Abstract:
Observed discharge is one the crucial and
common input parameter required in rainfall runoff model
development and importantly, it should be the true
representative runoff of that river basin. When the stream
flow regime is regulated, it needs to be modified or corrected
before setting up the rainfall runoff models. To examine the
fact, the study was carried out in the Kharun river, a tributary
of Seonath river in Chhattisgarh state having strongly regulated flow regime. The non-regulated flow time series at
Patherdihi was simulated from the regulated observed flow at
Patherdihi G/D site by developing Kharun river basin model
in Mike Basin software using the information on storages,
addition, diversion and return flows from irrigation, etc. The
MIKE 11 NAM rainfall runoff models were developed under
regulated flow condition and non-regulated (virgin) flow
conditions. Accuracy of the models were examined on the
basis of coefficient of determination (R2), Efficiency Index
(EI), Sum of Square of Error (SSE) and percentage
difference between observed and simulated discharge. When
the model was developed under non-regulated condition, the
R2 value increased from 0.78 to 0.858, Efficiency Index
increased from 0.71 to 0.86 and the difference between
observed and simulated discharge reduced from 6.2 to 3.5%
and the SSE of the model was found reduced by 25%
indicating improved agreement between simulated and
observed runoff. The model accuracy was seen higher when
the model was developed using the non-regulated flow data. It
could be concluded that the rainfall runoff model simulation
using regulated flow data as an input may lead to inaccurate
simulation of river flow.