ArticlePDF Available

Abstract and Figures

Across South Asia, the cost of rice cultivation has increased due to labor shortage. Direct seeding of rice is widely promoted in order to reduce labor demand during crop establishment stage, and to benefit poor farmers. To facilitate planning and to track farming practice changes, this study presents techniques to spatially distinguish between direct seeded and transplanted rice fields using multiple-sensor remote sensing imagery. The District of Raichur, a major region in northeast Karnataka, Central India, where irrigated rice is grown and direct seeded rice has been widely promoted since 2000, was selected as a case study. The extent of cropland was mapped using Landsat-8, Moderate Resolution Imaging Spectroradiometer (MODIS) 16-day normalized difference vegetation index (NDVI) time-series data and the cultivation practice delineated using RISAT-1 data for the year 2014. Areas grown to rice were mapped based on the length of the growing period detected using spectral characteristics and intensive field observations. The high resolution imagery of Landsat-8 was useful to classify the rice growing areas. The accuracy of land-use/landcover (LULC) classes varied from 84 percent to 98 percent. The results clearly demonstrated the usefulness of multiple-sensor imagery from MOD09Q1, Landsat-8, and RISAT-1 in mapping the rice area and practices accurately, routinely, and consistently. The low cost of imagery backed by ground survey, as demonstrated in this paper, can also be used across rice growing countries to identify different rice systems.
Content may be subject to copyright.
Mapping Direct Seeded Rice in
Raichur District of Karnataka, India
Murail Krishna Gumma, Deepika Uppala, Irshad A. Mohammed, Anthony M. Whitbread, and Ismail Ra Mohammed
Abstract
Across South Asia, the cost of rice cultivation has increased
due to labor shortage. Direct seeding of rice is widely pro-
moted in order to reduce labor demand during crop establish-
ment stage, and to benefit poor farmers. To facilitate planning
and to track farming practice changes, this study presents
techniques to spatially distinguish between direct seeded and
transplanted rice fields using multiple-sensor remote sensing
imagery. The District of Raichur, a major region in northeast
Karnataka, Central India, where irrigated rice is grown and
direct seeded rice has been widely promoted since 2000, was
selected as a case study. The extent of cropland was mapped
using Landsat-8, Moderate Resolution Imaging Spectroradi-
ometer (MODIS) 16-day normalized difference vegetation index
(NDVI) time-series data and the cultivation practice delineated
using RISAT-1 data for the year 2014. Areas grown to rice were
mapped based on the length of the growing period detected
using spectral characteristics and intensive field observa-
tions. The high resolution imagery of Landsat-8 was useful to
classify the rice growing areas. The accuracy of land-use/land-
cover (LULC) classes varied from 84 percent to 98 percent. The
results clearly demonstrated the usefulness of multiple-sensor
imagery from MOD09Q1, Landsat-8, and RISAT-1 in mapping the
rice area and practices accurately, routinely, and consistently.
The low cost of imagery backed by ground survey, as dem-
onstrated in this paper, can also be used across rice growing
countries to identify different rice systems.
Introduction
Agriculture is an important sector in India, contributing about
17.9 percent of the gross domestic product (GDP) (2014 figures)
(CIA, 2015). About 68 percent of the country’s population lives
in rural areas (FAOSTAT, 2013). India is one of the largest rice
growing nations accounting for one quarter of the global rice
area and produces around 125 million tons/year, with low
yields of around 2.85 t/ha (Siddiq, 2000). Given that rice is a
staple food crop in India and that the country has a popula-
tion that is growing at 1.2 percent per annum, a substantial
increase in rice production is the only way to guarantee food
security and to keep food prices low. Labor and fertilizer are
the two major input costs that are incurred while growing rice
when there is assured water availability (Savary et al., 2005).
Migration from rural to urban areas in search of work has led
to labor shortages and higher costs of agricultural operations
(Hossain, 1998; Savary et al., 2005). To overcome these con-
straints, promoting direct seeding in lieu of the traditional and
labor-intensive transplanting (De Datta, 1986; Naylor, 1992) is
being seen as a feasible alternative. However, there is a trad-
eoff with lower yields in rice that is direct seeded rather than
transplanted; so efforts are on to bridge this gap (Khush, 1995).
Accurate and timely information on rice systems, cropped
area, and yields are very important for sustainable food secu-
rity. Several studies have used remote sensing data to monitor
rice crop. Remote sensing technology provides a time saving
approach to estimate cropped area, intensity and other LULC
changes in a country (Badhwar, 1984; Gumma et al., 2011a;
Gumma et al., 2015; Lobell et al., 2003; Thenkabail et al.,
2009a; Thenkabail, 2010; Thiruvengadachari and Sakthiva-
divel, 1997). Many studies have reported the use of spatial-
temporal data to map irrigated areas, land-use, land-cover,
and crop type (Dheeravath et al., 2010; Goetz et al., 2004;
Gumma et al., 2014; Knight et al., 2006; Thenkabail et al.,
2005; Varlyguin et al., 2001; Velpuri et al., 2009) using MODIS
NDVI time-series data to map both agricultural areas (Biggs et
al., 2006; Gaur et al., 2008; Gumma et al., 2011c) and seasonal
crop area (Sakamoto et al., 2005). There are also studies that
have used Synthetic Aperture Radar (SAR) data to monitor
rice areas, particularly irrigated rice. (Le Toan et al., 1997)
used ERS-1 data as input to crop growth simulation models to
estimate production for study sites in Indonesia and Japan.
Similarly, (Shao et al., 2001) mapped rice areas using tempo-
ral RADARSAT data (1996 and 1997) for production estimates
in Zhaoqing in China. (Imhoff and Gesch, 1990) derived
sub-canopy digital terrain models of flooded forest using SAR.
(Kasischke and Bourgeau-Chavez, 1997) monitored the wet-
lands of South Florida using ERS-1 SAR imagery. (Leckie, 1990)
distinguished forest type using SAR and optical data. (Robin-
son et al., 2000) delineated drainage flow directions using SAR
data. (Townsend, 2001) mapped seasonal flooding in forested
wetlands using temporal RADARSAT SAR imagery. (Bouvet and
Le Toan, 2011) demonstrated how lower resolution wide-
swath images from advanced SAR (ASAR) data could be used to
map rice over larger areas. (Uppala et al., 2015) mapped rice
areas using single data hybrid polarimetric data from RISAT-1
SAR data. The advantage of rice mapping with SAR is that it
can overcome pervasive cloud cover across Asia during the
rainy season months when rice is cultivated. However, its
high cost inhibits its large-scale application.
A comprehensive plan to increase agricultural productivity
and manage labor can be drawn up based on near real-time
monitoring of rice systems along with weather and crop data
acquired from an analysis of remote sensing imagery. Remote
sensing platforms have proved to be ideal to take resource
inventories and monitor agriculture. Interdisciplinary ap-
proaches and methods have made it easier to analyze imagery
and extract information in ways unknown a few decades ago.
The goal of this study was to develop a method using mul-
tiple-sensor imagery (Landsat-8, RISAT-1 and MODIS time-series
data) to map the spatial extent of two different rice cultiva-
tion practices (direct seeding and transplanting) in Raichur
district, Karnataka State, India.
International Crops Research Institute for the Semi-Arid
Tropics (ICRISAT), Patancheru, Telangana, India (m.gumma@
cgiar.org).
Photogrammetric Engineering & Remote Sensing
Vol. 81, No. 11, November 2015, pp. 873–880.
0099-1112/15/873–880
© 2015 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.81.11.873
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
November 2015 873
11-15 November Peer Reviewed.indd 873 10/21/2015 1:18:30 PM
Study Area and Data Sets
Study Area
Raichur is one of largest and most productive rice growing
district in Karnataka state in southern India (Figure 1). The
district lies between 15°42'57" and 16°26'39" latitude and
76°14'47" and 77°35'39" longitude, with a land mass of 835,843
ha (cropped area 695,000 ha) (Figure 1) and bound in the
north and south by the rivers Krishna and Tungabhadra. Agro-
climatically, it falls under the Northeast Dry Zone of Karnataka
(semi-arid eco-subregion) with a long term average rainfall of
just above 600 mm distributed over kharif and rabi seasons.
Much of the district is well irrigated by the Tungabhadra Dam
on the Tungabhadra River, and the Narayanpura Dam on the
Krishna River, covering an irrigated area of 185,000 ha, 72.2
percent of which is from canals. The arable rainfed area occu-
pies 405,000 ha. The major crops grown in the district include
paddy, sorghum, groundnut, sunflower, cotton, and pulses
(chickpea and pigeonpea). The rice-rice cropping system is
dominant in most irrigated areas. The soils of the district are
mainly deep black clayey (47 percent) and red (34 percent).
