Use of Satellite Data for Watershed Management and

Use of Satellite Data for Watershed Management and Impact Assessment
R S Dwivedi1, K V Ramana1, S P Wani2, and P Pathak2
Over-exploitation of natural resources for meeting the increasing demand for food, fuel, and fiber of the
ever growing population has led to environmental degradation and calls for their optimal utilization based
on their potential and limitations. Information on the nature, extent, and spatial distribution of natural
resources is essential. Spaceborne multispectral measurements made at regular intervals hold immense
potential of providing such information in a timely and cost-effective manner, and facilitate studying
dynamic phenomenon. The geographic information system (GIS) provides an ideal environment for
integration of information on natural resources with the ancillary information for generating derivative
information which is useful in decision making. The study was taken up to generate the action plan for land
and water resources development and to monitor the progress of its implementation in the Adarsha
watershed, Kothapally, Ranga Reddy district, Andhra Pradesh, India. The approach involves generation
of thematic maps on various natural resources through a systematic visual interpretation of satellite data,
integration of such data with the ancillary information and generation of action plan in the GIS
environment, and monitoring vegetation development as a sequel to implementation of action plan by
generating Normalized Difference Vegetation Index (NDVI) from the Indian Remote Sensing Satellite (IRS1C/-1D) Linear Imaging Self-scanning Sensor (LISS-III) data. Soil erosion by water is the major land
degradation process operating in the watershed. There has been an improvement in the vegetation cover
owing to implementation of various soil and water conservation measures, which is reflected in the NDVI
images of pre- and post-implementation periods.
organic matter. In addition, rapid industrialization and
deforestation have led to building up of greenhouse
gases in the atmosphere resulting in global warming.
Degradation of vegetation by deforestation for timber
and fuel wood, shifting cultivation, and occasionally
forest fire is a very serious environmental problem.
Biodiversity conservation is equally important for the
sustainability of vegetation. Optimal utilization of
natural resources based on their limitations and
potential is, therefore, a prerequisite for sustained
agricultural production.
Soil erosion by water and wind is the major land
degradation process in the arid and semi-arid regions
of the world. Globally, about 1.965 billion ha of land
is subjected to some kind of degradation. Of this,
1.094 billion ha of land is subjected to soil erosion by
water and 549 million ha of land to soil erosion by
wind. On an average 25 billion tons of topsoil from
croplands is being washed into oceans. In India alone,
out of 329 million ha geographical area, 150 million
ha land is affected by wind and water erosion (GOI
1976). Annually about 6000 million tons of soil is lost
through soil erosion by water (Das 1985). Also,
shifting cultivation, waterlogging, and salinization
and/or alkalization have affected an estimated 4.36
million ha, 6 million ha, and 7.16 million ha of land
respectively (GOI 1976). Frequent floods and
drought further compound the problem. Soil
degradation contributes to an increase in atmospheric
carbon dioxide through rapid decomposition of
The Role of Remote Sensing
For optimal utilization of natural resources,
information on their nature, extent, and spatial
distribution is a prerequisite. Until the 1920s, such
information had been collected by conventional
surveys, which are labor-intensive, cost-prohibitive,
1. National Remote Sensing Agency (NRSA), Department of Space, Balanagar, Hyderabad, India.
2. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.
indices, which indicate the abundance and condition
of biomass. The index is typically a sum, difference,
ratio, or other linear combination of reflectance factor
or radiance observations from two or more
wavelength intervals. The vegetation indices thus
developed are highly correlated with the vegetation
density or cover; photosynthetically active biomass
(Tucker 1979, Wiegand and Richardson 1984); leaf
area index (Wiegand et al. 1979); green leaf density
(Tucker et al. 1985); photosynthesis rate (Sellers
1987); and amount of photosynthetically active tissue
(Wiegand and Richardson 1987). Landsat-TM data
have been used for deriving various vegetation
indices which in turn were used to assess the impact
of soil conservation measures in the treated
watersheds (NRSA 1996, 1999). The study reported
here was taken up to (i) generate the action plan for
sustainable development of land and water resources,
and (ii) assess the impact of the action plan in the
Adarsha watershed using IRS-1B/-1C and -1D
LISS-II and -III data (see Table 1).
