Use of remote sensing for generation of agricultural statistics

27: Use of Remote Sensing …

USE OF REMOTE SENSING FOR GENERATION OF
AGRICULTURAL STATISTICS
Prachi Misra Sahoo
Indian Agricultural Statistics Research Institute, New Delhi-110012

1. INTRODUCTION

Agricultural statistics plays an important role in national planning and judicious allocation of
limited resources to the different sectors of economy in the country. The domain of
agricultural statistics was initially based on two parameters the area and production
estimation of principal crops. The former is obtained through complete enumeration whereas
the latter through sample surveys. Subsequently, it was extended to the estimation of
livestock, fisheries, fruits, vegetables and other important products related to agriculture.
Apart from this, studies were carried out for estimation of cost of production/rearing, impact
of developmental projects in the field of agriculture and rural development. Most of the
methodologies developed for collecting agriculture statistics are based on stratified multistage
sampling design keeping in view the administrative setup and infrastructural facilities
available in the country.

In due course of time, many changes have taken place at national as well as at global level.
These are mainly due to changes in the level of technology, government policies and
structure of the population. Consequently, there are changes in data need and requirement.
Keeping in view the changing scenario, it has been realized that there is a need for reappraisal
of the methodologies developed for agricultural surveys. The recent technological
developments in the computer and space technology have shifted the emphasis of survey
research work towards newer emerging areas of Remote Sensing technology and
Geographical Information System (GIS). These advances in technology have given a new
direction to the survey research for generation of agricultural statistics. This article briefly
describes some of the important applications and perspectives of remote sensing for
generation of agricultural statistics.

2. GLOBAL SCENARIO

The use of space borne remote sensing data for large area crop survey was explored in USA
under Corn Blight Watch Experiment (CBWE) in 1971. In this experiment, first time a sound
statistical design was used for large-scale remote sensing program (NASA, 1974). An
experiment named Crop Identification Technology Assessment for Remote Sensing
(CITARS) was started in 1973 to quantify the Crop Identification Performance (CIP) with
several Automatic Data Processing (ADP) classifications using remote sensing (1973). An
attempt was made to forecast wheat crop production for major wheat growing regions of the
world under Large Area Crop Inventory Experiment (LACIE) during 1974-1977 (NASA,
1977). Later, a six-year programme of research and development named Agriculture and
Resource Inventory Survey Through Aerospace Remote Sensing (AGRISTARS) was taken
up in 1988. Since then, large-scale methodology development-cum-demonstration studies for
crop statistics have been carried out in Africa and Europe as well as in a number of other

276


countries like Argentina, Australia, Brazil, Canada, Japan etc. Currently, major programmes
are underway in Africa under Global Information and Early Warning System (GIEWS) and
in Europe under Monitoring Agriculture through Remote Sensing (MARS). The MARS
project has developed rapid crop survey procedure for Crop Growth and Monitoring System
(CGMS), which employ crop simulation models, agro-meteorological models, and real time
data for crop forecasting and assessment.

Remotely sensed satellite imagery was used for monitoring crop production, for estimating
losses due to drought in France in 1989, 1990 and 1991, and to monitor acreages and
potential crop yields throughout Europe (Lepoutre, 1991). A pilot remote sensing project
directed by the Joint Research Centre of the Commission of European Communities was
taken up with the objective to test remote sensing applications for agricultural statistics,
including differentiation, identification and measurement of the surface area of major crops,
and estimating crop production in five European countries (Meyer, 1991).