Satellite Images
Four Landsat-8 tiles were downloaded from the Earth Ex-
plorer, global land cover facility website (http://earthexplorer.
usgs.gov/). Images from the 2014 monsoon season and their
spectral characteristics are shown in Table 1. All the Landsat-8
tiles were converted into reflectance (http://landsat.usgs.gov/
documents/Landsat8DataUsersHandbook.pdf) to normalize
the multi-date effect (Markham and Barker, 1986; Thenkabail
et al., 2004) using the spatial modeler in ERDAS Imagine® (ER-
DAS, 2007). The 250-m, two-band MODIS data (centered at 648
nm and 858 nm; Table 1) collection 5 (MOD09Q1) were acquired
for every eight days during the crop-growing seasons from
January 2014 through December 2014. The data was acquired
in 12-bit (0 to 4,096 levels) and stretched to 16-bit (0 to 65,536
levels). Further processing steps are described in (Gumma et
al., 2015; Gumma et al., 2011b; Thenkabail et al., 2005).
RISAT-1, C-band (5.35 GHz) HH polarization, Medium Resolu-
tion SAR (MRS) temporal data sets were used to distinguish
direct seeded rice from transplanted rice. Data was acquired
every 25 days at a 37° angle of incidence spaced phenologi-
cally appropriate to identify rice fields and provide informa-
tion on data sets (Table 1).
Ground Survey Data
A ground survey was conducted during the monsoon season
for the crop year 2014 to 2015. Data was collected at 238 loca-
tions covering the major cropland areas in the study area. All
location-specific data were collected from 250 × 250 meter
plots and consisted of location coordinates (latitude, longi-
tude), land-use categories, land-use (percent), cropping systems
during the monsoon season (through farmer interviews), crop
types, and watering method (irrigated/rainfed). Samples were
within large contiguous areas of a specific land-use/land-cover.
The locations were chosen based on pre-classification classes
and local expertise. The local experts also provided information
on cropping systems, cropping patterns and cropping inten-
sity (single or double crop), irrigation application, and canopy
cover ( percent) for the previous years for these locations from
their recorded data. Overall, 116 spatially well-distributed data
points (Figure 1) were used for class identification and labeling;
of these, 46 data points were used for ideal spectra generation
and an additional 122 were used to assess accuracy.
Figure 1. Raichur District, Karnataka, India.
Table 1. CharaCTeriSTiCS of MulTiPle SaTelliTe SenSor DaTa uSeD in The STuDy
Sensor Dates Spatial
(m) Bands Band range
(µm)
Landsat-8 07 & 14
Nov 2014 30
1 0.433–0.453
2 0.450–0.515
3 0.525–0.600
4 0.630–0.680
5 0.845–0.885
6 1.560–1.660
7 2.100–2.300
RISAT-1
28 Jun 2014;
23 Jul 2014;
17 Aug 2014;
11Sep 2014;
06 Oct 2014
18 1 HH polarization
MODIS
(MOD09Q1)
2014
(16 days) 250
1 0.62–0.67
2 0.84–0.88
NDVI - 1 to + 1
874 November 2015
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
11-15 November Peer Reviewed.indd 874 10/21/2015 1:18:30 PM
Methods
Figure 2 presents an overview of the comprehensive meth-
odology used to map direct seeded and transplanted rice
cultivation practices using time series imagery from MOD09Q1
(spatial resolution 250 m) 16-day time-series NDVI, Landsat-8
(spatial resolution 30 m), and RISAT-1 (spatial resolution 18 m).
Rice growing areas were delineated initially with MODIS time-
series NDVI using spectral matching techniques (Gumma et
al., 2011b; Gumma et al., 2014; Thenkabail et al., 2007a). This
was used as a mask to clip Landsat-8 time-series imagery. The
subset of Landsat-8 was reclassified to separate rice from non-
rice areas due to the difference in resolution between MODIS
and Landsat-8. The Landsat-8 output provided a more ac-
curate delineation of rice growing area. This subset was again
used as a mask to clip out RISAT-1 temporal imagery.
Land-Use / Land-Cover Classication
The procedure began with image normalization of Landsat-8
data converted to top of atmosphere (TOA) reflectance using a
reflectance model implemented in ERDAS Imagine (http://land-
sat.usgs.gov/documents/Landsat8DataUsersHandbook.pdf).
The (Operational Land Imager) OLI band data can be con-
verted to TOA planetary reflectance using Reflectance rescaling
coefficients provided in the product metadata file. The follow-
ing equation was used to convert DN values to TOA planetary
reflectance for OLI data:
ρλ = Mρ Qcal + Aρ (1)
where: ρλ = TOA planetary reflectance (without correction of
solar angle), Mρ = Band specific multiplicative rescaling factor
from the metadata, Aρ = Band specific additive rescaling factor
from the metadata, and Qcal = the quantized and calibrated
standard product pixel values (DN).
TOA reflectance with correction for the sun angle is then:
ρλ =
ρλ
θ
'
sin(
)
SE
(2)
where: ρλ = TOA planetary reflectance, ρλ = TOA planetary
reflectance (without correction of solar angle), and θSE = the
local sun elevation angle provided in the metadata.
The MODIS stacked composite was classified using unsuper-
vised ISOCLASS cluster K-means classification algorithm fol-
lowed by successive generalization (Biggs et al., 2006; Gumma
et al., 2011c; Thenkabail et al., 2005). The unsupervised
classification algorithm (in ERDAS Imagine 2010) was applied
on a 12-band NDVI (monthly Maximum Value Composite)
MVC to obtain the initial 100 classes, followed by progressive
generalization (Cihlar et al., 1998). The unsupervised classifi-
cation was set at a maximum of 100 iterations with a conver-
gence threshold of 0.99 (Leica, 2010). Time-series NDVI spectra
were then plotted for each of the 100 classes and compared
with the ideal spectra to identify and label classes (Gumma et
al., 2014). However, the time-series NDVI profile helps gain an
understanding of the growth profile of different crops in addi-
tion to providing information on planting date, discrimination
between rice and other crops, early stage conditions (flooded
pixel showing low values initially), and discrimination be-
tween irrigation sources (e.g., irrigated versus rainfed). Class
identification and labeling were performed based on a suite of
methods and ancillary data, such as decision tree algorithms,
spectral matching techniques, Google Earth high-resolution
imagery and ground survey data (Gumma et al., 2014; Thenka-
bail et al., 2009b; Thenkabail et al., 2007b). The initial reduc-
tion in classes used a decision tree method (De Fries et al.,
1998) based on the temporal NDVI data. The decision tree is
based on NDVI thresholds at different stages in the season that
define vegetation growth cycle, and these algorithms help to
identify similar classes. The dates and threshold values were
derived from the ideal temporal profile (Gumma et al., 2014).
Using the ground survey data, Google Earth’s high-resolution
imagery along with spectral profiles of rice crops from MODIS
imagery, Landsat-8 imagery was classified using the super-
vised maximum likelihood classification algorithm.
The MODIS-derived rice area was used as the basis of the
maximum possible extent of rice area as one segment and
other LULC areas as another segment. This was used as a mask
to clip Landsat-8 time-series imagery. The subset of Landsat-8
was classified again to separate rice from non-rice areas. Both
segments were classified independently (to avoid mixed clas-
sification) using the protocols mentioned earlier. This led to
a more accurate delineation of rice growing area, but did not
show fragmented direct seeded rice areas. This was mainly
because direct seeded rice was sown in the early monsoon,
when there were no images due to heavy clouds. Landsat-8
rice area was again used as a mask to clip out RISAT-1 temporal
imagery. Separating the two practices of rice cultivation was
possible using RISAT-1 temporal imagery. This is the best com-
plimentary data to monitor croplands during the monsoon
season, and where there are continuous clouds.
Figure 2. Overview of the methodology for mapping different rice growing practices.
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
November 2015 875
11-15 November Peer Reviewed.indd 875 10/21/2015 1:18:36 PM
Delineating Direct Seeded Rice (DSR) Cultivation Using SAR Data
The RISAT-1 data was processed using ENVI software. The lev-
el-2, HH polarization data was imported in ENVI raster format
(.hdr). To reduce noise, Enhanced Lee Adaptive Speckle Filter
with kernel size of 5 × 5 was applied (Lopes et al., 1990).
Radiometric calibration for HH polarization amplitude image
was carried out by using the metadata provided along with the
imagery. The scaled values of digital numbers were converted
into backscattering coefficient using the equation given in
(Chakraborty et al., 2013; Laur et al., 2002). The temporal im-
ages were co-registered by fitting a second order polynomial
model to the manual GCPs with 50 well distributed points
throughout the scene. The temporal HH polarization RISAT-1
data was stacked. The rice crop delineated from Landsat-8 was
used to subset the RISAT-1 data stack for further processing.