and impractical in the inhospitable terrain. During the
1920s and early 1970s, aerial photographs were used
for deriving information on various natural resources
including lands subject to degradation by various
processes (Bushnell 1929, USDA 1951, Howard
1965, Iyer et al. 1975). Since the launch of the Earth
Resources Technology Satellite (ERTS-1), later
renamed as Landsat-1, in 1972, followed by Landsat2, -3, -4, and -5, SPOT-1, -2, -3, and -4, and the Indian
Remote Sensing Satellites (IRS-1A, -1B, -1C, and
-1D) with Linear Imaging Self-scanning Sensors
(LISS-I, -II, and -III), spaceborne multispectral data
collected in the optical region of the electromagnetic
spectrum have been extensively used in conjunction
with the aerial photographs and other relevant
information supported by ground truth, for deriving
information on geological, geomorphological, and
hydro-geomorphological features (Rao et al. 1996a,
Reddy et al. 1996); soil resources (Singh and Dwivedi
1986); land use/land cover (Landgrebe 1979,
Raghavaswamy et al. 1992, Rao et al. 1996b); forest
resources (Dodge and Bryant 1976, Unni 1992, Roy
et al. 1996); surface water resources (Thiruvengadachari
et al. 1996); and degraded or wastelands (FAO 1978,
Karale et al. 1988, Nagaraja et al. 1992, Dwivedi et
al. 1997a, 1997b). Futhermore, spaceborne multispectral data have been operationally used for
integrated assessment of natural resources and
subsequent generation of action plans for land and
water resources development and for assessment of
the impact of their implementation.
Biomass has been used as a surrogate measure to
evaluate the impact of the implementation of action
plan for land and water resources development. High
absorption of incident sunlight in the visible red
(600–700 nm) portion and strong reflectance in the
near-infrared (750–1350 nm) portion of the electromagnetic spectrum has been used to derive vegetation
Test Site
With an area of 1083 ha, Adarsha watershed in
Kothapally is bound by geo-coordinates 17°21’ to
17°24’ N and 78°5’ to 78°8’ E and forms part of
Shankarpally mandal (an administrative unit) of
Ranga Reddy district, Andhra Pradesh, India.
Vertisols and associated Vertic soils occupy 90% of
the watershed area. However, Alfisols do occur to an
extent of 10% of the watershed area. The main kharif
(rainy season) crops grown are sorghum, maize,
cotton, sunflower, mung bean, and pigeonpea. During
rabi (postrainy season) wheat, rice, sorghum,
sunflower, vegetables, and chickpea are grown. The
mean annual rainfall is about 800 mm, which is
received mainly during June to October.
Table 1. The details of remote sensing data used.
Path/row nos.
Date of pass
02-04-2000 & 29-11-1999
We have used the Indian Remote Sensing Satellite (IRS1B/-1C and -1D) Linear Imaging Self-scanning Sensor
(LISS-II and -III) and Panchromatic sensor (PAN) data
for deriving information on various natural resources
and for generation of action plans for land and water
resources development (Table 1). In addition, Survey of
India topographical maps at 1:50,000 scale, and
published soils and other resources maps and reports
were also used as collateral information.
The methodology involves database preparation,
generation of thematic maps on natural resources, and
their integration with the socioeconomic data to
arrive at a locale-specific prescription for land and
water resources development. The schematic diagram
of the approach is given in Figure 1. Once action plan
is implemented, the next logical step is to assess its
impact on environment and the beneficiaries.
Figure 1. Schematic diagram of the approach.
Database preparation
unit based on erosion status, land use/land cover, and
image elements, namely color, texture, shape, pattern,
and association. Soil composition of each
geomorphic unit was defined by studying soil profiles
in the field and classifying them based on
morphological characteristics and chemical analyses
data (USDA 1975, 1998).
