Merging and calibration of radar rain products for quantification of input uncertainty in urban drainage modelling for the Haute-Sûre catchment in Luxembourg

Authors

J. A. Torres-Matallana, F. Cecinati, V. Belos, and U. Leopold

Reference

in EnviroInfo 2017: From Science to Society: the Bridge provided by Environmental Informatics, Luxembourg, 2017

Description

Common practice in urban drainage modelling requires high temporal and spatial resolution of data inputs as land use, rainfall and runoff. In the case of rainfall input, the fast response to rainfall over small catchments has a direct effect on the temporal scale required to simulate with a good accuracy the hydrological and hydraulic dynamics processes. We present and illustrate a proposed work flow to build and calibrate a space-time geostatistical model of rain, using radar imagery as a covariate in regression-kriging based simulation. Then, a work flow for runoff and sewer system modelling is presented as well. These work flows are repeated as many times as the Monte Carlo simulation design requires to obtain an ensemble of rainfall input maps and time series of quantity variables (volume of the CSO tank, and CSO volume) and quality variables (loads and concentrations of COD and NH4) are produced to analyze how input uncertainty propagates to output uncertainty. We expect that these work flows represent a more realistic simulated time series of rainfall originating runoff that enters the sewer system, and we can do the Monte Carlo input uncertainty propagation and compare with results obtained in previous studies we did. In the future, we expect that validation will show a better job done not only in terms of mean error and root mean squared error of tank volume and overflow, but also in quantifying the uncertainty in the sewer system model outputs.