Received: 30 Aug 2019 – Discussion started: 05 Sep 2019
Abstract. Floods continue to devastate societies and their economies. Resilient societies commonly incorporate flood forecasting into their strategy to mitigate the impact of floods. Hydrological models which simulate the rainfall-runoff process are at the core of flood forecasts. To date operational flood forecasting models use areal rainfall estimates that are based on geographical features. This paper introduces a new methodology to optimally blend the weighting of gauges for the purpose of obtaining superior flood forecasts. For a selection of 7 Australian catchments this methodology was able to yield improvements of 15.3 % and 7.1 % in optimization and evaluation periods respectively. Catchments with a low gauge density, or an overwhelming majority of gauges with a low proportion of observations available, are not well suited to this new methodology. Models which close the water balance and demonstrate internal model dynamics that are consistent with a conceptual understanding of the rainfall-runoff process yielded consistent improvement in streamflow simulation skill.
How to cite. Wright, A. J., Robertson, D. E., Walker, J. P., and Pauwels, V. R. N.: Insights from a new methodology to optimize rain gauge weighting
for rainfall-runoff models, Hydrol. Earth Syst. Sci. Discuss. [preprint], https://doi.org/10.5194/hess-2019-450, 2019.
This paper details the development of a methodology to optimize the weighting of rainfall gauges for hydrologic simulation. In particular, catchments with a low gauge density and/or proportion of observations available are not well suited to this methodology. Application of this methodology with models that are consistent with a conceptual understanding of the rainfall-runoff process yield improvements of 7.1 % in evaluation periods.
This paper details the development of a methodology to optimize the weighting of rainfall gauges...