In Raichur, direct seeded rice cropping is practiced along with
transplanted rice which is most prevalent. Transplanted rice is
of two types depending on sowing date. The early transplanted
rice is sown in August and the late transplanted rice in Septem-
ber. SAR data is unique since it can distinguish between different
rice sowing dates. The rice crop generates a unique temporal
backscatter profile. It is clear from Figure 3 that during the estab-
lishment stage of transplanted rice, the backscatter is higher than
direct seeded rice due to tilling and low moisture. During trans-
plantation, backscatter is quite low due to standing water which
reflects minimal energy in the backscatter direction. Further,
the backscatter increases as the plant grows because of multiple
reflections from the crop canopy (Choudhury et al., 2012).
In the establishment phase of direct seeded rice, low
backscatter was observed due to moisture conditions. This
is the key to distinguishing 80 percent of the direct seeded
rice from transplanted fields. Also, some of the transplanted
rice fields cannot be distinguished from direct seeded fields
due to high standard deviation in transplanted fields because
of roughness. Additionally, early transplanted rice could be
distinguished at the transplantation stage due to very low
backscatter. However, it is difficult to discriminate between
direct seeded rice and late transplantation due to their similar
growth stages. The deviation of backscatter was found to be
low. The decision tree hybrid classification algorithm was
used here in order to separate the above mixed classes. June
is the crucial month to separate transplanted rice form direct
seeded rice (80 percent). The early transplanted rice was eas-
ily separated from directed seeded rice by applying decision
rules on the August imagery. This was possible because the
difference in backscatter was more than 2 dB. The late trans-
plantation was not separable from direct seeded rice because
of the low difference in backscatter. In order to resolve these
types of fields, unsupervised k-means algorithm with a con-
vergence threshold of 0.99 was applied on the remaining area
Figure 3. Temporal changes in backscatter coefcient in different
rice growing practices.
Plate 1. Land-use/land-cover during the 2014 monsoon season (kharif) in Raichur District, with rice classes temporal NDVI proles.
876 November 2015
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
11-15 November Peer Reviewed.indd 876 10/21/2015 1:18:39 PM
which included late transplanted rice and direct seeded rice.
By using this algorithm, 85 percent of late transplanted rice
fields were separated from direct seeded rice fields.
Results
Spatial Distribution of Land Use / Land Cover
Ten land-use/land-cover classes were identified using spectral
matching techniques using MODIS temporal imagery and Land-
sat-8 (Plate 1). With average rainfall in the order of 600 mm/
annum and large areas under rainfed cropping, canal irriga-
tion from Tungabhadra Dam provides insufficient irrigation
to grow rice. The total irrigated area accounts for 287,993 ha
(Table 2). The important rice growing taluks (administrative
units) are Manvi and Sindhanur where a large area is irri-
gated under rice. The adoption of direct rice seeding is due to
shortage of labor, and it is a new phenomenon and expanding
along with transplanted rice cultivation which is essentially
large in extent. The use of temporal RISAT-1 imagery helped
in identifying how rice was planted (direct seeded rice or
transplanted) and also ascertaining time of planting (early or
late transplanting). This information can be used to schedule
irrigation and estimate harvest dates for each type of practice.
It is clear that the head end of the canal has a larger extent
of transplanted rice than the tail end in Manvi and Sindha-
nur with irrigated mixed crops. Rainfed croplands occupy
the largest area of 351,895 ha in Raichur District. Pigeonpea
mixed with a cereal crop like sorghum is one of the important
and large cropping systems in the taluks of Lingsugur and
Raichur. Rangelands and shrublands are comparably large
which can be potentially used for rainfed agriculture.
Spatial Extent of Direct Seeded Rice and Transplanted Rice
Rice is a major crop in Raichur District, covering 130,343 ha in
2013, out of which transplanted rice covered 117,401 ha and
direct seeded rice 12,942 ha. The major rice area is located in
Manvi and Sindhanur taluks which are irrigated, depending on
ground water availability and rainfall. Of the total rice area, 52
percent is early transplanted rice, 38 percent is late transplant-
ed rice, and only 10 percent is direct seeded rice (Table 2 and
Plate 1). Variations within the direct seeded field are difficult
to identify due to the temporal deficit of the image acquisition.
Accuracy Assessment
The classification accuracies obtained from the 122 indepen-
dent ground data observation data points for the year 2014
were gathered from the study area (monsoon season). Accuracy
assessment was done using error matrix, a multi-dimensional
table in which the cells contain changes from one land-use
class to another. The statistical approach of accuracy assess-
ment consists of different multi-variate statistical analyses
(Congalton and Green, 1999; Jensen, 1996). The error matrix ac-
curacy varied from 75 percent to 92 percent across six cropland
areas. However, it must be noted that rice classes did not mix
with other classes. Overall, accuracy was 82 percent and kappa
was 0.78. So the uncertainty of about 18 percent is due to the
inter-mix among the other land-use/land-cover classes (Table 3).
Comparison with National Statistics and Other Studies
Statistics on taluk-wise rice area in Raichur District was ob-
tained from the Bureau of Economics and Statistics, Govern-
ment of Karnataka State, India. This data was used for com-
paring RISAT-1, Landsat-8, and MODIS-derived statistics. A high
correlation was found (Figure 4) between the two sources of
information on rice cultivation in the district.
Table 2. eSTiMaTeD area unDer DifferenT lanD uSe/lanD CoVer (lulC)
ClaSSeS in raiChur DiSTriCT
LULC area Area (ha) %
01. Irrigated-conjunctive-DS-rice 12,942 1.5
02. Irrigated-conjunctive-early TP-rice 67,418 8.0
03. Irrigated-SW-late TP-rice 49,983 5.9
04. Irrigated-conjunctive-SC-mixed crops 157,650 18.7
05. Rainfed-SC-pigeonpea/mixed crops 122,778 14.6
06. Rainfed-mixed crops/fallows 229,117 27.2
07. Rangelands/shrublands 152,442 18.1
08. Barren lands/shrublands/forests 37,060 4.4
09. Water bodies 9,241 1.1
10. Settlements 3,695 0.4
Total 842,326
Table 3. aCCuraCy aSSeSSMenT uSing grounD SurVey DaTa (Two DiMenSional error MaTrix)
Classified data
Reference data
Reference row total
Classified totals
Producers accuracy
Users accuracy
Kappa accuracy
01. Irrigated-conjuctive-
DS-rice
02. Irrigated-conjunctive-
early TP-rice
03. Irrigated-SW-late
TP-rice
04. Irrigated-conjunctive-
SC-mix
05. Rainfed-SC-
pigeonpea / mixed crops
06. Rainfed-mixed crops/
fallows
07. Other LULC
01. Irrigated-conjuctive-DS-rice 3 0 1 0 0 0 0 3 4 100% 75% 74%
02. Irrigated-conjuctive-early TP-rice 0 12 1 0 0 0 0 13 13 92% 92% 91%
03. Irrigated-SW-late TP-rice 0 0 7 1 0 0 0 9 8 78% 88% 87%
04. Irrigated-conjuctive-SC-mix 0 0 0 22 2 1 0 24 25 92% 88% 85%
05. Rainfed-SC-pigeonpea/mixed crops 0 0 0 1 19 2 1 27 23 70% 83% 78%
06. Rainfed-mixed crops/fallows 0 1 0 0 4 30 1 37 36 81% 83% 76%
07. Other LULC 0 0 0 0 2 4 7 9 13 78% 54% 50%
Total 3 13 9 24 27 37 9 122 122 84% 80% 77%
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
November 2015 877
11-15 November Peer Reviewed.indd 877 10/21/2015 1:18:39 PM
Discussion
Microwave remote sensing satellites carrying Synthetic Aper-
ture RADAR (SAR) enable imaging of the features of the Earth’s
surface both during the day and the night under all weather
conditions. These specific characteristics of the satellite are
useful in identifying crop information under cloudy skies
using standard methods. However, the objective stated in
this study demands a hybrid approach where optical and SAR
imagery are used to distinguish between the two rice growing
practices that exhibit similar backscatter mechanisms, except
at the establishment stage. Studies have been conducted to
come up with a robust classification method using X-band
SAR data to map rice crops, including for direct seeding in
different ecosystems and also to examine the relationship be-
tween crop characteristics and backscatter coefficient (Nelson
et al., 2014). However, the inherent property of C-band SAR
data, where the wavelength responds to roughness of soil by
including any background (soil backscatter), makes a differ-
ence by identifying a practice like direct seeded rice. The
basic premise that direct seeded rice requires less labor and
less water, making it economically less demanding option
and also a strategy to cope with monsoon failure, has some
lacunae. This is because the direct seeded rice fields are more
infested by weeds than transplanted rice fields (Rao et al.,
2007) as the puddled fields do not allow weed growth. This
leads to yield losses depending on weed infestation.