3. INDIAN PERSPECTIVE FOR GENERATION OF AGRICULTURAL
STATISTICS

In a country like India, with vast geographic spread and great diversity in its set up, the need
to apply remote sensing technology for national development was recognized early. The
pioneering experiment was of coconut root–wilt disease using colour-infrared aerial
photography (Dakshinamurti et al., 1971). Subsequently, major experiments using multi-band
aerial photography were carried out at space Applications Center, Ahmedabad for the survey
of agricultural resources under Agriculture Resources Inventory & Survey Experiment
(ARISE) in Anantpur (1974-75) and Patiala (1975-76); identification and classification of
paddy and sugarcane crops in Madhya Pradesh (1975-77), land use and forest inventory
survey of Panvhmahals district , Gujrat and Idukki district, Kerala (1976-1979) in
association with ICAR and other user organization. These experiments primarily aimed at the
development of methodologies and demonstration of the potential of remote sensing in
natural resources survey. Simultaneously, efforts were made for the indigenous development
of an airborne thermal scanner. Organizations such as Geological Survey of India, All India
Soil & Land Use Survey, National Bureau of Soil Survey and Land Use Planning, National
Geophysical Research Institute, etc., were using aerial photography / measurements in their
day to day work and had established interpretation facilities to cater to their needs.

In late seventies, National Remote Sensing Agencies (NRSA) was established in 1975
primarily for providing operational aerial survey services and carried out a number of
projects using aerial photographic data, airborne MSS and Landsat data particularly for the
states of Mizoram, Tripura and parts of Tamil Nadu, etc. Establishment of Landsat data
reception center at NRSA, Hyderabad in 1979 for receiving Landsat data acted as a catalyst
way and various central and state government departments started using the data as one of the
sources of the information .The reception center since then has been extensively used to
collect and disseminate data to Indian users. Paralarlly, considerable efforts went into
spacecraft system and their associated ground system, data products and required hardware
and software for application at various ISRO centres. Joint Experiment Programme (JEP)
taken up in 1978, in association with department of Agriculture & Co-operation, ICAR, and
Ministry of Steel and Mines as one such effort aimed at development of Indian Remote

277


Sensing Satellite. Extensive aerial data, photographic as well as that from ground experiments
were used to arrive at sensor specification for the Indian Remote Sensing Satellite (IRS).

Launch of two experimental earth observation satellites Bhaskara I and II (1979, 1981)
provided the necessary experience in handling a total remote sensing system on an
experimental level. These satellites carried a two band television camera operating in 0.54-
0.66 pm and 0.75-85 pm and a microwave radiometer acronymic SAMIR (Satellite
Microwave Radiometer). During this period, basic science studies to understand the spectral
behavior of natural objects such as crops, water, rock types and soils, etc., under various
environmental conditions were also carried out. This was made possible by the development
of an indigenous ground truth radiometer (GTR) having capability to operate in selected
spectral bands. These studies contributed in a significant way towards the definition of
Indian Remote Sensing programme as well. With the successful completion of Bhaskara
programme the capability to build operational satellites for remote sensing was well
established and this in conjunction with the experience gained through JEP laid the
foundation for the Indian Remote Sensing Satellite Programme.

Early eighties witnessed a spurt in the use of satellite data (Landsat) for various resource
applications in the country. End-to-end experiments were carried out to demonstrate the
capabilities of remote sensing in various spheres of natural resources management. A major
step was taken in 1982 when at the instance of PC–NNRMS (Preparatory Committee–
National Natural Resources Management System) fifty nine well defined experiments were
conducted to demonstrate the end utilization of remote sensing in various application areas.
Ground water targeting and mineral exploration were among the most striking examples of
these end to end experiments. To expand the scope of remote sensing data utilization, many
states established State Remote Sensing Application Centres to provide remote sensing inputs
for aiding the planning process under the overall umbrella of NNRMS. The Uttar Pradesh
Government set up first state remote sensing applications center at Lucknow.

In the era of late eighties many national level projects were undertaken jointly with state
remote sensing centres and other agencies. A nation-wide study was carried out during this
period pertained to wasteland mapping on 1:1M scale by NRSA. The study provided baseline
information on spatial distribution of wastelands and acted as a precursor detailed wasteland
mapping in the country, the information on which was critical for taking ameliorative steps
and bringing additional areas under tree plantation and agriculture. Similar study for land use
mapping for the entire country was also carried out during this period for facilitating agro-
climatic regional planning. Pilot studies were initiated during this decade for forecasting
crops using digital satellite data. To begin with, these studies were undertaken for Karnal
(Haryana) and Patiala (Punjab) for wheat, and Cuttack (Orissa) and Midnapore (W. Bengal)
for the rice. Efforts were also started to develop crop yield models based on RS and
metrological parameters. Another important area where remote sensing data was used on an
operational scale during late eighties was the forecasting of Potential Fishing Zone (PFZ)
using NOAA AVHRR data.

