Quantifying uncertainty in flood predictions due to river bathymetry estimation

Abstract. River bathymetry is important for accurate flood modelling but often unavailable due to the time-intensive and expensive nature of its acquisition. This leads to several proposed and implemented approaches for its estimation. However, the errors in measurements and estimations inherent in these methods, affecting the accuracy of the flood modelling outputs, are not extensively researched. We investigate the sensitivity of flood predictions to these errors in two formulas: the Uniform Flow and the Conceptual Multivariate Regression. Given channel slope, width and bank-full discharge, these formulas can be used to estimate bathymetry. However, errors in estimated bathymetry will affect the flood results. We employed a Monte Carlo framework to introduce random errors into these parameters drawn from a normal distribution with zero mean and a standard deviation set to 10 % of their best estimates. Using this process, we generated 50 simulated river bathymetries for each parameter along with an additional 50 where the errors were applied to all parameters simultaneously. The riverbeds generated from these bathymetries were combined with topographic LiDAR data to create model grids. Each grid was used in the hydrodynamic model LISFLOOD-FP to simulate the 2005 flood event in the Waikanae River area of New Zealand. We assessed the resulting flood predictions for their variability and sensitivity. The results indicate that, between the two methods, the combined errors in the parameters using the Uniform Flow formula are associated with greater uncertainty in flood depths (median error: 3.89 m, quartile range: 2.36 to 7.78 m) and extents (208.72 ha, 206.59 to 209.58 ha), compared to Conceptual Multivariate Regression (depth: 3.61 m, 2.32 to 7.37 m; extent: 207.82 ha, 206.42 to 208.48 ha). Among the parameters, the width errors correspond to the highest uncertainty, while the slope errors correspond to the lowest.