In addition, derivative maps, namely land
capability and land irrigability maps were generated
based on information on soils and terrain conditions
according to criteria from the All India Soil and Land
Use Survey Organization (All India Soil and Land
Use Survey 1970). Land capability classification is an
interpretative grouping of soils mainly based on:
(i) inherent soil characteristics, (ii) external land
features, and (iii) environmental factors. The
groupings enable one to get a picture of (i) the
hazards of the soils to various factors which cause soil
damage and deterioration or lowering in fertility, and
(ii) its potentiality for production. The interpretation
of soil and land conditions for irrigation is concerned
primarily with predicting the behavior of soils under
the greatly altered water regime brought about by the
introduction of irrigation. For arriving at land
irrigability classes, soil characteristics, namely,
effective soil depth, texture of the surface soil,
permeability, water-holding capacity, course
fragments, salinity and/alkalinity, presence of hard
pan in the surface, topography, and surface and subsurface drainage are considered.
Land use/land cover maps have been prepared
using monsoon (kharif) and winter (rabi) crop
growing seasons and summer period satellite data for
delineating single-cropped and double-cropped areas
apart from other land use and land cover categories.
Furthermore, micro-watersheds and water bodies
have been delineated and the drainage networks have
also been mapped. Slope maps showing various slope
categories have been prepared based on contour
information available at 1:50,000 scale topographical
sheets. Rainfall data were analyzed to study the
rainfall distribution pattern in time and space.
Demographic and socioeconomic data were analyzed
to generate information on population density,
literacy status, economic backwardness, and the
availability of basic amenities.
The first step in generating the multi-sensor data sets
is the geo-referencing of the image to a common map
grid. When merging higher resolution data with the
lower resolution images, usually high resolution
image (here PAN data with 5.8 m spatial resolution) is
used as a reference for respective enhancement of the
lower resolution (LISS-III data with 23.5 m spatial
resolution) data (Cliché et al. 1985). To begin with,
the Survey of India topographical maps at 1:50,000
scale were scanned on a Contex FSS-800 scanner at
300 dots per square inch (dpi). The digital LISS-III
was later co-registered to digital, topographic
database on a Silicon Graphics Octane work station
using 20 tie points (ground control point) and imageto-image registration algorithm. The IRS-1D PAN
digital data was subsequently co-registered to LISSIII data following similar approach. Subsequently, the
LISS-III data was resampled to 6 m pixel dimension
using nearest neighborhood algorithms for further
processing. The IRS-LISS-II data was also digitally
co-registered to IRS-1C LISS-III data and resampled
to 24 m pixel dimension. The three bands, namely
0.52–0.59 µm, 0.62–0.68 µm, and 0.77–0.86 µm of
LISS-III data were digitally merged with PAN using
Brovey transformation algorithm. The Brovey
transformation is a formula-based process that works
by dividing the band to display in a given color by the
sum of all the color layers, i.e., red, green, and blue
and then multiplying by the intensity layer.
Generation of thematic maps
Thematic maps on hydrogeomorphological conditions,
soil resources, and present land use/land cover have
been generated through a systematic visual
interpretation of IRS-1B/-1C/-1D LISS-II and -III
data in conjunction with the collateral information in
the form of published maps, reports, wisdom of the
local people, etc. supported by ground truth. The
information derived on the lithology of the area and
geomorphic and structural features in conjunction
with recharge condition and precipitation was used to
infer groundwater potential of each lithological unit.
Soil resource maps of the area have been prepared by
delineating sub-divisions within each geomorphic
Generation of action plan
DN represents digital number in respective spectral
bands. The equation produces NDVI values in the
range of –1.0 to 1.0, where negative values generally
represent clouds, snow, water, and other nonvegetated surfaces, and positive values represent
vegetated surfaces.
The generation of an action plan essentially involves
a careful study of thematic maps on land and water
resources, both individually as well as in
combination, to identify various land and water
resources regions or Composite Land Development
Units (CLDU) and their spatial distribution, potential
and limitations for sustained agriculture and other
uses, and development of an integration key. It was
achieved by scanning the thematic maps on a
CONTEX FSS 800 black and white scanner at 400
dpi. It was followed by vectorization, projection to
real world coordinates, editing map compilation and
unionizing the thematic boundaries in a geographic
information system (GIS) domain using ARC/INFO
version-7 software. Each CLDU was studied
carefully and a specific land use and soil and water
conservation practice was suggested based on its
sustainability. Subsequently, taking landform as a
base an integration key in terms of potential/
limitations of soils, present land use/land cover, and
groundwater potential, and suggested alternate land
use/action plan was developed.