Towards the end of eighties, remote sensing activities in India received a tremendous boost
with the launch of Indian Remote Sensing Satellite-1A (IRS 1A) in March 1988. This carried
two cameras LISS-I and LISS-II (spatial resolution of 72.5 m and 36.25 m respectively) in

278


identical four spectral bands providing repetitively of 22 days. Realizing the utility of remote
sensing data in a variety of application areas various Ministries of Govt. Of India and State
Govt. Departments started making increasing use of remote sensing derived information in
agricultural management process. By this time remote sensing centres were also established
in majority of the states. Crop Acreage and Production Estimation (CAPE) project which was
undertaken for a few districts/states for wheat and rice in the eighties was extended to more
crops in their major producing areas in the nineties at the request of the Department of
Agriculture and Cooperation.

Ninties witnessed launch of a series of IRS satellites (IRS 1B/P2 and IRS IC/ID, IRS P3, IRS
P4 – Oceansat) by ISRO and simultaneous operationalization of many application area.
Realizing the importance of RS technology various user ministries such as Ministry of
Agriculture (CAPE), Ministry of Environment and Forests (Environmental Impact
Assessment, Nation-wide Wetland Mapping. Coastal Zone Studies, etc.), Department of
Ocean Development, etc. sponsored national level projects Local Agencies such as Urban
Development Authority, Municipal Corporations, etc., and NGO’s also started making use of
remotely sensed data in their planning schemes.

Providing national level multiple wheat production forecasts using multi date WiFS data has
been very much appreciated by the end user in this period. RADARSAT data is being
investigated to provide national level kharif rice production forecasts. Realizing that remote
sensing can not be a stand alone system for crop production forecasts, the concept of
Forecasting Agricultural output using Space, Agro-meteorology and Land based observation
(FASAL) has been evolved. To execute this project, Department of Agriculture &
Cooperation is establishing National Crop Forecasting Center (NCFC). Studies related to
Soil, Water management, Land-use planning have also been carried on using satellite data in
this period. Use of satellite data in disaster management is gaining momentum.

Some of the important applications of remote sensing in management of agriculture are
discussed briefly in following sections.

4. APPLICATIONS FOR GENERATION OF AGRICULTURAL STATISTICS
USING REMOTE SENSING

Crop production statistics are of vital importance to a country such as India, where the
agricultural production is highly susceptible to the vagaries of monsoon. These statistics
consist of two major components: (i) acreage under the crop and (ii) crop yield per unit area.
The traditional approach of crop estimation in India involves a complete enumeration (except
a few states where sample surveys are employed) for estimating crop acreages and the yield
surveys based on crop cutting experiments for estimating crop yield. The crop production
estimates are obtained by taking product of crop acreage estimates and the corresponding
crop yield estimates.

279


4.1 Crop Acreage Estimation

Importance of crop production information was realized in India as early as 1884, when the
government initiated such a programme for wheat crop. Currently, the use of remotely sensed
(RS) data is being investigated for crop production forecasting all over the world. The
intrinsic ability of spectral reflectance data to identify crops and distinguish one from another
is very helpful in estimating crop acreages. Acreage estimation using RS data has been
demonstrated in various parts of the world (Mac Donald 1984, Renondo et al., 1985). A
number of studies have been carried out on remote sensing based acreage estimation in India.
Earlier investigations with spectral data used visual interpretation technique. Single date
Landsat MSS digital data and supervised classification approach for wheat acreage estimation
of Karnal district was used by Dadhwal and Parihar (1985). Larger study areas (a group of
district or a state) were later taken up for study for which sample segment approach
consisting of 10x10 km segment, 10% sampling fraction and stratified sampling was
suggested by Dadhwal and Sridhar (1986). Studies on rice acreage estimation using remote
sensing satellite data have been taken up since 1986-87 for the state of Orissa (Panigrahi et.al.
1991). Work carried out so far in India has demonstrated that even with single-date satellite
data it is possible to estimate pre-harvest acreages of major crops, particularly in single crop
dominated regions, with sufficient accuracy. Rai et. al., (2004) have carried a study on land
use statistics through integrated modeling using GIS. Misra et.al (2005) developed an
integrated approach for estimation of crop acreage using remote sensing data, GIS and field
survey for hilly region. A spatial sampling procedure incorporating spatial dependence of
neighbouring units has been proposed by Misra et. al (2006).

