Systematic analysis of uncertainty in flood inundation modelling

摘要: Recent evaluations of 2D models have analysed uncertainty in data inputs into flood models, but treated the model code as a black box. In this work, influence numerical representation on results is evaluated. The purpose not only to understand significance physical scheme results, also importance respect other known sources uncertainty, particular boundary conditions, calibrated parameters such Manning’s friction values, DEM accuracy and more subjective forms associated with choices used by modellers constructing building representation. To further explore impact that level has output, were using risk exposure based measures. methods included vulnerability weighted measures use damage curves from Multi Coloured Manual. A series Monte Carlo tests undertaken for range over 3 test cases LISFLOOD-FP code. was chosen it several formulations solving floodplain flow within its framework, each different representation. The two urban events, culvert overtopping event Glasgow canal embankment failure Coventry, river Mexborough, Yorkshire rural domain. provided wider hydraulic conditions are reflected events typically assessed inundation ensure effect bias removed results. The indicated choice most critical affecting particularly case. However, interaction between factors certain scenarios, becomes less results. identified areas variations sets schemes traditional evaluation techniques. Recent