Nimbus TOMS
method Sof
was. used^ for the
present study. *This:-TOMS- .IPAR
data set consists of monthly
average estimates, at a spatial
resolution OF 1°*<12 Cdegres from
data from
through: the
Dye {19919
sensor
Eck and
669N- ‘to 166 "S “Latitude. “The data
for the Indian region Was
extracted ‘from (the uglobal’ ‘data
set’ Cand 2 interpolated "to match
the 39^: km.cresolutiormn" of “NDVI
data.
3.2 Climatic data
Time seriesoiclimatic data" for
India including 'daily rainfall,
maximum and minimum temperature
were extracted from Global
Summaries prepared by National
Climatic Data: Center, USA. "This
contained daily observations
for: more than 200nistations Jof
India for::1977-1991. a But: cCom-
plete:.data-cofor only about 975
stations.in India was:available
for-ethe-year 989 which: was
used in the present study. The
daily data was converted to
monthly rainfall and monthly
average mean temperature.
4. MAPPING AGRICULTURAL AREAS:
the
the
One of
faced in
major challenge
estimation of
agricultural productivity is
the mapping of agricultural
areas. "Therefore,o^efforts were
made to develop an automated
technique for the identifi-
cation” and mapping of tagricul-
tural areas based on NDVI-
climatologucal modeling. The
conceptanisy based ron: thesgfact
that «he NDVI'of*- the (natural
vegetation 1s expected to show
a positive correlation with-the
climatic. ofactors « Oofvsthe area,
but: ‘notiithe NDVI ofVthe:dgri-
cultural crops: which ‘are sarti-
ficialbly" mañaged “by € œupplying
terms of
Therefore,
addicioóonal inputs ‘in
water and nutrients.
there is a possibility of
identifying the agricultural
pixels sasooutiiersjin the NDVI-
climatological relationship
(Hooda and Dye, 1995).
The NDVI and climatit ‘data for
the year 1989, a .normalijcyear
with respect to monsoon
effecting Indian agriculture,
was used for the present] study:
The = ‘point climatic data for
about 75 meteorological
stations was correlated against
the average NDVI in /a 3*3'pixel
window around the same
location. Relationship was
tried for different crop
growing seasons of winter,
summer ‘and monsoon as well. as
on annual basis.
No relationship between
and mean "temperature > could be
observed in the present. study
The ‘possible ’reason-‘could be
that India is a tropical
country “andvitemperature 1istingt
a limiting factor for the
growth ¢ of {vegetation fore moss
of the year. Relationship
between NDVI and"5rainfall Mg
different Seasons also
not be^ observed but ->the-“ annual
integrated NDVI0 : did0 ‘show f
Togar ithmié relationship with
the annuali: rainfall. ‘However,
some outlier pixels showing
very high NDVI^át'"-low rainfall
were also noticed. The rela-
tionship improved signifi.
cantly after removing these
outlier ‘pixedlssricBased ‘on, ‘this
analysis a pixel was classified
as agricultural:pixel if,
NDVI
could
ENDVI=0.0042*ann. rainfall+0.5
Since this
high NDVI
rainfall, at
to assume
technique identifies
pixels at low
would "be-ological
these pixels as
irrigated agricultural pixels
because only irrigated'aoucmops
can show high NDVI-evem: atc ow
rainfall due to availability of
water ‘through irrigation.c Thus,
one’ ofc thew: limitations «of niche
technique ,is'"^that it^omay not
separate Out dry land
agricultural areas cas>well/%s
some 'of*trhe irrigatedv areas iin
the high rainfall eastern
region of the country. "However,
when compared with the
available irrigated areas map
of ithe (country, .the’ctechnique
Seems to give ^a fair idea of
the major‘ irrigated "areas ‘in
the- country.. The? net irrigated
area reported in the country is
only 1397290 sg km.;… but’ the
net sown area with reasonably
assured water supply is
reported as 726170 Sq. km.
(Anonymous, 139897) compared to
750016 sq.’ km. observed based
upon the present technique.
Thus, the NDVI-climatological
technique proved -quite useful
in quickly generating an
irrigated agricultural areas
map. This ‘map Was cCused ^as à
mask ‘to’ extractidifferent data
sets for only agricultural
areas of India.
5. AGRICULTURAL PRODUCTIVITY
ESTIMATION
estimating
different
Use of PEM for
productivity. involves
steps as detailed below:
5.1 Fraction of IPAR absorbed
by vegetation (fAPAR)
vegetation index
The. spectral
produced by
measurements
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996