RISAT-1 SAR imagery was a viable and useful option as it
overcomes cloud cover as well as provides a convenient
temporal coverage to distinguish between these two practices
right at establishment stage. Mixed classes were resolved
based on extensive ground information collected during field
trips and tools like Google Earth Explorer. A multi-sensor
approach was a useful strategy to fulfill the objective of this
study, especially to arrive at accurate estimates of the prac-
tice of growing direct seeded rice in Raichur District. The
MODIS, 16-day NDVI data was primarily used to delineate the
rice cropped areas taking advantage of its temporal availabil-
ity. The spectral profiles from MODIS temporal data provide
important information on different aspects of the crop, such
as its source of water, phenological stage, and stress. The role
of Landsat-8 imagery was to generate a high accuracy rice
cropped area map to match the resolution of RISAT-1 imagery
and in turn transfer information from MODIS to RISAT-1.
Karnataka State’s sub-district divisions are large and few.
For instance, Raichur has only five taluks. Statistics gener-
ated from this study on taluk-wise rice cropped area were
compared with those from the government department. A
high correlation was observed between these two sources of
information (Figure 4). Statistics on direct seeded rice was not
available from the government departments since it is a new
practice in this district. However, this study was successful
in identifying direct seeded rice because of the difference in
crop establishment dates compared to the other practice of
transplanting rice.
Conclusions
Mapping the cropped area of direct seeded rice using a multi-
sensor approach, specifically the use of RISAT-1 imagery, is
important as it can penetrate cloud cover during the rainy
season. Optical remote sensing methods may not be success-
ful in discriminating land-use classes based on management
practices. Hybrid approaches have always proved useful in
atypical cases such as this study. Even though there are three
types of direct seeded rice cultivation, an attempt was made
to map only the dry seeded sowing practice prevalent in
Raichur District. Sindhanur and Manvi are the two taluks in
the district where the practice is followed. Using the temporal
SAR data along with optical data such as MODIS, a similar
approach can be used to map direct seeded rice cultivation
during post-rainy (rabi) season as well. The spatial distribu-
tion and quantification of the extent of direct seeded rice
will not only help breeders to understand the locale-specific
constraints to yields such as weeds, but also provide remedies
to bridge the yield gaps. In turn, this will help in overcoming
the labor shortage and also help in conserving water for other
uses. In this context, it is proposed to study and examine how
we can differentiate between rice establishment practices not
only at the time of establishment using SAR imagery, but also
when there is a large outbreak of weeds in direct seeded rice
cultivation.
Acknowledgments
This research was supported by the CGIAR Research Program
Water, Land and Ecosystems. The authors would like to thank
Smitha Sitaraman and Amit Chakravarty, science editors/pub-
lisher at ICRISAT for his assistance with editing. The authors
also thank Dr. AN Rao (IRRI) and Dr. Gajanan Sawargaonkar
for their valuable feedback of the rice classification system
and sub-district wise statistics.
References
Badhwar, G.D., 1984. Automatic corn-soybean classification using
landsat MSS data - I. Near-harvest crop proportion estimation,
Remote Sensing of Environment, 14(1-3):15–29.
Biggs, T.W., P.S. Thenkabail, M.K. Gumma, C.A. Scott, G.R.
Parthasaradhi, and H.N. Turral, 2006. Irrigated area mapping in
heterogeneous landscapes with MODIS time series, ground truth
and census data, Krishna Basin, India, International Journal of
Remote Sensing, 27(19):4245–4266.
Bouvet, A., and T. Le Toan, 2011. Use of ENVISAT/ASAR wide-swath
data for timely rice fields mapping in the Mekong River Delta,
Remote Sensing of Environment, 115(4):1090–1101.
Chakraborty, M., S. Panigrahy, A. Rajawat, R. Kumar, T. Murthy, D.
Haldar, A. Chakraborty, T.
Kumar, S. Rode, and H. Kumar, 2013. Initial results using RISAT-1
C-band SAR data, Current Science (Bangalore), 104(4):490–501.
Choudhury, I., M. Chakraborty, S.C. Santra, and J.S. Parihar, 2012.
Methodology to classify rice cultural types based on water
regimes using multi-temporal RADARSAT-1 data, International
Journal of Remote Sensing, 33(13):4135–4160.
CIA, 2015. The World Factbook, URL: https://www.cia.gov/library/
publications/the-world-factbook/fields/2012.html (last date
accessed: 24 September 2015).
Cihlar, J., Q. Xiao, J. Beaubien, K. Fung, and R. Latifovic, 1998.
Classification by progressive generalization: A new automated
methodology for remote sensing multichannel data, International
Journal of Remote Sensing, 19(14):2685–2704.
Congalton, R.G., and K. Green, 1999. Assessing the Accuracy of
Remotely Sensed Data: Principles and Practices, New York:
Lewis.
De Datta, S., 1986. Technology development and the spread of direct-
seeded flooded rice in Southeast Asia, Experimental Agriculture,
22(04):417–426.
De Fries, R.S., M. Hansen, J.R.G. Townshend, amd R. Sohlberg, 1998.
Global land cover classifications at 8 km spatial resolution: The
use of training data derived from Landsat imagery in decision
tree classifiers, International Journal of Remote Sensing,
19(16):3141–3168.
Dheeravath, V., P.S. Thenkabail, G. Chandrakantha, P. Noojipady,
G.P.O. Reddy, C.M. Biradar, M.K. Gumma, and M. Velpuri, 2010.
Irrigated areas of India derived using MODIS 500 m time series
for the years 2001-2003, ISPRS Journal of Photogrammetry and
Remote Sensing, 65(1):42–59.
ERDAS, 2007. ERDAS Field Guide, Volume 1, October 2007.
878 November 2015
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
11-15 November Peer Reviewed.indd 878 10/21/2015 1:18:39 PM
FAOSTAT, 2013. URL: http://faostat.fao.org/site/666/default.aspx (last
date accessed: 24 September 2015).
Gaur, A., T.W. Biggs, M.K. Gumma, and H. Turral, 2008. Water
scarcity effects on equitable water distribution and land use
in Major Irrigation Project - A Case study in India, Journal of
Irrigation and Drainage Engineering, 134 (1):26–35.
Goetz, S.J., D. Varlyguin, A.J. Smith, R.K. Wright, S.D. Prince,
M.E. Mazzacato, J. Tringe, C. Jantz, and B. Melchoir, 2004.
Application of multitemporal Landsat data to map and monitor
land cover and land use change in the Chesapeake Bay
watershed, Analysis of Multi-temporal Remote Sensing Images
(P. Smits and L. Bruzzone, editors), World Scientific Publishers,
Singapore, pp. 223–232.
Gumma, M.K., S. Mohanty, A. Nelson, R. Arnel, I.A. Mohammed,
and S.R. Das, 2015. Remote sensing based change analysis of
rice environments in Odisha, India, Journal of Environmental
Management, 148(0):31–41.
Gumma, M.K., P.S. Thenkabail, A. Maunahan, S. Islam, and A.
Nelson, 2014. Mapping seasonal rice cropland extent and area
in the high cropping intensity environment of Bangladesh
using MODIS 500m data for the year 2010, ISPRS Journal of
Photogrammetry and Remote Sensing, 91(5):98–113.
Gumma, M.K., D. Gauchan, A. Nelson, S. Pandey, and A. Rala, 2011a.
Temporal changes in rice-growing area and their impact on
livelihood over a decade: A case study of Nepal, Agriculture,
Ecosystems & Environment, 142(3-4):382–392.
Gumma, M.K., A. Nelson, P.S. Thenkabail, and A.N. Singh, 2011b.
Mapping rice areas of South Asia using MODIS multitemporal
data, Journal of Applied Remote Sensing, 5, 053547.
Gumma, M.K., P.S. Thenkabail, I.V. Muralikrishna, M.N. Velpuri, P.T.
Gangadhararao, V. Dheeravath, C.M. Biradar, S. Acharya Nalan,
and A. Gaur, 2011c. Changes in agricultural cropland areas
between a water-surplus year and a water-deficit year impacting
food security, determined using MODIS 250 m time-series data
and spectral matching techniques, in the Krishna River basin
(India), International Journal of Remote Sensing, 32(12):3495–
3520.
Hossain, M., 1998. Sustaining Food Security in Asia: Economic,
Social, and Political Aspects - Sustainability of Rice in the
Global Food System, Pacific Basin Study Center, Davis,
California, and International Rice Research Institute, Manila,
The Philippines, pp. 19–44.
Imhoff, M.L., and D.B. Gesch, 1990. The derivation of a sub-canopy
digital terrain model of a flooded forest using synthetic aperture
radar, Photogrammetric Engineering & Remote Sensing,
56(8):1155–1162.
Jensen, J.R., 1996. Introductory Digital Image Processing: A Remote
Sensing Perspective, Prentice Hall, Upper Saddle River, New
Jersey.
