o study the 
et al, 1973, 
assessment 
sand, 1986; 
subsidence 
al data in 
a study was 
mining on 
1-deposition 
ase used in 
with path - 
d Survey of 
ound truth 
atic visual 
ISS-III and 
'aphic maps 
data fusion 
1 Survey of 
re digitized 
rated using 
M was used 
r deriving 
] analysis, 
ial data sets 
> identified 
ng sets for 
lihood per- 
nitially, the 
need to be 
sova et al. 
1e erosion- 
the spatial 
of sediment 
nodel is as 
(1) 
ea 
of steepest 
rojected to 
the runoff, 
s was used 
neration of 
model in 
)vided by 
'orporating 
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring”, Hyderabad, India, 2002 
necessary modifications. The maximum rainfall recorded in the 
area during the past 30 years was used for predicting runoff, 
which in turn used for computing erosion. Since the objective of 
the study was to assess the impact of mining on agricultural lands, 
the erosion from the mining areas only was computed while 
setting the detachment from othef® areas to zero, such that the 
deposition zones show only areas where the deposition from 
overburden material is taking place. Subsequently, the erosion - 
deposition index was arbitrarily sliced into various regions for 
better appreciation of the magnitude of topographic influence on 
erosion - deposition process. The erosion - deposition grid was 
overlaid onto land use / land cover grid to identify the cultivated 
areas where deposition of eroded material from the mining areas 
is taking place. For validation of the results, a ground truth 
campaign was subsequently launched. Runoff samples from 
various pour points were simultaneously collected for a single 
rainfall event to objectively compare the sediment concentration 
in the runoff vis-a-vis land use / cover pattern of the catchment of 
a pour point. It would have been better, in fact, if a systematic 
runoff and sediment measurements were to be made after 
establishing gauging stations, which is a cost-prohibitive apart 
from involving a lot of logistics. 
The positive values in the resultant erosion-deposition image 
(Fig-2) indicate the areas experiencing erosion while negative 
values indicate the deposition zones. The area statistics indicate 
that an estimated 219ha of total area is subject to deposition of 
material from the mining areas. Out of which, only 10ha of land is 
under agriculture. In the subsequent ground truth mission that was 
launched to validate the results, it was observed that the 
depositional areas near the foothill zones are matching well with 
the predicted ones. However, in the cultivated areas, the 
deposition of sediments from mining areas had been taking place 
earlier also. Due to adoption of appropriate soil conservation 
measures, the deposition of sediments has been considerably 
dropped during recent years. 
The information collected from farmers reveals that the acidity in 
the cultivated areas has been considerably increased (pH lowered) 
in pockets where deposition of sediment from mining took place 
with an attendant reduction in the paddy yield. Further, the subtle 
variations in the topography which plays a key role in erosion — 
deposition process could not be derived from the DEM that was 
generated by interpolation of contours (10m interval) from the 
1:25,000 scale topographic maps. 
Field observations reveal the fact that the sediment concentration 
from mining areas is very high as compared to undisturbed areas, 
which are protected well with the natural vegetation cover. 
Further, an observation from the forest watershed attests this fact. 
For instance, the sediment concentration of 5.98 g/l was observed 
in mining areas whereas it was 0.01 g/l in case of full forest cover 
pointing thereby to the role of surface cover in arresting the soil 
loss. The concentration of sediment is nearly proportional to the 
extent of mine/ ore dumps. In addition, an increase in the 
sediment concentration in the runoff water has also been observed 
with an increase in the over burden material from mining areas in 
the catchment. The soil separate analysis of sediments from 
mining areas shows that the coarse sand and gravel constitute the 
major component. In contrast, the areas where conservation 
practices have been adopted, only finer material (clay and fine 
silt) has been observed in the runoff. Since the overburden 
733 
material is composed mainly of coarse sediments, conservation 
practices that trap the sediment are necessary to reduce their 
loading into runoff, preventing thereby the siltation of water 
bodies and streamlets. 
2.3. Delineation and Monitoring of Aquaculture 
Owing to association with water, which exhibits the maximum 
absorption of incident radiation in the near-IR region, areas where 
aquaculture is practiced could be delineated very effectively and 
their dynamics studied using optical sensor data from Landsat- 
MSS and TM, and SPOT -MLA (Quader et al., 1986; Shahid et 
al., 1992; Vibulstesth et al, 1993 Venkataratnam et-al., 1997). 
Furthermore, more often than not, aquaculture is practiced along 
the coast and delta by utilizing brackish water. However, 
wherever available, fresh water is also used for aquaculture 
especially for raising fish. Co-existance of fresh water and 
brackish water aquaculture, which is associated with prawn 
culture, poses a problem with respect to delineation of its 
components. In the East coast of India, fish ponds and prawn 
ponds have characteristic shape, which enables their segregation 
apart from spectral information. The study was taken up to 
delineate fish and prawn culture ponds and monitoring the extent 
of aquaculture in part of Krishna district of Andhra Pradesh, 
southern India using Landsat-TM, SPOT-MLA and IRS-1C LISS- 
III data. 
The approach essentially involves geometric correction, database 
preparation and systematic on-the—screen visual interpretation of 
both concurrent as well as historical space-borne multispectral 
and multi-temporal digital data. For sub-categorization of 
aquaculture areas, LISS-III and PAN data were resampled to 6m 
spatial resolution and were fused using principle component data 
fusion technique, which was subsequently used for further 
interpretation. 
Initially, the spaceborne multispectral data of 1986 were‘ 
displayed onto colour monitor and the areas where aquaculture 
is practiced were delineated manually as a vector coverage vis-a- 
vis ground truth. In the satellite image the clusters of aquaculture 
ponds could be seen in different shades of the blue along the 
swales and the coastal plains, and follow the palaeo coastlines. 
Similar exercise was carried out for satellite data acquired during 
1988, 1997 and 2001 data sets too. 
The area statistics of aquaculture areas was subsequently 
generated for all the four periods. The results clearly bring out the 
fact that a very large chunk of crop land especially paddy lands 
has been converted into aquaculture. Consequently, area under 
aquaculture has increased five folds during the period 1986 to 
2001 i.e., 3,250 ha in 1986 versus 17,674 ha in 2001, at the cost 
of paddy lands. Such a phenomenon is highly deleterious to the 
environment especially to surface and ground water apart from 
soils. 
Since the damage caused by prawn cultivation is more serious, 
owing to the presence of abundant quantity of salts, than fish 
cultivation, it is necessary to segregate them out. In order to have 
a further insight into identification of type of aquaculture, the 
IRS-1C LISS-III and PAN merged data were used. During ground 
truth mission it was observed that the ponds wherein prawn 
farming is practiced are narrow and long when compared to