Abstract:
A grid based Regional Scale Sediment Yield Model RSSYM was setup different
catchments of Indus basin using coarse resolution grid data. When coarse grid data is
used, the land slope decreases and sediment delivery is reduced. In this study the
slope averaging effect was minimized by using the fractal analysis approach. The
fractal constant (α) and fractal dimension (D) equations were correlated with the
standard deviation on elevation by Zhang et al. (2000). These equations for fractal
constant and fractal dimension were developed using Digital Elevation Model of 1
km2 resolution (DEM) for local topography. The slope was computed for 250 m and
50 m downscaling from 1 km2 grid. A comparison was made between fractal constant
equation and fractal dimension to scale down the slope parameter. Both equations
were incorporated in RSSYM. The model was applied to Phulra catchment. The
results showed that fractal dimension equation gives better results as compared to
fractal constant. So the fractal dimension equation was incorporated in RSSYM.
The RSSYM was applied on three catchments of Indus basin namely Soan, Gariala
and Phulra. For Phulra catchment of Siran river the total observed sediment was 0.304
million tons whereas the simulated sediment was 0.291 million tons. The coefficient
of efficiency (COE) was 0.85 and the coefficient of determination was 0.83 which
shows that there is a good correlation between the observed and simulated values.
In Gariala catchment the observed sediment was 6.01 million tons whereas simulated
sediment was 6.80 million tons. The coefficient of efficiency for Gariala catchment
was 0.91 and the coefficient of determination observed and simulated hydrograph was
0.81. Similarly for the Soan catchment the observed sediment was 10.61 million tons
and simulated sediment was 12.29 million tons. The coefficient of determination and
coefficient of efficiency for this catchment was 0.95 and 0.98 respectively.
The snowmelt runoff model SRM was applied to Astor catchment to test the
applicability of temperature index approach for Indus basin. The terrain is difficult to
measure the hydrological and hydraulics data. Most of the data was available at the
outlet of the catchments. The input data included daily temperature and precipitation,
dividing the catchment in to different zone depending on the elevation difference. The
elevation zones were generated from the DEM of the area. The snow depletion curves
were made for the snow covered area derived from the satellite data analysis on
monthly basis. The model output was discharge hydrograph. The COE was 0.91
which shows that there is a good correlation between the observed and simulated
values. The statistical test showed that model performance was good. The results of
SRM model encouraged to use temperature index approach for snowmelt runoff
estimation in the Indus basin.
A snowmelt runoff model was developed using temperature index approach. The
model was coupled with regional scale sediment yield model RSSYM. The coupled
model was named as modified regional scale sediment yield model (MRSSYM). The
MRSSYM model was applied to Astor and Gilgit catchments.
For the Astor
catchment total observed sediment load was 3.98 million tons whereas the simulated
sediment load was 4.34 million tons. The coefficient of the model was 0.89 whereas
the coefficient of determination was 0.83. Similarly for Gilgit catchment the measured
and simulated sediment loads were 4.50 and 4.48 million tons respectively. The
coefficient of efficiency and coefficient of determinations were 0.95 and 0.88
respectively. From the results it can be concluded that MRSSYM can be applied with
confidence to various catchments of Indus basin where runoff is due to snowfall and
snow melting.