Kasischke, E.S., and L.L. Bourgeau-Chavez, 1997. Monitoring south
Florida wetlands using ERS-1 SAR imagery, Photogrammetric
Engineering & Remote Sensing 63(3):281–291.
Khush, G.S., 1995. Breaking the yield frontier of rice, GeoJournal,
35(3):329–332.
Knight, J.F., R.L. Lunetta, J. Ediriwickrema, and S. Khorram, 2006.
Regional scale land-cover characterization using MODIS-NDVI
250 m multi-temporal imagery: A phenology based approach,
GIScience and Remote Sensing, 43(1):1–23.
Laur, H., P. Bally, P., Meadows, J. Sanchez, B. Schaettler, E. Lopinto,
and D. Esteban, 2002.
Derivation of the backscattering coefficient θ0 in ESA ERS SAR PRI
products, ESA, Noordjiwk, The Netherlands, ESA Document ES-
TN-RE-PM-HL09(2).
Le Toan, T., F. Ribbes, L.-F. Wang, N. Floury, K.-H. Ding, J.A. Kong, M.
Fujita, and T. Kurosu, 1997. Rice crop mapping and monitoring
using ERS-1 data based on experiment and modeling results, IEEE
Transactions on Geoscience and Remote Sensing, 35(1):41–56.
Leckie, D., 1990. Synergism of synthetic aperture radar and visible/
infrared data for forest type discrimination, Photogrammetric
Engineering & Remote Sensing, 56(9):1237–1246.
Leica, 2010. ERDAS Field Guide, Volume 4, October 2010.
Lobell, D.B., G.P. Asner, J.I. Ortiz-Monasterio, and T.L. Benning, 2003.
Remote sensing of regional crop production in the Yaqui Valley,
Mexico: Estimates and uncertainties, Agriculture, Ecosystems &
Environment 94(2):205–220.
Lopes, A., R. Touzi, and E. Nezry, 1990. Adaptive speckle filters
and scene heterogeneity, IEEE Transactions on Geoscience and
Remote Sensing, 28(6):992–1000.
Markham, B.L., and J.L. Barker, 1986. Landsat MSS and TM Post -
Calibration Dynamic Ranges, Exoatmospheric Reflectances and
At-Satellite Temperatures, Earth Observation Satellite Company,
Lanham, Maryland, Landsat Technical Notes, 1, August.
Naylor, R., 1992. Labour-saving technologies in the Javanese rice
economy: Recent developments and a look into the 1990s,
Bulletin of Indonesian Economic Studies, 28(3):71–91.
Nelson, A., T. Setiyono, A.B. Rala, E.D. Quicho, J.V. Raviz, P.J.
Abonete, A.A. Maunahan, C.A. Garcia, H.Z.M. Bhatti, and
L.S. Villano, 2014. Towards an operational SAR-based rice
monitoring system in Asia: Examples from 13 demonstration
sites across Asia in the RIICE Project, Remote Sensing,
6(11):10773–10812.
Rao, A.N., D.E. Johnson, B. Sivaprasad, J.K. Ladha, and A.M.
Mortimer, 2007. Weed management in directθseeded rice,
Advances in Agronomy (L.S.Donald, editor), Academic Press,
pp. 153–255.
Robinson, C., F. El-Baz, M. Ozdogan, M. Ledwith, D. Blanco, S.
Oakley, and J. Inzana, 2000.
Use of radar data to delineate palaeodrainage flow directions in
the Selima Sand Sheet, Eastern Sahara, Photogrammetric
Engineering & Remote Sensing, 66(6):745–753.
Sakamoto, T., M. Yokozawa, H. Toritani, M. Shibayama, N. Ishitsuka,
and H. Ohno, 2005. A crop phenology detection method using
time-series MODIS data, Remote Sensing of Environment, 96(3-
4):366–374.
Savary, S., N.P. Castilla, F. Elazegui, and P.S. Teng, 2005. Multiple
effects of two drivers of agricultural change, labour shortage and
water scarcity, on rice pest profiles in tropical Asia, Field Crops
Research, 91(2):263–271.
Shao, Y., X. Fan, H. Liu, J. Xiao, S. Ross, B. Brisco, R. Brown, and G.
Staples, 2001. Rice monitoring and production estimation using
multitemporal RADARSAT, Remote Sensing of Environment,
76(3):310–325.
Siddiq, E., 2000. Bridging the Rice Yield Gap in India, URL: http://
www.fao.org/docrep/003/X6905E/x6905e09.htm (last date
accessed: 24 September 2015).
Thenkabail, P.S., 2010. Global croplands and their importance for
water and food security in the twenty-first century: Towards an
ever green revolution that combines a second green revolution
with a blue revolution, Remote Sensing, 2(9):2305–2312.
Thenkabail, P., C. Biradar, P. Noojipady, V. Dheeravath, Y. Li, M.
Velpuri, M. Gumma, G.
Reddy, H. Turral, X. Cai, J. Vithanage, M. Schull, and R. Dutta,
2009a. Global irrigated area map (GIAM) for the end of the last
millennium derived from remote sensing, International Journal
of Remote Sensing, 30(14):3679–3733.
Thenkabail, P.S., C.M. Biradar, P. Noojipady, V. Dheeravath, Y. Li, M.
Velpuri, M., Gumma, O.R.P. Gangalakunta, H. Turral, X. Cai, J.
Vithanage, M.A. Schull, and R. Dutta, 2009b. Global irrigated
area map (GIAM), derived from remote sensing, for the end of
the last millennium, International Journal of Remote Sensing,
30(14):3679–3733.
Thenkabail, P., P. GangadharaRao, T. Biggs, M. Gumma, and H. Turral,
2007a. Spectral matching techniques to determine historical
land-use/land-cover (LULC) and irrigated areas using time-
series 0.1-degree AVHRR Pathfinder datasets, Photogrammetric
Engineering & Remote Sensing, 73(10):1029–1040.
Thenkabail, P.S., P. GangadharaRao, Y. Biggs, M.K. Gumma, and
H. Turral, 2007b. Spectral Matching techniques to determine
historical land use/land cover (LULC) and irrigated areas using
time-series AVHRR Pathfinder datasets in the Krishna River
Basin, India, Photogrammetric Engineering and Remote Sensing,
73(10):1029–1040.
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
November 2015 87 9
11-15 November Peer Reviewed.indd 879 10/21/2015 1:18:39 PM
Thenkabail, P.S., M. Schull, and H. Turral, 2005. Ganges and Indus
river basin land use/land cover (LULC) and irrigated area
mapping using continuous streams of MODIS data, Remote
Sensing of Environment, 95(3):317–341.
Thenkabail, P.S., E.A. Enclona, M.S. Ashton, C. Legg, and M.J. De
Dieu, 2004. Hyperion, IKONOS, ALI, and ETM+ sensors in the
study of African rainforests, Remote Sensing of Environment,
90(1):23–43.
Thiruvengadachari, S., and R. Sakthivadivel, 1997. Satellite remote
sensing for assessment of irrigation system performance: A
case study in India, Research Report 9, Colombo, Sri Lanka:
International Irrigation Management Institute.
Townsend, P.A., 2001. Mapping seasonal flooding in forested
wetlands using multi-temporal Radarsat SAR, Photogrammetric
Engineering & Remote Sensing, 67(7):857–864.
Uppala, D., R.V. Kothapalli, S. Poloju, S.S.V.R. Mullapudi, and V.K.
Dadhwal, 2015. Rice crop discrimination using single date
RISAT1 hybrid (RH, RV) polarimetric data, Photogrammetric
Engineering & Remote Sensing, 81(7):557–563.
Varlyguin, D., R. Wright, S.J. Goetz, and S.D. Prince, 2001. Advances
in land cover classification: A case study from the mid-Atlantic
region, American Society for Photogrammetry and Remote
Sensing, Proceedings of the Annual Conference, St. Louis,
Missouri.
Velpuri, N.M., P.S. Thenkabail, M.K. Gumma, C.B. Biradar, P. Noojipady,
V. Dheeravath, and L. Yuanjie, 2009. Influence of resolution in
irrigated area mapping and area estimations, Photogrammetric
Engineering & Remote Sensing, 75(12):1383–1395.
880 November 2015
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
11-15 November Peer Reviewed.indd 880 10/21/2015 1:18:39 PM
... Near real time satellite imagery helps in identifying droughts and floods for quick decisions (Gumma et al., 2017). Numerous studies have been conducted on monitoring croplands and natural resources using remote sensing and geographical information systems supported by secondary information (Rao et al., 2001;Gumma et al., 2009Gumma et al., , 2015Gumma et al., , 2018bQiu et al., 2013). Several studies mapped water bodies, flooded areas and soil moisture regimes using multiple data sets including MODIS, Landsat and sentinel (Feyisa et al., 2014;Gumma et al., 2015;Qiu et al., 2015). ...