4.2 Crop Yield Estimation:

Various yield models have been discussed in literature to establish relationship between
yields and weather, soil or biometrical characters of the plants, several studies have been
taken up to establish the relationships between spectral reflectance and the crop yield. The
basic approach in the past has been to develop a transformation of the multi-band spectral
response as a measure of vegetation vigor and relate it to some agronomic quantity such as
leaf area index, wet or dry biomass or grain yield. Several vegetation indices have been
developed and shown to be well correlated with these agronomic variables.

Crop yield estimation surveys based on crop cutting experiments are conducted throughout
the country for obtaining precise estimates of average yield for all major crops. Sukhatme and
Panse (1951) gave the estimation procedure of estimating average yield and crop production
based on crop cutting experiments under general crop yield estimation surveys. Use of
satellite data along with survey data of crop yield from General Crop Estimation Surveys
(GCES) based on crop cutting experiments for obtaining improved estimators of crop yield
has been undertaken at IASRI since 1990 (Singh et. al. 1992). Singh and Goyal (1993) have
used spectral vegetation indices like Normalized Difference Vegetation Index (NDVI) to
obtain improved crop yield estimators. Singh et. al. (2000 a) have developed post-stratified
estimators of crop yield using spectral data in the form of vegetation indices for stratification
of cropped area. Global Positioning System (GPS) has been used for collecting the data for
identification of crop plots of the survey data. Singh et al. (2002) have given small area
estimates of crop yield at Tehsil/Block level. Two small area estimators namely Direct and

280


Synthetic estimators have been developed from general crop yield estimation surveys on crop
cutting experiments. Rai et. al (2007) developed spatial models for crop yield estimation.
This procedure is also extended for estimation of crop yield at small area levels.

4.3 Crop Yield Modeling

Reliable and timely forecast of crop production is of crucial economic importance. The
advent of remote sensing technology during seventies provided an immense potential to
improve upon the existing pre-harvest forecasting models. The fact that spectral reflectance
data can be timely available for the entire crop growth period with almost equal accuracy can
be effectively utilized in the development of better yields forecasting models.

Many studies have been carried out to relate plant biometrical characters and spectral
parameters. By relating the reflectance data of individual crops, in specific wavelength
regions to canopy growth or vigor, it is possible to predict yield estimates using appropriate
modeling techniques backed up by adequate ground truths. The availability of regular space
observations has simulated the development of computerized agriculture information systems
in several countries. Development of reliable crop-yield models with minimal data has
become a major thrust area in our country also. Studies related to use of data from different
spectral regions, generation of models using growth-profile parameters and incorporation of
agro-meteorological information in the simple yield models have been carried on widely.

Over the past decade, a number of crop yield forecasting models using remote sensing inputs
have been developed and used in making forecasts. These include (a) Single date RS-based
models (b) Spectral profile related growth parameters derived from multi-date data (c)
Combinations of different parameters like trend, RS and meteorological parameters either by
including all in a multiple linear regression equation or by optimal combination of different
estimates. Singh and Ibrahim (1996) examined the use of multi-date satellite spectral data for
crop yield modeling using Markov chain model. Saha (1999) used satellite data and GIS for
developing several yield models for forecasting. Singh et. al. (2000 b) have given spectral
models and integrated models using spectral data and farmers eye estimate for forecasting
crop yield. Besides this, microwave data, owing to its all weather capabilities, have also been
used for crop forecasting.

Recently on the recommendation of an Expert Group under the chairmanship of Director
IASRI, the Ministry of Agriculture has established a National Center for Forecasting of Crops
(NCFC) and a major project on ' Forecasting Agricultural output using Space, Agro
meteorology and Land based observation' (FASAL) is under way to develop methodology to
obtain accurate and timely crop forecasts.

