Stochastic Hybrid Event Based and Continuous Approach to Derive Flood Frequency Curve

This study proposes a methodology that combines the advantages of event-based and continuous models, for derivation maximum flow hydrograph volume frequency curves, by combining stochastic weather generator (the advanced generator, abbreviated as AWE-GEN) with fully distributed physically based hydrological model TIN-based real-time integrated basin simulator, tRIBS) runs both simulation. The is applied to Peacheater Creek, 64 km2 located in Oklahoma, United States. First, set 5000 years’ hourly forcing series generated using AWE-GEN. Second, simulation 50 years climate tRIBS. Simultaneously, separation storm events performed applying exponential method 5000- 50-years series. From years, mean soil moisture top 10 cm (MSM10) layer at an time step extracted. Afterwards, from times MSM10, values associated all within are Therefore, each event has initial value (MSM10Event). Thus, probability distribution MSM10Event month year obtained. Third, five major terms total depth simulated framework tRIBS, assigning state Monte Carlo framework. Finally, annual hydrographs obtained peak-flow volume, curves derived. To validate method, results hybrid compared those deriving flood analyzing dependence between volume. Independence rainfall prior conditions been proved. proposed can reproduce univariate good agreement curve Nash–Sutcliffe coefficient 0.98, whereas 0.97. permits generate but reducing computation on order months hours.