... Numerous studies have been conducted on monitoring croplands and natural resources using remote sensing and geographical information systems supported by secondary information (Rao et al., 2001;Gumma et al., 2009Gumma et al., , 2015Gumma et al., , 2018bQiu et al., 2013). Several studies mapped water bodies, flooded areas and soil moisture regimes using multiple data sets including MODIS, Landsat and sentinel (Feyisa et al., 2014;Gumma et al., 2015;Qiu et al., 2015). Temporal satellite imagery and spectral analysis were successfully used in monitoring croplands and flooded areas in various studies (Gumma et al., 2014;Dong et al., 2015;Gumma et al., 2019), including at watershed and higher scales (Khan et al., 2001;Gumma et al., 2016). ...
... (3) thermal noise removal to remove additive noise in subswaths; (4) radiometric calibration to compute the backscatter intensity and (5) terrain correction to compute σ 0 on the basis of Digital Elevation Model (DEM). VV polarized images were considered as advantageous for flood mapping when using Sentinel-1 data (Gumma et al., 2015;Twele et al., 2016). Monthly composite images were computed from the preprocessed images to carry the monthly pattern of flooding during 2017 (eight images in a year). ...
Article
Full-text available
The dry lowlands of Ethiopia are seasonally affected by long periods of low rainfall and, coinciding with rainfall in the Amhara highlands, flood waters which flow onto the lowlands resulting in damage to landscapes and settlements. In an attempt to convert water from storm generated floods into productive use, this study proposes a methodology using remote sensing data and geographical information system tools to identify potential sites where flood spreading weirs may be installed and farming systems developed which produce food and fodder for poor rural communities. First, land use land cover maps for the study area were developed using Landsat-8 and MODIS temporal data. Sentinel-1 data at 10 and 20 m resolution on a 12-day basis were then used to determine flood prone areas. Slope and drainage maps were derived from Shuttle RADAR Topography Mission Digital Elevation Model at 90 m spatial resolution. Accuracy assessment using ground survey data showed that overall accuracies (correctness) of the land use/land cover classes were 86% with kappa 0.82. Coinciding with rainfall in the uplands, March and April are the months with flood events in the short growing season ( belg ) and June, July and August have flood events during the major ( meher ) season. In the Afar region, there is potentially >0.55 m ha land available for development using seasonal flood waters from belg or meher seasons. During the 4 years of monitoring (2015–2018), a minimum of 142,000 and 172,000 ha of land were flooded in the belg and meher seasons, respectively. The dominant flooded areas were found in slope classes of <2% with spatial coverage varying across the districts. We concluded that Afar has a huge potential for flood-based technology implementation and recommend further investigation into the investments needed to support new socio-economic opportunities and implications for the local agro-pastoral communities.
... Hence, proper management is important to avoid these factors which limit and reduce the yield in direct seeded rice. While many studies using SAR have been conducted to accurately map and monitor rice areas, very few attempts have been made to further classify rice areas by crop establishment method (Gumma et al., 2015;Yang et al., 2014;Yang et al., 2017;Fikriyah et al., 2019). Both studies by Gumma et al. (2015) and Yang et al. (2014) found that TP and DS rice have different backscatter responses during crop establishment. ...
... While many studies using SAR have been conducted to accurately map and monitor rice areas, very few attempts have been made to further classify rice areas by crop establishment method (Gumma et al., 2015;Yang et al., 2014;Yang et al., 2017;Fikriyah et al., 2019). Both studies by Gumma et al. (2015) and Yang et al. (2014) found that TP and DS rice have different backscatter responses during crop establishment. Gumma et al. (2015) used RISAT-1 HH polarisation, with 25 day revisit cycle while Yang et al. (2014) used RADARSAT-2 quad polarisations with 24 day revisit cycle. ...
... Both studies by Gumma et al. (2015) and Yang et al. (2014) found that TP and DS rice have different backscatter responses during crop establishment. Gumma et al. (2015) used RISAT-1 HH polarisation, with 25 day revisit cycle while Yang et al. (2014) used RADARSAT-2 quad polarisations with 24 day revisit cycle. Both are C-band SAR data. ...
Article
Full-text available
Occurrence of pests and diseases are influenced by several factors including weather, landscape and field-level factors such as crop management practices including crop establishment method. In this paper, we adopted and applied a method using Sentinel-1A (S-1A) Synthetic Aperture Radar (SAR) intensity to discriminate between rice fields that are transplanted and direct seeded to come up with a robust method for automated classification of crop establishment method. Multi-temporal S-1A C-band dual polarization images at 20m resolution covering the wet cropping season over four provinces in the Philippines were acquired from March to November 2018. Field measurements, observations and interviews were conducted on 186 sample fields and mean backscatter values for each of the sampled fields were generated from S-1A data acquired during the season. The reported dates of land preparation and estimated dates of crop growth stages were matched with the corresponding SAR acquisition dates. We used the Mann-Whitney U test to identify growth stages for which there are significant differences in backscatter values between transplanted and direct seeded rice. The results are generally consistent with the findings of a previous study conducted in one province in the Philippines in the dry season of 2017. We found, however, some inconsistencies in terms of the polarization where the significant differences were observed. These findings demonstrate the possibility of discriminating transplanted from direct seeded rice using SAR temporal data but suggests further fine tuning in the methodology is needed for different locations and seasons.
... Only a few studies have considered SAR data to distinguish between crop establishment methods. Gumma et al., (2015) demonstrated the use of RISAT-1 (C-band, HH polarisation, and 25-days revisit time) to distinguish TP and DDS rice in India. This study showed that early TP and DDS rice have different backscatter responses during the establishment period. ...
... Prior studies have used SAR data to observe the temporal backscatter characteristics of TP and DS rice using RISAT-1 data (C-band, HH polarisation, 25-days temporal resolution) (Gumma et al., 2015) and RADARSAT-2 data (C-band, full polarisation, 24-days temporal resolution) (Yang et al., 2014). However, these studies had limited acquisition frequency. ...
... Our finding that TP and WDS rice is best discriminated through the observation in the early growing season confirms those obtained by Yang et al. (2014) who investigated the seedling stage as the primary time to distinguish transplanted hybrid rice and direct seeded Japonica rice (by RADARSAT-2, C-band, quad polarimetric data) based on the higher dB from the seedlings in transplanted rice. We also agree with the results from Gumma et al. (2015) who identified that TP and DS rice have distinct backscatter during transplanting based on C-band RISAT-1 data at HH polarisation. However, our results are the first to identify opportunities to discriminate between TP and WDS across different land management activities and crop stages. ...
Article
Improved rice crop and water management practices that make the sustainable use of resources more efficient are important interventions towards a more food secure future. A remote sensing-based detection of different rice crop management practices, such as crop establishment method (transplanting or direct seeding), can provide timely and cost-effective information on which practices are used as well as their spread and change over time as different management practices are adopted. Establishment method cannot be easily observed since it is a rapid event, but it can be inferred from resulting observable differences in land surface characteristics (i.e. field condition) and crop development (i.e. delayed or prolonged stages) that take place over a longer time. To examine this, we used temporal information from Synthetic Aperture Radar (SAR) backscatter to detect differences in field condition and rice growth, then related those to crop establishment practices in Nueva Ecija (Philippines). Specifically, multi-temporal, dual-polarised, C-band backscatter data at 20m spatial resolution was acquired from Sentinel-1A every 12 days over the study area during the dry season, from November 2016 to May 2017. Farmer surveys and field observations were conducted in four selected municipalities across the study area in 2017, providing information on field boundaries and crop management practices for 61 fields. Mean back-scatter values were generated per rice field per SAR acquisition date. We matched the SAR acquisition dates with the reported dates for land management activities and with the estimated dates for when the crop growth stages occurred. The Mann-Whitney U test was used to identify significant differences in backscatter between the two practices during the land management activities and crop growth stages. Significant differences in cross-po-larised, co-polarised and band ratio backscatter values were observed in the early growing season, specifically during land preparation, crop establishment, rice tillering and stem elongation. These findings indicate the possibility to discriminate crop establishment methods by SAR at those stages, suggesting that there is more opportunity for discrimination than has been presented in previous studies. Further testing in a wider range of environments, seasons, and management practices should be done to determine how reliably rice establishment methods can be detected. The increased use of dry and wet direct seeding has implications for many remote sensing-based rice detection methods that rely on a strong water signal (typical of transplanting) during the early season.
... The dry-seeded sowing practice in Raichur District of Karnataka state was estimated to be about 13,000 ha (Gumma et al., 2015). DSR is a common practice among farmers in West Singhbhum and Saraikela -Kharsawan Districts of Jharkhand due to the uncertainty of monsoons, water shortages and labour scarcity (Barla et al., 2021). ...
