall forecasting based
10.1016/j.asr.2006.02.025 on satellite image sequence analysis. Comput.
Smart, D.R., M.L. Whiting and C. Stockert, 2007. Remote Geosci., 33: 20-30. DOI:
sensing of grape K deficiency symptoms using leaf 10.1016/j.cageo.2006.05.010
level hyperspectral reflectance. Proceedings of Zhang, N., M. Wang and N. Wang, 2002. Precision
Western Nutrient Management Conference, March 6- agriculture: A worldwide overview. Comput. Elect.
8, Salt Lake City, UT., pp: 19-24. Agric., 36: 113-132. DOI: 10.1016/S0168-
http://cropandsoil.oregonstate.edu/sites/default/files/ 1699(02)00096-0
WERA103/2007_Proceedings/WNMC07.p19.Smart.p Zhao, D., K.R. Reddy, V.G. Kakani and V.R. Reddy,
df 2005. Nitrogen deficiency effects on plant growth,
Stroppiana, D., M. Boschetti, P.A. Brivio and S. Bocchi, leaf photosynthesis and hyperspectral reflectance
2008. Plant nitrogen concentration in paddy rice from properties of sorghum. Eur. J. Agron., 22: 391-403.
field canopy hyperspectral radiometry. Field Crop DOI: 10.1016/j.eja.2004.06.005
Res., 111: 119-129. DOI:
10.1016/j.fcr.2008.11.004
Tellaeche, A., X.P. BurgosArtizzu, G. Pajares, A. Ribeiro
and C.F. Quintanilla, 2008. A new vision-based
approach to differential spraying in precision
agriculture. Comput. Elect. Agric., 60: 144-155. DOI:
10.1016/j.compag.2007.07.008
Tilling, A.K., G.J. O’Leary, J.G. Ferwerda, S.D. Jones,
and G.J. Fitzgerald, 2007. Remote sensing of
nitrogen and water stress in wheat. Field Crop Res.,
104: 77-85. DOI: 10.1016/j.fcr.2007.03.023

55