vai al contenuto della pagina vai al menu di navigazione

Rainfall-runoff modelling

Keywords: rainfall-runoff modelling; parameterisation; automatic calibration algorithms, data availability scenarios
Fig. 1 Feedforward neural network for rainfallrunoff forecasting (Author: E. Toth).
Fig. 2 Conceptual scheme of the distributed rainfall-runoff model AFFDEF (Author: E. Toth).
Fig. 3 A modular approach that uses system-theoretic rainfall-runoff models according to the situation characterising the forecast instant. (E. Toth)

The assessment of the parameters characterising a rainfall-runoff model is crucial for a reliable simulation of streamflow values. The problem is particularly critical for systemic models, that are based exclusively on the information attainable from the calibration data, but also the choice of a physically-based approach does not generally overcome the need to calibrate at least a part of the model parameters.

The research aims at obtaining indications on the quantity and quality of the calibration data that are needed for a reliable and efficient automatic parameterisation of rainfall-runoff models. Extensive experiments of calibrations and validations of rainfall-runoff models for different scenarios of historical information availability: different meteorological input variables, different methods for estimating the meteorological fields, different spatial and temporal scales of both input and output variables, different length and information content of the calibration record. Such aspects have been considered in relation to a variety of rainfall-runoff models, of deeply different nature: a physically-based distributed model, a conceptual lumped model, systemic, data-driven models based on Artificial Neural Networks.

Furthermore, the Whittle maximum likelihood estimator was proposed for calibrating the parameters of hydrological models. This method may represent a valuable opportunity in the context of ungauged or scarcely gauged catchments. In fact, the only information required for model parameterization is the spectral density function of the actual process simulated by the model. When long series of calibration data are not available, the spectral density can be inferred by using old and sparse records, regionalization methods or information on the correlation properties of the process itself.

Finally, an innovative regional parameterisation approach is proposed, based on the match, in the optimisation process, of a set of streamflow statistics. Such an approach allows the parameterisation of the model also for ungauged basins, based on the regionalisation of the selected statistics as a function of the climatic and geomorphologic characterisation of the watershed.

Main publications

Brath A., A. Montanari, E. Toth, Analysis of the effects of different scenarios of historical data availability on the calibration of a spatiallydistributed hydrological model. Journal of Hydrology. vol. 291, pp. 232 - 253 ISSN: 0022- 1694, 2004.

Castiglioni, S. Lombardi, L., Toth, E., Castellarin, A., Montanari, A., Calibration of rainfallrunoff models in ungauged basins: A regional maximum likelihood approach, Advances in Water Resources, 33, 1235–1242, 2010.

Lombardi, L., Toth, E., Castellarin, A., Montanari, A., Brath, A., Calibration of a rainfallrunoff model at regional scale by optimising river discharge statistics: performance analysis for the average/low flow regime, Physics and Chemistry of the Earth, 42–44, 77–84, 2012.

Montanari A., E. Toth, Calibration of hydrological models in the spectral domain: An opportunity for scarcely gauged basins?. Water Resources Research. vol. 43, pp. W05434 - . ISSN: 0043- 1397, 2007.

Moretti, G., e Montanari, A., Affdef: A spatially distributed grid based rainfall-runoff model for continuous time simulations of river discharge, Environmental Modelling & Software, Vol. 22(6), 823 – 836, 2007.

Mount, N.J., Maier, H.R., Toth, E., Elshorbagy, A., Solomatine, D., Chang, F-J., Abrahart, R.J., Data-driven modelling approaches for sociohydrology: opportunities and challenges within the Panta Rhei Science Plan, Hydrological Sciences Journal, 61:7, 1192-1208, DOI: 10.1080/02626667.2016.1159683, 2016.

Toth E., A. Brath, Use of spatially-distributed or lumped precipitation inputs in conceptual and black-box models for runoff forecasting. In: A. Brath A. Montanari E. Toth (Eds). Recent advances in peak river flow modelling, prediction and real-time forecasting - Assessment of the impacts of land-use and climate changes. (pp. 247 - 261). Ed. Bios (Italy), 2004.

Toth, E., Brath A., Multistep ahead streamflow forecasting: Role of calibration data in conceptual and neural network modeling, Water Resour. Res., 43, W11405, doi:10.1029/2006WR005383, 2007.

Toth, E., Classification of hydro-meteorological conditions and multiple artificial neural networks for streamflow forecasting, Hydrology and Earth System Sciences, 13, 1555–1566, 2009.

Research projects

5-years research contract (2013-2018) with the water supply consortium Romagna Acque – Società delle Fonti SpA

Italian Research Project of National Relevance 2005, “Characterisation of average and extreme flows in ungauged basins by integrated use of data-based methods and hydrological modelling”, financed by the Italian Ministry of University and Research (MIUR).

Italian Research Project of National Relevance 2008, “Uncertainty assessment of rainfall and streamflow measurements and impacts on the management of water scarcity conditions”, financed by the Italian Ministry of University and Research (MIUR).