IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring”, Hyderabad, India, 2002 
fishponds. Hence, an attempt was made to delineate them based 
on their shape. Since the bunds of aquaculture ponds are very 
clear and sharp in the LISS-III and PAN- merged image, the 
dataset was used to digitize the boundaries of ponds. 
Subsequently, the boundaries were labeled and topology was 
constructed. A simple shape parameter, defined below, was used 
to segregate prawn culture ponds from fish ponds : 
Area of the polygon 
Shape factor = (2) 
Perimeter of the polygon 
  
Due to long and narrow shape, as pointed earlier, the prawn 
culture areas exhibit a lower shape factor as compared to fish 
ponds. A few ground truth observations were used. to fix the 
threshold value. A threshold value of 18.0. was found to be 
optimal for segregation of fish ponds from those of prawn 
cultivation areas. The segmentation of aquaculture areas using 
shape parameter along with LISS-III and PAN merged data is 
appended as Fig-3. Accuracy estimation made through total 
enumeration approach indicated that the fish and prawn ponds 
could be segregated with a fairly good accuracy (82.696) using 
shape parameter (Table-1). Further refinement in the exercise may 
help improving the accuracy of delineation. 
3. CONCLUSIONS 
From the foregoing, it is evident that the information derived from 
remote sensing data when used in conjunction with GIS can play 
an important role in the detection, delineation and monitoring of 
degraded lands. The output thus generated from thermal IR (10.4- 
12.5um) measurements made in the night have shown quite 
encouraging results with respect to detection of waterlogged areas 
due to rising ground water table. Satellite data acquired at the end 
of crop season enables detection of waterlogging quite efficiently. 
However, delineation of irrigated areas from potentially 
waterlogged areas is still a challenging task. Further, potential of 
multiple channel thermal data and passive microwave need to be 
exploited in this endeavor. Similar studies need to be carried out 
under different agro-climatic conditions to validate the results 
obtained in this study. 
Table- 1 Confusion matrix for prawn and fish ponds. 
  
  
  
  
  
  
  
  
a REFERENCE DATA 
Ss = Prawn Fish Total 
e = Prawn 282 114 396 
o Fish 107 764 871 
Total 389 878 1267 
  
  
  
  
Overall accuracy = (2824-764) / 1267 - 82.696 
Producers Accuracy: Prawn - 282/389 - 72.590 
Fish 767 / 878 2 87.4% 
User's Accuracy : Prawn = 282/396 - 71.296 
Fish = 767/871 - 88.196 
Il 
For assessment of land degradation due to mining, spatially 
distributed models have been found quite useful. Since the 
information on vegetation and soil plays a key role in arresting 
soil loss by water, such information derived from satellite data 
734 
forms a valuable input in studying the spatial distribution of 
erosion — deposition pattern in a watershed. It also emerges from 
the study that the prediction of erosion-deposition pattern made 
using UPDEP model has been reasonably well. Since the 
topographic information derived from contours is interpolated and 
could not explain micro-variâtions in topography, stereo pairs 
from high resolution satellite data could form a better database for 
detailed level delineation of erosion-deposition zones. Further, in 
the present erosion-deposition model there is no provision for 
incorporating the conservation practices like silt trapping tanks, 
gully plugs, check dams, etc., and hence a continuous cell by cell 
model that accounts erosion-deposition rates need to be 
developed. 
Developments in automated object recognition techniques based 
on shape and color could help in digital classification of the fish 
and prawn ponds, which is very important from environmental 
health point of view. Attempts need to be made to integrate 
thematic information in Geographic Information System (GIS) 
environment to model and quantify the impact of aquaculture on 
soil environment. 
ACKNOWLEDGEMENTS 
The authors are profusely thankful to Dr. R.R. Navalgund, 
Director and Dr. A.Bhattacharya, Dy. Director, NRSA for 
providing financial support and suggestions made during the 
course of study. We are also thankful to Dr. L. Venkataratnam, 
Group Director, Agriculture & Soils Group for evincing keen 
interest. Thanks are also due to Mrs. G. Sujatha and E. 
Shankaraiah for laboratory support. 
REFERENCES 
Anonymous, 1976. Report of the National Commission on 
Agriculture. Chapter-V, Resources Development. Ministry of 
Agriculture, Govt. of India, New Delhi. 
Chase, M.E., 1969. Airborne remote sensing for groundwater - 
studies in prairie environment. Canadian Journal of Earth Sci., 6, 
pp. 737-741. 
Concklin, H. C., 1957. Hanunoo agriculture. FAO, Forestry 
Development paper No.12, FAO Rome, pp.109. 
Das, D.C., 1985. Problem of soil erosion and land degradation in 
India. Proceedings of the National Seminar on Soil Conservation 
and Watershed Management, New Delhi. 
Heilman, J.L. and Moore, D.G., 1982. Evaluating depth to 
shallow groundwater using Heat Capacity Mapping Mission 
(HCMM) data. Photogrammetric Engineering and Remote 
Sensing, 48(12): 1903-1906. 
Huntley, D., 1978. On the detection of shallow aquifers using 
thermal infrared imagery. Water Resources Research, pp. 1075 - 
1083. 
Irons, J. R. Lachhowski, H., and Peterson, C., 1980. Remote 
sensing of surface mines: A comparative study of sensor systems. 
Proceedings of the 14" International Symposium on Remote 
Sensing of the Environment. San Jose, Costa Rica, April, 1990, 
re Al ^ at = 
ba FA US OP ped 
DN fr wena o JAN