Natural resources
Lithologically, the watershed comprises of basalt and
laterites. The moderately dissected plateau which is
interspersed with structural valleys constitutes the
major landform. While the undissected plateau has
poor groundwater potential, the dissected plateau has
poor to moderate potential. Structural valley has good
groundwater potential depending on the nature of the
fracture. Whereas Vertic Haplaquepts have developed
over structural valleys the dissected plateau support
the development of shallow soils namely Lithic
Ustochrepts and Lithic Ustorthents. Vertic
Ustochrepts, however, do occur in local depressions
within the dissected plateau. The watershed is mainly
used for raising both kharif and rabi crops. A few
pockets of land, however, is wasteland mostly in the
form of land with/without scrub. The land under kharif
crops constitute the major land use and land cover
category followed by double cropped land (Fig. 2).
Implementation of action plan
The action plan and/alternate land use practices and
drought-proofing activities emerging from this approach
have been implemented by the district/mandal
authorities using the state-of-the-art technology for
each action item to fully exploit the contemporary
developments in agriculture, science, and technology.
Action plan
Since the watershed very often experiences drought,
apart from alternate land use based on potential and
limitations of natural resources, various droughtproofing measures such as vegetative barriers,
contour bunding, stone check-dams, irrigation water
management, horticulture, groundwater development
with conservation measures, and fodder and
silvipasture in marginal lands have been undertaken.
The suggested optimal land use practices are
intensive agriculture, intercropping system, improved
land configuration, agro-horticulture, horticulture
with groundwater development, and silvipasture.
Impact assessment
Since vegetation condition is the reflection of soils
and hydrological conditions which are altered in the
event of implementation of suggested action plan, it
has been taken as a surrogate parameter for
assessment of the impact of such treatment in the
watershed. The Normalized Difference Vegetation
Index (NDVI) values from near infrared (NIR) and
red (R) band responses in the IRS-1B/-1C/-1D LISSII and -III data were generated on a Silicon Graphics
work station as follows:
NDVI (output DN) =
Implementation of action plan
Various soil and water and conservation measures,
e.g., broad-bed and furrows, contour planting,
NIR (DN) – R (DN)
NIR (DN) + R (DN)
Figure 2. LISS-III and PAN merged image and land use map of Adarsha watershed, Kothapally, India.
waterways and drainage channels, field bunding,
wasteland development, storage of excess water
through construction of check-dams, dug out ponds,
gabion structures, gully plugging, and increased
cropping intensity have been undertaken in the
watershed. In addition, integrated nutrient and pest
management trials have also been conducted.
groundwater level, well density and yield, cropping
pattern and crop yield, occurrence of hazards, and
socioeconomic conditions. Land use/land cover
parameters include: changes in the number and aerial
extent of surface water bodies, spatial extent of forest
and other plantations, wastelands, and cropped area.
As mentioned earlier, NDVI has been used to
monitor the impact of the implementation of action
plan. A close look at the NDVI images of 1996 and
2000 reveals an increase in the vegetation cover
which is reflected in improvement in the vegetation
cover (Fig. 3). The changes in the vegetation cover
can be seen in the satellite image as variations in the
red-colored patches, and in the NDVI images as
changes in yellow and pink colors. The spatial extent
of moderately dense vegetation cover which was 129
ha in 1996 has risen to 152 ha in 2000. Though the
satellite data used in the study depicts the terrain
Impact assessment
Soon after implementation of the suggested action
plan, the area undergoes transformation, which is
monitored regularly. Such an exercise not only helps
in studying the impact of the program, but also
enables resorting to mid-course corrections, if
required. Parameters included under monitoring
activities are land use/land cover, extent of irrigated
area, vegetation density and condition, fluctuation of
conditions during 1996, implementation activities
started only in 1998. It is, therefore, obvious that it
will take considerable time for detectable changes in
the terrain and vegetation conditions.
