loading page

Flood-duration-frequency modelling with adaptive tail behaviour: A Bayesian approach
  • +1
  • Danielle Barna,
  • Kolbjørn Engeland,
  • Thordis L. Thorarinsdottir,
  • Chong-Yu Xu
Danielle Barna
Norwegian Water and Energy Directorate

Corresponding Author:[email protected]

Author Profile
Kolbjørn Engeland
Norwegian Water and Energy Directorate
Author Profile
Thordis L. Thorarinsdottir
Norwegian Computing Center
Author Profile
Chong-Yu Xu
University of Oslo
Author Profile


Flood frequency analysis is a statistical approach for estimation of design flood values. Design flood values give estimates of flood magnitude within a given return period and are essential to making adaptive decisions around land use planning, infrastructure design, and disaster mitigation. Flood magnitude is here typically taken as peak flow from an instantaneous discharge series. However, this univariate approach can be somewhat artificial as a flood event is not described by its peak flow alone. A relatively simple extension of traditional flood frequency models can be found in flood-duration-frequency, or QDF, models. QDF models take flood magnitude to be a product of peak flow and duration and are analogous to intensity-duration-frequency curves for precipitation. In an application to 12 locations in Norway, we assess how three different QDF models capture relationships between floods of different duration. Incorporating dependence on return period in the ratio between growth curves improves modeling of both short-duration events and events with long return periods. This model extension further expands the models' ability to simultaneously model a wide range of flood durations. Overall, we find the choice of durations used to fit the QDF model is a highly influential aspect of the modeling process. Users should be aware that the choice of which durations to fit the model with will always be a qualitative choice that is only partially mitigated by adding extra flexibility to the models.