Abstract:
The main objective of the research was to reconstruct flow regimes of Borkena
catchment in ungauged locations for irrigation development using rainfall-runoff
modeling. Four black-box-type rainfall-runoff models, namely, the Simple Linear
Model, the seasonally based Linear Perturbation Model, the wetness-index
based Linearly Varying Gain Factor Model, and the Artificial Neural Network
Model, along with the conceptual Soil Moisture Accounting and Routing Model,
were used to test the hydrological response of the Borkena catchment.
The performance of these hydrological models for the study area was tested
and the best candidate model for the catchment response prediction was
selected. Artificial neural network model was selected as a robust rainfall-runoff
model to obtain the estimated flow of the gauged station which is the basis to
transfer flow data to ungauged sites on the catchment. The R
2
during calibration
and verification was 97.57% and 91.30% respectively.
On the basis of the selected model 15 days 75% dependable flow derived from
the flow duration was used as the basis of estimating dependable low flows at
ungauged locations of the catchment using area ratio method of transferring
flows.
