Rainfall-runoff model parameter conditioning on regional hydrological signatures: application to ungauged basins in southern Italy

Abstract

Parameter estimation for rainfall-runoff models in ungauged basins is a key aspect for a wide range of applications where streamflow predictions from a hydrological model can be used. The need for more reliable estimation of flow in data scarcity conditions is, in fact, thoroughly related to the necessity of reducing uncertainty associated with parameter estimation. This study extends the application of a Bayesian procedure that, given a generic rainfall-runoff model, allows for the assessment of posterior parameter distribution, using a regional estimate of ‘hydrological signatures’ available in ungauged basins. A set of eight catchments located in southern Italy was analyzed, and regionalized, and the first three L-moments of annual streamflow maxima were considered as signatures. Specifically, the effects of conditioning posterior model parameter distribution under different sets of signatures and the role played by uncertainty in their regional estimates were investigated with specific reference to the application of rainfall-runoff models in design flood estimation. For this purpose, the continuous simulation approach was employed and compared to purely statistical methods. The obtained results confirm the potential of the proposed methodology and that the use of the available regional information enables a reduction of the uncertainty of rainfall-runoff models in applications to ungauged basins.

Improved Medium- and Long-Term Runoff Forecasting Using a Multimodel Approach in the Yellow River Headwaters Region Based on Large-Scale and Local-Scale Climate Information

“…Djibo et al [12] used a probabilistic approach statistical seasonal streamflow forecasting over West African Sahel; the input variables were sea level pressure, relative humidity, and air temperature. Biondi and De Luca [13,14] considered a simple lumped and conceptual rainfall-runoff model for design flood estimation in gauged and ungauged catchments in southern Italy, by considering as input 500-years of 20-min synthetic rainfall data, derived from a daily rainfall generator and a specific downscaling procedure for Southern Italy. Previous studies primarily focused on precipitation and antecedent runoff as the inputs for runoff forecasting, however, less attention was paid to fully incorporate large-scale climate variables and other local-scale climate variables as the inputs.…”

Bivariate Return Period for Design Hyetograph and Relationship with T-Year Design Flood Peak

“…The catchment response has been modelled using a lumped and event-based RR model [26,32] composed by the Soil Conservation Service Curve Number (SCS-CN) method [33] for rainfall excess estimation, and by Nash's model [34] for runoff estimation at basin outlet. Although the SCS-CN method is a popular (due its relative simplicity and its reliance on a limited number of parameters, resulting in a robust tool for those catchments that are partially or poorly gauged) and ubiquitous means for estimating storm runoff from a given rainfall event, i.e., as an alternative to empirical or physically-based infiltration models like Green-Ampt or Horton's equations, it should be noted that its usage at the sub-daily time scale is controversial [35].…”

“…As a final step of this study, the RR model, introduced in Section 4, was calibrated with the available time series of rainfall and discharge data according to a procedure placed in the context of Bayesian inference described in Biondi and De Luca [32]. Specifically, the assessment of posterior distributions of model parameters were derived on the basis of the closeness of simulated hydrological signatures obtained with different parameter sets to those derived from actual at-site observations.…”

Predicting Runoff Signatures Using Regression and Hydrological Modeling Approaches

“…They include indices that reflect streamflow distribution such as mean annual runoff and flow percentiles and indices that reflect streamflow dynamics such as the slope and concavity index of the flow duration curve. These indices are important for many applications including characterizing and classifying catchments (Ssegane et al, 2011;Wagener et al, 2007), predicting runoff (Shu & Ouarda, 2012;Zhang et al, 2014), water resources assessment, planning and management (Biondi & De Luca, 2017;Smakhtin, 1999), and ecological and environmental flow assessments (Beck & Birch, 2012;Bozec et al, 2005;Newcomer et al, 2012).…”