)ps need 
ia block, 
ssing the 
r mango 
1as been 
ars were 
aximum 
the land 
that IRS 
tribution 
a, which 
increased 
failed to 
table sites 
| view, the 
nating the 
oups and 
ramework 
Ct. 
Path — 
t, Uttar 
rdinates 
y 78%11° 
of India 
toposheets on 1: 50,000 scale. The Siyana block covers about 
18079 ha area, encompassing 35 villages and one town (Census, 
1991). The Ganga River flows in the eastern side of the block. 
The climate of the area is sub-tropical monsoon type with 
moderately cold and dry winter and hot summer. There are two 
major soil orders in the study area viz. Entisols and Inceptisols 
and eight soil series identified in the study area. 
  
  
  
  
  
  
Fig. 1 : Location map of study area 
4. METHODOLOGY 
For the present study, IRS-1D, LISS-III data of summer season 
(May 2000) was used, because during this period, no major crops 
were present in the field. The image of study area was 
geometrically rectified with respect to Survey of India (SOI) 
topographical maps on 1:50,000 scale. Tehsil map on 1:63,360 
scale from Dept. of Census was used for extracting study area 
boundary and preparation of village map. Tehsil map of study 
area was brought on 1:50,000 scale using affine transformation 
technique of ARC/INFO software. Projected study area boundary 
was overlaid on rectified image and study area was extracted. 
False color composite was generated using near infrared, red and 
green band of LISS-III data. Using ground truth information 
training sites were marked for young mango orchard, (orchard 
with age >5 to < 15 years) old mango orchard (orchard with age > 
15 years), current fallow, crop, water body, aquatic weed and 
settlement. 
Maximum likelihood supervised classification algorithm was used 
for mango orchard area estimation. The area covered by aquatic 
weeds alongwith the canal and the ponds were masked before 
classification of image. Accuracy of digital classification of 
various landuse classes was assessed using the confusion matrix. 
GIS analysis techniques were used for village- wise mango 
orchard acreage estimation. Land evaluation is the process of 
estimating the potential for alternative kind of use. FAO 
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring”, Hyderabad, India, 2002 
framework of land evaluation (FAO, 1983) was employed in this 
study to assess the potential soil suitability of mango crop using 
GIS based integration analysis. For this purpose, available soil 
map of the Siyana block (RSAC-U.P, 1998) was used. The soil 
suitability of mango was assessed based on land qualities of soil 
For this, criteria of land qualities rating for mango orchard 
developed by Agriculture and Soils Division of Indian Institute of 
Remote Sensing, Dehradun were used. (2001). 
5. RESULTS AND DISCUSSION 
The study has clearly demonstrated the usefulness of IRS-1D, 
LISS-III data for acreage estimation of mango orchards and 
identifying the two age groups of mango orchards, viz. Orchards 
having trees between 5 to 15 years of age (young orchard) and 
orchards having trees more than 15 years of age (old orchard). 
The acreage estimation of different landuse categories on the 
basis of IRS-1D, LISS-III data is given in Table: 1. Existing 
mango orchard distribution map is presented in Fig. 3. The total 
26.75 per cent mango orchard area was found, in which young 
orchard is 5.95 % and old orchard is 20.8096 of the total study 
area (18079 ha). The spectral response of mango orchard along 
with other land cover classes is presented in fig: 2 in the form of 
mean D.N. values of spectral band. The false colour composite 
(FCC) image of IRS-1D, LISS-III of the study area was digitally 
classified. The classified images presented in figure 3. It was 
found that satellite data of summer season is suitable for mango 
acreage estimation. 
  
  
  
  
  
  
  
  
  
  
  
Landuse Classes Area (ha) % 
Old Mango orchard (>15 | 3756.29 20.8 
ears) 
Young mango orchard (5-15 1075.23 5.95 
years 
Current fallow 8704.90 48.2 
Crop 4109.07 22.7 
Settlement 347.92 1.93 
Aquatic weed 4.97 0.03 
Canal ! 53.4 0.3 
Water body 27.28 0.15 
Total 18079.06 100 
  
  
  
  
Table-1: Area Statistics of Mango and other landuse classes 
  
  
  
SPECTRAL RESPONSE CURVE 
150 pz —0—ciop 
2 1001. —iit— fallow 
S sui 
z 501 mango old 
= 3 
0 —% mango 
bandi band2 band3 young 
—3J*- settlement 
BANDS 
  
  
  
  
  
Fig 2: Spectral Response Curve of mango and other classes.