Enhancement and comprehensive evaluation of the Rating Curve Model for different river sites

SummaryThis study presents an enhancement of the Rating Curve Model (RCM) proposed by Moramarco et al. (2005) which was initially formulated to assess discharge at a downstream river site where only stage is monitored while the flow is recorded at an upstream section and significant lateral inflow can occur along the river branch. The original model formulation is here of fact extended for upstream discharge estimate by assuming the flow known at a downstream site and stages measured at both ends. In this new configuration, the model can be applied to river reaches with negligible lateral inflow or to river reaches with significant lateral contribution but where a kinematic flow regime holds. The new model formulation is tested by considering two case studies, both selected in the Upper-Middle Tiber River basin, in central Italy. For the first case study, the model is applied to 22 flood events observed along three branches of the Tiber River with accurate rating curves at ends and having the same upstream river site. RCM successfully simulates the discharge hydrographs observed at the upstream section closely capturing both the peak rate and the time to peak, with average absolute errors less than 5% and 0.36 h, respectively. The model accuracy is found independent of the intermediate basin area and, hence, of the lateral inflow contribution. The rating curve computed in this upstream section by using the shortest reach is slightly overestimated whereas the ones derived by applying the model to the other two reaches are almost coincident and slightly underestimated, but for all of them the errors are less than 5%.The second case study concerns the estimation of the upstream rating curve at a river site, where discharge measurements are available for very low stages alone, starting from the flow known at two different downstream equipped sections. The RCM application to recorded flood events provides two very similar stage-discharge relationships with a maximum difference, between the reproduced rating curves, of about 5 m3/s for the higher stage values.Moreover, the RCM accuracy is shown to be satisfactory for river reaches located in physiographic regions (northern Italy and Slovak Republic) different from the one used for the model development and testing and considered in the two case studies mentioned above.Finally, the model capability to take implicitly account of the unsteady effects (i.e. rating loop) is shown through experimental data recorded by an ultrasonic flowmeter at a gauged site of the Tiber River as well as by considering a numerical test.

► We present a deep investigation on the Rating Curve Model (RCM). ► RCM estimates the rating curve at river sites where only stage is recorded. ► RCM is enhanced for upstream discharge estimate. ► The model quantifies the lateral inflow and represents unsteady flow effects. ► RCM accuracy is shown for several case studies in different physiographic areas.