... The DSR is now spreading to Sindhanur, Gangavati areas (Gumma et al., 2015) and is becoming a widespread rice cultivation practice in Karnataka (Gurupadappa et al.,2018). Working in that area, one of us (A. ...
Article
Full-text available
In India, rice is predominantly grown as puddled transplanted rice (PTR) under irrigated or assured rainfall conditions. The share of groundwater in net irrigated area, as compared to the area under surface irrigation, is more than 60% at present. The over-exploitation of groundwater through the explosion of tube wells has raised sustainability issues. India's Central Groundwater Board has warned of critically low groundwater availability by 2025. Rice cultivation under PTR is labour and energy-intensive. The rising costs of labour and energy in India is making PTR less profitable. PTR is also not very environment-friendly due to its relatively higher methane emissions. Due to the above concerns, the shift of rice cultivation to direct-seeding (DSR) has been well researched and developed in India. The technology has also been actively promoted and disseminated for farmers to adopt across many Indian states. The advantages of the DSR system can be obtained only by alleviating the significant constraints, including weed problems and issues related to crop nutrition. The research carried out at different agro-ecological conditions in India has amply proved that the adoption of improved DSR technologies results in several advantages over PTR. The benefits include savings in labour (40-45%), water (30-40%), fuel/energy (60-70%), and reductions in greenhouse gas emissions. In this paper, we briefly discuss the historical aspects of DSR in India, the advantages of DSR, the reasons for inadequate adoption of DSR during the pre-pandemic period, the farmers' adoption of DSR during the pandemic making the crisis an opportunity. We also discuss the potential and research/extension needs for further upscaling DSR in India during the post-pandemic period.
... The dry-seeded sowing practice in Raichur District of Karnataka state was estimated to be about 13,000 ha (Gumma et al., 2015). DSR is a common practice among farmers in West Singhbhum and Saraikela -Kharsawan Districts of Jharkhand due to the uncertainty of monsoons, water shortages and labour scarcity (Barla et al., 2021). ...
... The DSR is now spreading to Sindhanur, Gangavati areas (Gumma et al., 2015) and is becoming a widespread rice cultivation practice in Karnataka (Gurupadappa et al.,2018). Working in that area, one of us (A. ...
Article
Full-text available
In India, rice is predominantly grown as puddled transplanted rice (PTR) under irrigated or assured rainfall conditions. The share of groundwater in net irrigated area, as compared to the area under surface irrigation, is more than 60% at present. The over-exploitation of groundwater through the explosion of tube wells has raised sustainability issues. India's Central Groundwater Board has warned of critically low groundwater availability by 2025. Rice cultivation under PTR is labour and energy-intensive. The rising costs of labour and energy in India is making PTR less profitable. PTR is also not very environment-friendly due to its relatively higher methane emissions. Due to the above concerns, the shift of rice cultivation to direct-seeding (DSR) has been well researched and developed in India. The technology has also been actively promoted and disseminated for farmers to adopt across many Indian states. The advantages of the DSR system can be obtained only by alleviating the significant constraints, including weed problems and issues related to crop nutrition. The research carried out at different agro-ecological conditions in India has amply proved that the adoption of improved DSR technologies results in several advantages over PTR. The benefits include savings in labour (40-45%), water (30-40%), fuel/energy (60-70%), and reductions in greenhouse gas emissions. In this paper, we briefly discuss the historical aspects of DSR in India, the advantages of DSR, the reasons for inadequate adoption of DSR during the pre-pandemic period, the farmers' adoption of DSR during the pandemic making the crisis an opportunity. We also discuss the potential and research/extension needs for further upscaling DSR in India during the post-pandemic period.
... Also, Kontgis, Schneider, and Ozdogan (2015) remarked that many of the problems caused by missing data related to cloud cover and SLC-off gaps can be overcome by using long temporal series covering multiple years, emphasizing that flooded scenes are particularly important for the characterization of the rice fields. Alternatively, Synthetic Aperture Radar (SAR) can be used in combination with optical images to reduce the impact of cloud cover (Gumma et al. 2015). ...
... The literature suggests a few ways to tackle this problem. In areas where cloud cover is common, it is usual to combine multispectral images from different sensors (Hao et al. 2014), as well as to integrate multispectral images with data provided by SAR sensors (Esch et al. 2014;Karila et al. 2014;Gumma et al. 2015;Whitcraft et al. 2015;Kussul et al. 2016). SAR sensors are much less affected by weather phenomena (Inglada et al. 2016) and are sensitive to land surface characteristics and leaf structure and canopy architecture, which can help identifying crops (Forkuor et al. 2014). ...
Preprint
Full-text available
Classifying crop areas is very important for production forecasting, formulation of public policies, management of natural resources, among others. Manual classification is a labour intensive, expensive and error prone process, making the search for alternative options a priority. As more high quality satellite images become available, automating (at least partially) the classification process using image processing and machine learning seems a viable choice. However, there are many challenges yet to be overcome which prevent this kind of strategy to be used in practice. The objectives of this review were: 1) to identify those challenges and how they were tackled by different authors over the course of the last few decades and 2) to function as a source of ideas for future research on the subject of automatic segmentation and classification of agricultural areas in remotely sensed images, with focus on satellite data comprising the visible and near infrared spectral bands.
... Satellite remote sensing provides a time-saving and influential approach for monitoring agricultural areas and other land cover features. [2][3][4][5][6] Optical remote sensing is a viable approach to map rice-growing areas effectively at regional and global levels because of its potential for large-area coverage and repeated observations. 4,[7][8][9] Because most of the rice grows in rainy and cloudy regions, it is immensely difficult to attain cloud-free optical images during critical rice-growing seasons. ...
... The launch of the first Indian spaceborne hybrid polarimetric SAR system Radar Imaging Satellite-1 (RISAT-1) by the Indian Space Research Organization in April 2012 brought more opportunities for mapping and monitoring rice crops. 3,16 It carries a hybrid polarimetric SAR payload operating at C-band that supports right circular transmit and coherent linear receive mode as well as other standard modes. The imaging capability of RISAT-1 in HH, HV, VH, VV, and circular polarizations ensured its broad aptness. ...
Article
Full-text available
Updated and accurate information of rice-growing areas is vital for food security and investigating the environmental impact of rice ecosystems. The intent of this work is to explore the feasibility of dual-polarimetric C-band Radar Imaging Satellite-1 (RISAT-1) data in delineating rice crop fields from other land cover features. A two polarization combination of RISAT- 1 backscatter, namely ratio (HH/HV) and difference (HH−HV), significantly enhanced the backscatter difference between rice and nonrice categories.With these inputs, a QUEST decision tree (DT) classifier is successfully employed to extract the spatial distribution of rice crop areas. The results showed the optimal polarization combination to be HH along with HH/HV and HH −HV for rice crop mapping with an accuracy of 88.57%. Results were further compared with a Landsat-8 operational land imager (OLI) optical sensor-derived rice crop map. Spatial agreement of almost 90% was achieved between outputs produced from Landsat-8 OLI and RISAT-1 data. The simplicity of the approach used in this work may serve as an effective tool for rice crop mapping.
... Sentinel The process of labeling class identification was done based on spectral matching techniques (SMTs) (Gumma et al. 2018;Gumma et al. 2016;Gumma et al. 2015). Initially, 60 classes from the unsupervised classification were grouped based on spectral similarity or closeness of class signatures. ...
Technical Report
Full-text available
Ahmednagar is the largest district of Maharashtra in terms of area and population. It lies in the central part of the state of Maharashtra which is having common boundaries with seven adjoining Districts. The total geographical area of the district is 17.41 lakh ha. The net cropped area is 12,56,500 ha, out of which an area of 3,30,000 ha. (26.27 %) is under canal (84,000 ha) and well irrigation. About 9,26,500 ha. (73.73 %) area is rain fed. The area under Kharif crops is 4,60,000 ha. (36.6 per cent) while 7,58,000 ha (60.32 per cent) area is under Rabi crops. A multiple cropping system is followed on 1,10,500 ha area. A total of 8.73 per cent area of the district is under forest. The climate of the district is hot and dry, on whole extremely genial and is characterized by a hot summer and general dryness during major part of the year except during south-west monsoon season. Ahmednagar district receives average 566 mm. rainfall. The major rainfall received during month of June to September. The average temperature ranges between 9 0c (during Dec.) to 41 0C (during April and May). The soil types of the district are broadly divided into four categories namely coarse shallow soil; medium black soil; deep black soil and reddish soil occupying about 38, 41, 13 and 8 percent of the cultivated area respectively. In the first two categories, soil moisture is the predominant limiting factor affecting productivity of crops particularly under rainfed condition. Godavari and Bhima are the major rivers in the district. Godavari river flows through the northern border of Ahmednagar district. Major Kharif crops grown in the district are Cotton, Maize, Bajra, Sugarcane, and Soybean and during Rabi season are Jowar, Wheat, Soybean and Pulses.
