«MENT 
r optimum 
equired for 
from space 
aps soils at 
of multiple 
ote sensing 
state, India. 
yugh visual 
respect to 
ilating LPI 
roductivity 
fferent soil 
| respect to 
uctivity in 
le benefits 
iment had 
it 1:50,000 
oping units 
mandal of 
lia. It lies 
to 82555 
) hectares. 
t site is a 
arnokites / 
te with a 
ing south- 
mum and 
8°C to 22° 
rothermic 
onths and 
aries drain 
st to east. 
deciduous 
tribals are 
The crops 
er, jowar, 
the crops 
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring", Hyderabad, India, 2002 
4. DATA USED 
In this study the data used consists of satellite data, ancillary 
information and topographical maps of Survey of India. 
Remotely sensed data in the form of FCC (false colour 
composite) prints at 1:50,000 scale, from IRS-IB LISS-II 
sensors — A1, A2, B1 and B2 of 26" February 1993 (figure-2) 
were used to map soils and land use of the study area through 
visual interpretation technique and ground information. Survey 
of India topographical maps and published data were used as 
ancillary information. The soil and land use maps were 
prepared at 1:50,000 scale. ARC / INFO GIS package was used 
to create database and for computation of land productivity and 
for generation of outputs 
  
Figure-1 IRS-IB LISS FCC of part of Paderu mandal 
Visakapatnam district of Andhra Pradesh State, India 
5. METHODOLOGY 
The methodology adopted for the study consists of preparation 
of soil and land use maps using satellite data at 1:50,000 scale 
with filed observations, land productivity assessment through 
parametric approach of Riquier et al.,(1970), creation of data 
base in ARC / INFO GIS and integrated analysis of soil / land 
cover information to arrive at land productivity of different soil 
mapping units of the study area. The various steps involved at 
each stage are discussed below: 
5.1. Preparation of soil map 
The soil map of the test site was prepared through systematic 
visual interpretation of FCCs of IRS-IB LISS-II at 1:50,000 
scale, systematic soil profile studies in the field, analysis of soil 
samples and classification of soils as per soil taxonomy (Soil 
Survey Staff, 1992). The final soil map was prepared with 
appropriate legend showing the relationship between 
physiography and soils in the test site. The soils were classified 
at family level and their associations. 
5.2 Preparation of the land use / land cover map 
The FCC prints of satellite data were preliminarily interpreted 
on light table based on photo-elements like tone, texture, size, 
shape etc., and various land use / land cover classes were 
delineated by placing a translucent film on the FCCs. The 
classification scheme was adopted from manual of land use / 
land cover mapping using satellite imagery (NRSA,1989). 
Ground truth was collected both kharif and rabi seasons for 
land use / land cover classes delineated from satellite data. The 
final map was prepared by incorporating the filed observations 
through cartographic work with an appropriate legend. 
5.3 Land productivity assessment 
The land productivity of the test site was assessed with respect 
to crops, pasture and forest / tree species following a parametric 
approach of Riquier et al.,(1970) where the land productivity 
Index (LPI) was computed using a multiplicative model using 
nine soil and site parameters namely moisture (H), drainage 
factor (D), depth factor (P), texture / structure (T), base 
saturation (N), soluble salt concentration (S), organic matter 
content (O), mineral exchange capacity (A), and mineral 
reserve (M). 
LPIZHxDxPxTxNxSxOxAxM 
Each of these factors is rated on a scale from 0 — 100 for crops, 
pasture and forest / tree species separately and their percentages 
are multiplied to arrive at LPI. The resultant LPI also lies 
between 0 — 100 and is set against a scale placing the soil in one 
or other five productivity classes as shown in table —1. 
Table - 1 Actual — P, and Potential — P^ ratings of LPI 
values for productivity classes 
  
  
  
  
  
  
Actual -P Productivity LPI Range Potential — P* 
class 
1 Excellent 100 - 65 I 
2 Good 64 — 35 II 
3 Average 34 — 20 HI 
4 Poor 19-8 IV 
5 Extremely poor 7-0 V 
  
  
  
  
  
  
GIS has potential for modeling the above land productivity 
parameters through integrated analysis of multi-parameter data 
that includes different spatial inputs such as -soils, land use and 
other non-spatial data like soil physical and chemical 
properties, Originally the model was proposed to study the 
individual soil types. In the current experiment GIS tools were 
used not only for calculating LPI values for soil types but also 
for deriving area weighted LPI values for soil mapping units 
and in the generation of Land Productivity map for the study 
area. In the present study actual LPI (P), potential LPI (P^) and 
coefficient of improvement for various soil mapping units were 
computed for crops, pasture and forest / tree species using ‘aml’ 
utilities in GIS. The coefficient of improvement (C) was 
computed for all the soil mapping units by taking the ratio of 
Potential LPI (P’) upon Actual LPI (P). 
6. RESULTS AND DISCUSSIONS 
6.1 Soils resources 
The soil map of the test site was prepared at 1:50,000 scale by 
establishing the physiography — soil relationship. Altogether 8 
major physiographic units, namely structural hills, residual 
hills, piedmont, pediment, inselberg , buried pediplain, lateritic 
hill, and 
755