1. Aynalem T. Tsegaw and Knut Alfredsen
Comparison of continous time rainfall-runoff model, event based model and
statistical methods to estimate floods at ungauged rural small catchments
Aynalem T. Tsegaw and Knut Alfredsen
(Tel.: [+47] and Fax: [+47] ); addresses:
2. Back ground of estimation methods
2.1 Rational and Statistical methods
Advantages
Based on simple assumptions and simple to use
Uses the observed data for analysis
Disadvantages
Contains parameters which are dificult to estimate (e.x. runoff coefficient)
Do not consider antecedent moisture condition
Based on stationarity assumption
Based on questionable assumptions (ex.P200 = Q200)
2.2 Continous time rainfall-runoff model –DDD
Incorporates the moisture and ground condition prior to flood producing preci. event
Able to include the contribution of ground water
Produces hydrographs(not only peak flood)
Disadvantages /challenges
Lack of availability of high spatial resolution of geographical data
Lack of availability of high spatial and temporal resolution of hydro-meteorological data
3. Methodology used
3.1 Ungauged rural small catchments
( < = 50km2)- Regionalization of DDD
Have short response times (few hours or even sub-hourly )
It is difficult to setup hydrological models
Model parameters need to be transfered from gauged catchments
3.2 Distance distribution dynamics(DDD) model parameters-Regionalized hourly
Multiple regression, physical similarity and combined methods
4. Results
4.1 Regionalization methods
Multiple regression is best of all the three regionalization methods tested
Physical similarity performed the poorest
4.2 Flood estimation methods
Objectives:
1. 1 To compare flood estimation methods
Event based model- Rational formula
Statistical method – NIFS formula
Continous time rainfall-runoff model –DDD
1.2 To select appropriate mthod to study:
Impacts of future landuse changes
Impacts of future climate changes
200years, 1 hr duration precipitatin 23.4mm/hr
Case1. August 2013 when:
The subsurface is 32.2% saturated
No snow mel
Case2. August 2013 when:
The subsurface is 100%
saturated
No snow melt
Case3. May 2013 when:
The subsurface is 100% saturated
There is snow melt
Rational NIFS formula DDD
Small rural ungauged catchment (1.5km2)-used for case study
5. Conclusion
Continuous rainfall-runoff hydrological models have the potential in studying flood risk management in small rural ungauged catchments under future land use and climate with better understanding of the hydrology leading up to the flood event.
Acknowledgements
We acknowledge the Research Council of Norway in funding the research work under the project