• A flexible modelling framework for the application of CSM to European countries, including flexible rainfall generation models, rainfall-runoff models, and objective selection criteria depending on data availability.
• A catalogue of data requirements for the application of CSM.• A framework for assessing the uncertainty of CSM estimates.
Estimation of design floods using Continuous Simulation Modelling (CSM) has emerged as a very active research topic across academic institutions in Europe. CSM is based on the use of rainfall-runoff models, of various complexities, for transforming a precipitation data into river flows. By coupling a rainfall-runoff model with a stochastic rainfall model, Monte Carlo simulations can generate long series of synthetic rainfall which can be transformed into river flow from which the flood frequency characteristics can be deducted.
Compared to the statistical approach, CSM is both more numerically complex and require significantly more input data, but has a number of advantages not available from a direct statistical analysis. Firstly, observed flow records are often limited and cover relatively short periods of time while rainfall data (driving CSM) are generally longer and more abundant. Secondly, if the chosen rainfall-runoff model is sufficiently flexible, it can potentially be applied to any catchment regardless of complexity, which is not always possible using simpler techniques where significant approximations might be necessary. Finally, the use of more physically based rainfall-runoff models should lead to more reliable predictions in ungauged catchments and for sensitivity analysis offlood characteristics to changes in land-cover, urbanisation and climate change.
The use of CSM is still largely confined to academia. FloodFreq will propose a framework for the application of CSM to European countries, by explicitly referring to the benchmark datasets compiled in D.2.1 described above. Stochastic rainfall models and rainfall-runoff models will be suggested for the use in CSM depending on data availability. A framework for assessing the uncertainty of the estimates will be developed, based on multivariate statistical fitting of estimation errors and selected explanatory variables.
Minute of WG3 meeting - Bratislava (November 25, 2011)
Minutes of Budapest WG3 meeting (June 16-17, 2011)