Abstract:
Floods are the most critical among all the natural calamities in world causing vast damages to
life and property. In Ethiopia many areas are under flood problem and Jimma is like other
tropical cities in Ethiopia facing flash flood so flood hazard and risk mapping is important. The
HEC-HMS hydrologic modeling system software requires hydrological data for rainfall runoff
modeling. The hydrological data of 16 years (i.e. 1995-2011) were collected from Ethiopian
meteorological agency and ministry of water, irrigation and energy. Normal ratio method was
used for filling missing value of precipitation data and data consistency was checked up using
double mass curve. Other parameters like curve number and basin lag time were generated
using Hydrologic Engineering Center-Geo spatial Hydrologic Modeling System which is an
interface between Arc Geographic Information System and HEC-HMS. Soil Conservation
Service-Curve Number, Soil Conservation Service-Unit Hydrograph, monthly constant and
Muskingum methods were chosen for precipitation loss modeling, excess precipitation
transformation to direct runoff, base flow modeling and flood routing respectively. Among the
collected a 16 years hydro-meteorological data for rainfall runoff modeling, 10 events (1995-
2005) were used for model calibration and 6 events (2006-2011) were used for model validation.
The model performance was evaluated using performance measuring techniques including Nash
Sutcliff Efficiency (NSE) and Coefficient of Determination (R2
). Nash Sutcliff Efficiency during
calibration and validation was 0.77 and 0.7 respectively whereas coefficient of determination
during these two processes was 0.96 and 0.99 respectively. After the evaluation of model
performance, it was concluded that HEC-HMS indicated good performance for Awetu sub-basin
rainfall runoff modeling. Flood frequency analysis conducted using HEC-HMS’ frequency storm
method for 25, 50,250 and 500 year return periods. The peak flood for each respective return
periods were 13.2, 14.9, 18.9 and 21 m
3
/s respectively. Flood inundation mapping was modeled
for peak flood of each return period using HEC-RAS and inundation area 58.9ha, 60ha, 60.5ha
and 71.7 ha respectively.