5. REMOTE SENSING APPLICATION FOR INTEGRATED RESOURCE
MANAGEMENT FOR SUSTAINABLE AGRICULTURE PRODUCTION

Adoption of appropriate strategies for achieving integrated sustainable development of land
and water resources is the only answer to improve agricultural productivity in a rational

281


manner. The integrated sustainable development is defined as growth oriented development
to meet the needs of the present as well as the needs of the future generations without causing
any degradation to the ecology and environment. Sustainable agricultural production could be
achieved only through an understanding of the mutual interdependencies of land and water
resources (both renewable and non-renewable) and identification of the constraints/ecological
problems at the micro level.

The synoptic view provided by satellite remote sensing offers a technologically appropriate
method for integrating the land and water resources information and for identifying agro-
climatically coherent zones. Once such zones are identified, locale specific prescriptions
could be arrived at through the effective use of space based remote sensing data merged with
other collateral socio-economic data by use of geographic information systems.

It is in this context the Indian experience of using satellite remote sensing for integrated
resources development at micro level become relevant. It involves stock taking of land and
water resources through a series of surveys, carried out in phases, using a combination of
conventional and remote sensing techniques. The first phase consists of collection of
conventional data and their evaluation. The second stage involves preparation of a set of
resource maps using remotely sensed data on (i) surface water bodies (ii) ground water
potential zones, (iii) potential zones for ground water recharge, (iv) existing land use and
distribution of wastelands and (vi) an integrated land and water resource map giving high
priority areas for development of agriculture, fuel and fodder, soil conservation and
afforestation. The final stage is to develop a package of appropriate strategies to address the
local resource management and environment problems.

6. REMOTE SENSING APPLICATION FOR PRECISION FARMING

Precision farming aims to improve crop performance and environmental quality. It is defined
as the application of technologies and principles to manage spatial and temporal variability
associated with all aspects of agricultural production. In other words, precision farming is the
matching of resource application and agronomic practices with soil attributes and crop
requirements as they vary across a field. Thus, the concepts of precision farming include: (i)
Variations occur in crop or soil properties within a field (ii) These variations are noted, and
often mapped (iii) Management actions are taken as a consequence of the spatial variability
within the field.

Though, the 20th century agriculture had been characterized by the increase in land and labor
productivity, the use of external inputs, an increase in efficiency and efficacy of external
inputs, it has also been associated with the stimulation of uniformity in agricultural
production areas and the negative side-effects of agriculture. This techniques, by appreciating
the variability within the field and adopting management practices to cater the variability, are
serving the dual purpose of enhancing productivity and reducing ecological degradation. The
real value from precision farming is that the farmer can perform more timely tillage, adjust
seeding rates, fertilizer application according to soil conditions, plan more crop protection
programs with more precision, and know the yield variation within a field. These benefits can
enhance the overall cost effectiveness of crop production. Many technological developments,

282


which occurred in 20th century contributed to the development of the concept of precision
farming which includes GPS, GIS and high resolution remote sensing satellite data.

REFERENCES

Dadhwal, V.K. and Parihar, J.S. (1985). Estimation of 1983-84 wheat acreage of Karnal
(Haryana) using Landsat MSS digital data. Scientific Note, IRS-UP/SAC/CPF/
SN/9/85, SAC, Ahmedabad.

Dadhwal, V.K. and Sridhar,V.N. (1986). Sampling approach for remote sensing based crop
inventory. Scientific Note, RSAM/SAC/CAPE/SN/01/86.

Dakshinamurti, C., Krishnamurthy, B., Summanwar, A. S., Shanta, P. and Pisharoty, P.R.
(1971). Remote Sensing for coconut wilt. Proc. Seventh Int. Symp on Remote Sensing
of Envirn. May 17-21, 1971. 25-29.

Lepoutre, D. (1991). Satellite imagery for following crop production. Comptes-Rendus-de-
l'Academie-d'Agriculture-de-France, 77(6), 75-84.

Mac Donald R. B. (1984) A summary of the history of the development of automated remote
sensing of automated remote sensing for agricultural applications. IEEE Trans. Geosci.
Rem Sens. 22, 473-481.