We would like to place on record our sincere thanks to
Dr D P Rao, Director, National Remote Sensing
Agency (NRSA), Department of Space, Government
of India, Hyderabad, India for evincing keen interest,
providing necessary guidance and facilities, and
active involvement in this study. Thanks are also due
to Prof. S K Bhan, Deputy Director (Applications),
NRSA for his timely and kind advice at various stages
of the study. Thematic and action plan maps of the
watershed provided by Dr A Perumal, Head,
Integrated Survey Division, NRSA have been quite
useful in finalizing the manuscript.
The study vividly demonstrates the potential of
spaceborne multispectral data in deriving information
on natural resources. The GIS provides an ideal
environment for integration of data on natural
resources with the ancillary information and
facilitates generation of action plan for development
of land and water resources. After implementation of
action plan, multi-temporal satellite data help in
monitoring its success and progress. The change in
vegetation cover in the Adarsha watershed as a result
of adopting soil conservation measures during 1996
to 2000, is an indicator of the success of
implementation of such action plans. High spatial
resolution panchromatic and multispectral data from
IKONOS-II and the future earth observation missions
such as Resourcesat-1, Cartosat-1 and -2, Quick Bird,
Almaz-1B, etc. may further enhance our capability of
generating farm-level action plan for land and water
resources development, and to study the success and
progress of the implementation of such action plans.
All India Soil and Land Use Survey. 1970. Soil survey
manual. New Delhi, India: All India Soil and Land Use
Survey, Ministry of Agriculture, Government of India.
Bushnell, T.M. 1929. Aerial photographs of Jennigs
county. Proceedings of Indiana Academy of Science
Cliché, G., Bonn, F., and Teillet, P.M. 1985. Integration of
the SPOT panchromatic channel into its multispectral
mode for image sharpness. Photogrammetric Engineering
and Remote Sensing 51(3):311–316.
Das, D.C. 1985. Problem of soil erosion and land
degradation in India. Presented at the National Seminar on
Figure 3. Satellite and NDVI images of Adarsha watershed, Kothapply, India.
Soil Conservation and Watershed Management, 17–18 Sep
1985, New Delhi, India.
NRSA (National Remote Sensing Agency). 1999.
Monitoring the impact of soil conservation measures using
remote sensing. Project Report. Hyderabad, India: NRSA.
Dodge, A.G., and Bryant, E.S. 1976. Forest type mapping
with satellite data. Journal of Forestry 74(8):526–531.
Raghavaswamy, V., Gautam, N.C., Rao, D.P., and
Nagaraja, R. 1992. Land use/land cover mapping and
monitoring urban sprawl using IRS data - A few examples.
Pages 120–127 in Natural resources management: A new
perspective (Karale, R.L., ed.). Bangalore, India: National
Natural Resources Management System (NNRMS),
Department of Space.
Dwivedi, R.S., Kumar, A.B., and Tiwari, K.K. 1997a.
The utility of multi-sensor data for mapping eroded lands.
International Journal of Remote Sensing 18(11):2303–
Dwivedi, R.S., Ravi Sankar, T., Venkataratnam, L.,
Karale, R.L., Gawande, S.P., Rao, K.V.S.,
Senchaudhary, Bhaumik, K.R., and Mukharjee, K.K.
1997b. The inventory and monitoring of eroded lands
using remote sensing data. International Journal of Remote
Sensing 18(1):107–119.
Rao, D.P., Bhattacharya, A., and Reddy, P.R. 1996a. Use
of IRS-1C data for geological studies. Current Science
(Special Section: IRS-1C) 70(7):619–623.
Rao, D.P., Gautam, N.C., Nagaraja, R., Ram Mohan, P.
1996b. IRS-1C applications in land use mapping and
planning. Current Science (Special Section: IRS-1C)
FAO (Food and Agriculture Organization of the United
Nations). 1978. The application of Landsat imagery soil
degradation mapping at 1:50000 scale. FAO Summary
Report. Rome, Italy: FAO.
Reddy, P.R., Vinod Kumar, K., and Seshadri, K. 1996.
Use of IRS-1C data in ground water studies. Current
Science (Special Section: IRS-1C) 70(7):600–605.
