Hydrologic and Hydraulic Modelling in the Meolo-Targa Drainage Basin

Lowland drainage basins are hydraulically complex because small topographic gradients, artificial drainage networks, local storage, and pumping-station operation strongly influ ence the water flow and potential flooding events. This study developed and calibrated a two-dimensional HEC–RAS Rain-on-Grid model of the Meolo–Targa drainage basin in the Province of Rovigo, northeastern Italy. The model provides a unified representation of rainfall-runoff generation, as well as of overland and channel flow, the main drainage net work, culverts, road embankments, and the operation of the Meolo pumping station. The 17 September 2022 rainfall event was used for event-scale calibration against observed water-surface elevations upstream of the pumping station. Calibration was first performed for the period preceding pump activation, allowing the rainfall–runoff response to be evaluated independently of pump operation. This initial phase allowed the estimation of rainfall-generation controlling parameters (Curve Num bers in the SCS runoffmodel), theparametersofthemodelsoflocalhydraulicphenomena, and hydraulic routing parameters. The final configuration achieved an NSE of 0.80 and an RMSE of 0.0084 m during the pre-pump period. The analysis of the system behavior after pump activation showed that the simulated response was strongly affected by the representation of the Meolo channel bed, connected channel storage, and the uncertain Pump1head–dischargerelationship. TheLiDAR-based channel representation produced relatively rapid filling of the channel system and of the intake basin, causing an excessive and unrealistic pump cycling, whereas the channel bed configurationmodifiedusingafew surveyed cross-sectionsproduced slowerreplenishment andfewer cycles with respect to observations. An intermediate channel-bed representation was therefore adopted, based on the modeled behavior. The final calibrated configuration used a channel Manning’s coefficient of 0.048 together with calibrated pump-control set tings. The NSE remained 0.75 when the evaluation was extended to a relatively short pe riod after pump activation and decreased to 0.48 over the complete 24-hour event, mainly because of differences in the timing of later pump-induced drawdown and recovery cycles. Beyond reproducing key features of the event-scale hydrologic and hydraulic response of a small mechanically drained lowland basin, this thesis shows the usefulness of a de tailed hydrologic–hydraulic model for understanding the dynamics of low-slope drainage systems, their main controlling factors, and the properties that require both accurate sur veying and direct monitoring. The study demonstrates that HEC–RAS 2D Rain-on-Grid can reproduce the main event-scale hydrologic and hydraulic behavior of a small mechan ically drained lowland basin, while also showing that model performance is strongly lim ited by uncertainty in channel geometry and pumping-station characteristics. The results of the thesis, in fact, point to specific sectors where a better knowledge of system proper ties may bring significant benefits for its management. Additional channel cross-sections, measurements of inflow to the pumping station, and a verified Pump 1 performance curve are identified as the most important data for further model improvement and validation.