Meyer, R.J.(1991). Pilot project on remote sensing applied to agricultural statistics in Europe.
In Options-Mediterranean’s.- Serie -A,-Seminaires - Mediterranean. 4: 57-63.

Sahoo, P. M., Singh, R. and Rai, A. (2006): Spatial sampling procedures for agricultural
surveys using geographical information system. J. Ind. Soc. Agril. Stat. 60 (2)134-143.

Sahoo, P. M., Singh, R. and Rai, A., Handique B.K. and Rao C. S. (2005). Integrated
approach based on remote sensing and GIS for estimation of area under paddy crop in
north-eastern hilly region’ J. Ind. Soc. Agril. Stat. 59(2), 151-160.

NASA, Johnson Space Centre, (1974). Corn blight watch experiment: Summary report, V.3
NASASP-353, NTIS, Springfield, VA.

NASA, Johnson Space Centre, (1977). LACIE: wheat yield models for the United States:
NASAJSC-0043, rev A NTIS, Springfield, VA.

Panigraphy,S., Parihar, J.S. , Patel, N.K. , Dadhwal, V.K. , Medhavy, T.T. , Ghose, B.K. ,
Ravi, N. , Pant, K.C. , Panigrahy, B.K. , Sridhar, V.N., Mohanty, R.R. , Nanda, S.K. ,
Tripathy, D.P. , Mishra, P.K. , Bhatt, H.P., Oza, S.R. , Sudhakar, S. , Sudha,K.S. ,
Kumar, P. and Das, N.K. (1991). “Rice acreage estimation for Orissa using Remotely
Sensed data”. J. Soc. of Rem. Sens. I9(1), 17-26.

Rai A., N.K. Gupta and Randhir Singh (2007) Small area estimation of crop production
using spatial models. Mod. Asst. Stat. Apl. , 2(2), 89-98,

Rai Anil, Srivastava A.K., Singh R, and Jain V.K. (2004) A study of land use statistics
through integrated modelling using geographic information system. IASRI, New Delhi
Publication.

283


Redondo F.V., Lacprogent C., Gargantini C., Anter M. and Fonda C. (1985). Estimating
wheat cultivated area within large productivity region in Argentina using Landsat data.
Proc. 19th Int. Symp. Rem. Sens. Environ. 361-367.

Saha, S. K. (1999) Crop yield modeling using satellite remote sensing and GIS- current status
and future prospects. Proc. Geoinformatics- Beyond 2000. Int. Conf on Geoinformatics
for natural resource assessment, monitoring and management. Dehradun. 9 -11 March,
1999.

Singh, R., Goyal, R.C. Saha, S.K. and Chhikara, R.S. (1992) Use of satellite spectral data in
crop yield estimation surveys. Int. J. Rem. Sens. 13(14), 2583-2592.

Singh, R. and Goyal, R.C. (1993) Use of remote sensing technology in crop yield estimation
surveys. Project Report, IASRI, New Delhi.

Singh, R. and Ibrahim, AEI (1996) Use of spectral data in markov chain model for crop yield
forecasting. J. Ind. Soc. Rem Sens. 24(3), 145-152

Singh, R. Semwal, D.P., Rai, A. and Chhikara, R.S. (2002) Small area estimation of crop
yield using remote sensing satellite data. Int. J. Rem. Sens. 23(1): 49-56.

Singh, R., Semwal, D.P., Rai, A. and Chhikara, R.S. (2000a) Small area estimation of crop
yield using remote sensing satellite data. Accepted for publication in Int. J. Rem. Sens.

Singh, R. and Goyal, R.C. (2000 b) Use of remote sensing technology in crop yield
estimation surveys. Project Report, IASRI, New Delhi.

Sukhatme. P.V. and Panse, V.G. (1951). Crop surveys in India-II. J. Ind. Soc. Agrl. Stat. 2,
95-168.

284

Use of remote sensing for generation of agricultural statistics

More Related Content

What's hot (19)

Similar to Use of remote sensing for generation of agricultural statistics (20)

Use of remote sensing for generation of agricultural statistics