International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004 
  
rainfed cultivation lands and settlements in the area. The 
reasons for the abandonment of agricultural lands might be due 
to: 1) migration of dwellers; 2) low production from the lands; 
and 3) shifting cultivation. It is very common practice seen in 
Nepal (Awasthi er al., 2002). 
Similarly, in 1999, all the landuse/landcover classes excluding 
class 3 had increased by 0.2 km in whole watershed, whereas, 
class 3 in TOWs, SOWs and FOWSs had decreased by 0.3, 0.1 
and 0.1 km”, respectively. This is because of the significant 
expansion of floodplains near the confluence with Rapti River, 
which does not attribute/share to any sub-watershed (Figure 1). 
The percentage areas of class 3 contributed from FOWs, SOWs 
and TOWs to whole watershed, are 6.7 96, 8.6 % and 37%, 
respectively, in 1988; and 10.8%, 10.5% and 40%, respectively, 
in 2003. These results suggest the development and expansion 
of landuse/landcover class 3 areas (mostly floodplains) are 
predominantly larger in higher order watersheds, which cover 
main channel and tributaries. Two third order watersheds which 
are derived from two main tributaries of Thadokhola River, 
Bhundrun and Goganghari, occupies whole channel area of 
these two tributaries. Thus, it is obvious that TOWs comprise 
larger area of floodplains. In contrary to that SOWs covers only 
few portions of these two tributaries since SOWs contribute 
only 56% of its total area to TOWs, thus have less floodplain 
areas in compare to TOWs. 
The dynamics of expansion and contraction landuse/landcover 
class 3 are mostly due to the internal trading among the landuse 
classes: cultivation lands, sparse forest and floodplains. 
Specifically, expansion and development of new floodplains are 
observed in the rice fields that were cultivated along the river 
banks, and some parts of sparse forests. 
Development of floodplains/eroded lands is hardly seen on 
dense forests. The reasons for expansion of landuse/landcover 
class 3 are mainly because of soil erosions taken place on both 
cultivated lands and on sparse forest floors. The red soil of the 
study area is very susceptible to rain and wind erosion. From 
the field survey, it was confirmed that the sparse forest floors 
are very poorly vegetated. Collection of litters for composts, 
fuels and fodders, and damage to understudy vegetation due to 
grazing, have impaired the surface of forest floor from resisting 
erosive forces exerted during rainy period. Furthermore, the 
root morphology of sal (Shorea robusta), which is the main 
component of forest in the study area, has extremely low effects 
in protecting surface erosion. Sal possess stake-like main root 
which is very thick and strong enough to penetrate hardpans of 
red soils (Laterite soil) but, it has enormously few lateral and 
fine/hair roots which are very useful in protecting surface 
erosions (Shrestha ef al., 2001). 
4. CONCLUSION 
The geomorphometric analyses revealed that the watersheds are 
susceptible to surface erosion, but inert to sediment 
transportation. Landuse/landcover changes incorporated with 
geomorphometric parameters analysis is useful in investigating 
watershed characteristics, in which a wide variety of analytical 
results can be effectively integrated using GIS for explaining 
the current situation and proposing future plans for integrated 
soil conservation and watershed managements. The ability of 
GIS to integrate the digital maps of land use derived from 
interpretation of satellite images with DEM also provides a new 
opportunity for analysing patterns of long term landuse 
568 
dynamics with respect to the geomorphometric characters. 
Furthermore, several similar studies on different parts of 
Siwaliks need to be carried out in order to have conclusive 
characteristics of whole Siwaliks regions. 
4. REFERENCES 
Awasthi, K.D., Sitaula, B.K., Singh, B.R. and Bajaracharya, 
R.R., 2002. Land-use change in two Nepalese watersheds: GIS 
and Geomorphometric analysis, Land Degrad. Develop. 13:495- 
513. 
Shrestha, M.B., 2001. Study on restoration of vegetation for 
conservation of the dilapidated mountaineous regions of Nepal, 
Ph.D thesis, United graduate school of agricultural sciences, 
Gifu university, Japan PP 243. (in Japanese, unpublished). 
Shrestha, M.B., Yamadera, Y. and Miyazaki, T., 2001. A study 
on root morphology of the plants that were introduced to 
stabilize steep roadcut slopes of the Siwalik region in Nepal, 
Journal of the Japanese Society of Revegetation Technology, 
vol. 27 (2). Pp. 416-429. (in Japanese with English summary). 
KEY W 
ABSTR 
Data int 
data. Th 
study c: 
differen 
solution 
construc 
Each stı 
In recen 
multiple 
intellige 
Llinas, 
of inten 
1997; I 
Clark ai 
1997). 1 
from di: 
judgmer 
domains 
absolute 
become 
Howeve 
multiple 
One of t 
determir 
Dependi 
divided 
and deci 
1997). T1 
directly 
level fu 
extractec 
available 
fusion : 
interpret 
source o 
Several 
applied, 
various 
remote x