... The study was successful due to the increased observations from the Landsat 8 images, made available by sidelaps in high altitude regions. In an another study, Gumma, Uppala, Mohammed, Whitbread, and Mohammed (2015) used two dates Landsat-8 images combined with temporal RISAT-1 and time series MODIS data to classify direct seeded rice areas. This study was successful due to the use of multi-source data and a hybrid algorithm. ...
Article
Full-text available
Repeated monitoring of paddy rice is essential for government agencies and policy makers to maintain the balance of supply and demand for rice. Recent studies have mostly concentrated on the mapping of paddy rice with temporal satellite imagery during growing seasons. Given the phenological variation within paddy rice fields and spectral confusion between paddies and other vegetation classes, our ability to identify paddy rice fields with temporal imagery remains limited. The objective of this study is to develop new phenology and textural-based strategies to detect paddies with HJ-1A, MODIS and PALSAR FNF imagery. Two phenology-based strategies that track the seasonal trajectory of crops and one textural-based strategy that contains image surface characteristics are presented. With the proposed strategies, temporal, spectral and textural features were investigated for paddy rice detection. The results indicate that the phenology-based strategies could reveal the phenological variation within paddy rice and significantly improved the detection accuracy. Seasonal amplitude, grey level co-occurrence matrix entropy and spectral features of the heading stage were proven to be important in identifying paddy rice. It was concluded that the combination of HJ-1A, MODIS and PALSAR FNF imagery are promising in facilitating the rapid mapping of paddy rice at a regional scale.
Article
Full-text available
In India, agriculture monitoring is severely hampered by frequent cloud cover. Crop health assessment from the early growing stage onwards is vital for accurate and timely yield prediction. In this study, time series of Sentinel-1A SAR images over central India have been processed to quantify kharif crops’ growth rate. Sentinel-1 and Sentinel-2 images acquired on the same day have been used to compare the radar backscattered energy at the VH channel with normalized difference vegetation index (NDVI). A good coefficient of determination (R² = 0.824) was found between backscatter and NDVI. It attests that the NDVI can be used in combination with SAR backscatter during the kharif season. In addition, k-means clustering classification of Sentinel-1 images indicated that the total area covered by paddy, soybean, and other crops were 103,506.86 ha, 85,390.93 ha, and 71,667.02 ha, respectively. The classification result has been validated with ground information, which has indicated an overall accuracy of 83.47%. The work indicated that radar signals’ temporal behavior is sensitive to the health status of the crops from sowing to harvesting stages. Sentinel-1 SAR images can be used to analyze kharif crops during the whole phenological cycle. The approach may serve as a solution for assessing both the health and spatial distribution of kharif crops.
Article
Full-text available
Palaeodrainage directions in the Selima Sand Sheet (centered on 22.5°N, 29°E) were determined using high-resolution, multi-wavelength, multi-polarization Spaceborne Imaging Radar (SIR-C) data and the Global Land One-kin Base Elevation (GLOBE) Project Digital Elevation Model (DEM). The combined use of these two data sets shows that both large flood features and later superimposed drainage channels of variable morphology all drain NE and ENE from northwest Sudan toward the Kharga depression in southern Egypt. This is supported by drainage directions deduced from the USGS Global Topography (GTOPO) DEM. These directions are opposite to those of the Trans-African Drainage System (TADS) model in which the large flood features are considered to flow southwest across northeastern Africa into the Chad Basin. Instead, the results show that an internal drainage basin operated in the gently undulating terrain of the Selima Sand Sheet (probably during the Cenozoic period), and that the slope of the North African plate remained generally northeastward during those times. Further, the northeastern parts of the Selima Sand Sheet are likely to be the primary area for ground-water accumulation in southern Egypt.
Article
Full-text available
Rice is the most important food grain crop in India and contributes to more than 40 percent of the country's food grain production. Spaceborne remote sensing offers economically viable and accurate production and area statistics. The utility of optical remote sensing in mapping rice cropped area is limited by persistent cloud cover during monsoon season. Temporal availability of SAR data has facilitated an operational procedure to monitor the rice crop. The current study discriminates rice crop, using single date hybrid polarimetric data available from RISAT-1 SAR. This was subjected to Raney m-δ, m-χ decompositions, and supervised classification was performed. The accuracy was estimated using the field points. The results were compared with rice map generated using optical sensor Resourcesat-2 LISS-IV and statistical data. The spatial agreement between the estimate from RISAT-1 and LISS-IV data was found to be 85 percent. The class kappa value was 0.94 and 0.92 for LISSIV and RISAT-1, respectively.
Article
Full-text available
This study presents a methodology to classify rice cultural types based on water regimes using multi-temporal synthetic aperture radar (SAR) data. The methodology was developed based on the theoretical understanding of radar scattering mechanisms with rice crop canopy, considering crop phenology and variation in water depth in the rice field, emphasizing the sensitivity of SAR to crop geometry and water. The logic used was the characteristic decrease in SAR backscatter that is associated with the puddled or transplanted field due to specular reflection for little exposure of crop, with increase in backscatter as the crop growth progresses due to volume scattering. Besides, the multiple interactions between SAR and vegetation/water also lead to an increase in backscatter as the crop growth progresses. Classification thresholds were established based on the information provided by each pixel in each image, the pixel’s typical temporal behaviour due to crop phenology and changing water depth in rice field and their corresponding SAR signature. Based on this logic, the study site (i.e. South 24 Paraganas district, West Bengal) was classified into three major rice cultural types, namely shallow water rice (SWR; 5 cm ≤ water depth ≤ 30 cm), intermediate water rice (IWR; 30 cm ≤ water depth ≤ 50 cm) and deep water rice (DWR; water depth > 50 cm) during the kharif season. These three types represent most of the traditional rice-growing areas of India. The methodology was validated with the field data collected synchronously with the satellite passes. Classification results showed an overall accuracy of 98.5% (95.5% kappa coefficient) compared with a maximum-likelihood classifier (MLC) with an overall accuracy of 95.5% (84.2% of kappa coefficient) with 95% confidence interval. The relationship between field parameters, especially exposed plant height and water depth with SAR backscatter, was explored to design empirical models for each of the three rice classes. Significant relationships were observed in all the rice classes (coefficient of determination, R2, value more than 0.85) even though they had similar growth profiles but varied with water depth. The two main conclusions drawn from this study are (i) the importance of multi-temporal SAR data for the classification of rice culture types based on water regimes and (ii) the advantages and flexibility of the knowledge-based classifier for classification of RADARSAT-1 data. However, being empirical, the approach needs modification according to the current rainfall pattern and rice-growing practice.
Article
Five visible, three near-infrared, and one mid-infrared airborne MSS bands and eight airborne synthetic aperture radar (SAR) bands were acquired during summer over a forest test site near Chalk River, Ontario, Canada. Radar bands were better at discriminating softwood species than were the visible/infrared bands. Red and white pine had relatively low microwave backscattering, whereas jack pine and white spruce had high backscattering. The hardwoods tested had intermediate to high backscattering. Use of a near-infrared or mid-infrared band is important for distinguishing hardwoods from softwoods. Visible and mid-infrared bands are useful for discriminating open areas with either sparse, moderate, or dense ground vegetation from forested areas. -from Author
Article
Imaging radars provide information that is fundamentally different from sensors that operate in the visible and infrared portions of the electromagnetic spectrum. The Indian Space Research Organisation (ISRO) has launched a multi-mode, multi-polarization Synthetic Aperture Radar (SAR) on-board Radar Imaging SATellite-1 (RISAT-1) on 26 April 2012. Various data products from RISAT-1 SAR are now going through calibration-validation (cal-val) phase and soon will be available for the global users for operational and research purposes. In this regard, algorithms are being developed to retrieve various parameters in diverse application areas. This article deals with the in-house algorithm development for studying different resources using initial available data of RISAT-1.
Article
We present a first assessment of UAS-derived orthoimagery and digital elevation data in the context of newly-released accuracy standards for digital geospatial data developed by the American Society for Photogrammetry and Remote Sensing. We outline results from two case studies using a commercially-available UAS, photogrammetry software, and an array of ground control and check points. Radial horizontal and vertical root-mean square-errors (RMSE) were calculated as 0.05 m and 0.06 m, respectively, for one site, and 0.08 m and 0.03 m, respectively, for the other. Under the 1990 ASPRS standards, both surveys meet the requirements for Class 1 accuracy at the 1:500 map scale and at the 0.50 m contour interval. Under the newly-developed ASPRS standards, the reported errors fulfill the requirements for both horizontal and vertical mapping at the 10 cm RMSE level. Overall, these results provide initial direction for practitioners considering UAS surveying in the context of accuracy standards.