GOI (Government of India). 1976. Report of the National
Commission on Agriculture. Parts V, IX, and Abridged
Report, Ministry of Agriculture and Irrigation. New Delhi,
India: GOI.
Roy, P.S., Dutt, C.B.S., Jadhav, R.N., Rangnath, B.K.,
Murthy, M.S.R., Gharai, B., Lakshmi, V.U., Kandya,
A.K., and Thakker, P.S. 1996. IRS-1C data utilization for
forestry applications. Current Science (Special Section:
IRS-1C) 70(7):606–613.
Howard, J.A. 1965. Small scale photographs and land
resources in Nyamweziland, East Africa. Photogrammetric
Engineering 32:287–293.
Sellers, P.J. 1987. Canopy reflectance, photosynthesis and
transpiration: The role of biophysics in the linearity of their
interdependence. Remote Sensing of Environment
Iyer, H.S., Singh, A.N., and Ram Kumar. 1975. Problem
area inventory of parts of Hoshiarpur district through
photo-interpretation. Journal of the Indian Society of
Remote Sensing 11:79–85.
Singh, A.N., and Dwivedi, R.S. 1986. The utility of
Landsat-MSS as an integral part of database for small scale
soil mapping. International Journal of Remote Sensing
Karale, R.L., Saini, K.M., and Narula, K.K. 1988.
Mapping and monitoring ravines using remotely-sensed
data. Journal of Soil Water Conservation, India 32:75.
Thiruvengadachari, S., Jonna, S., Hakeem, K.A., Raju,
P.V., Bhanumurthy, V., Rao, G.S., Paul, P.R., Shankar,
E.S., and Jayaseelan, A.T. 1996. Improved water
management: IRS-1C contribution. Current Science
(Special Section: IRS-1C) 70(7):589–599.
Landgrebe, D. 1979. Monitoring the earth’s resources
from space – Can you really identify crops by satellites?
Presented at the National Computer Conference, New York
City, USA, June 7, 1979.
Nagaraja, R., Gautam, N.C., Rao, D.P., and
Raghavaswamy, V. 1992. Application of Indian Remote
Sensing Satellite (IRS) data for wasteland mapping. Pages
170–174 in Natural resources management: A new
perspective (Karale, R.L., ed.). Bangalore, India: National
Natural Resources Management System (NNRMS),
Department of Space.
Tucker, C.J., Townshend, J.R.G., and Goff, T.E.. 1985.
African land cover classification using satellite data.
Science 227:369–375.
NRSA (National Remote Sensing Agency). 1996.
Evaluation study of treated sub-watersheds in Kundah,
Lower Bhawani and Tungabhadra catchments using
remote sensing techniques. Project Report. Hyderabad,
India: NRSA.
Unni, N.V.M. 1992. Forestry and ecology applications of
IRS-1A data. Pages 108–119 in Natural resources
management: A new perspective (Karale, R.L., ed.).
Management System (NNRMS), Department of Space.
Tucker, C.J. 1979. Red and photographic infrared linear
combinations for monitoring vegetation. Remote Sensing
of Environment 8:127–150.
USDA (United States Department of Agriculture). 1951.
Soil survey manual. Agriculture Hand Book No. 18.
Washington, DC, USA: USDA.
Wiegand, C.L., and Richardson, A.J. 1984. Leaf area,
light interception, and yield estimates from spectral
component analysis. Agronomy Journal 82:623–629.
Wiegand, C.L., and Richardson, A.J. 1987. Spectral
component analysis. Rationale and results for three crops.
International Journal of Remote Sensing 8:1011–1032.
USDA (United States Department of Agriculture). 1975.
Soil taxonomy - A basic system of soil classification for
making and interpreting soil surveys. Agriculture Hand
Book No. 436. Washington, DC, USA: USDA.
Wiegand, C.L., Richardson, A.J., and Kanemasu, E.T.
1979. Leaf area index estimates for wheat from
LANDSAST and their implications for evapotranspiration
and crop modeling. Agronomy Journal 71:336–342.
USDA (United States Department of Agriculture). 1998.
Key to soil taxonomy. Washington, DC, USA: USDA.