A process-based multivariate framework for reservoir safety evaluation for Slovakia

Changing climate and changes in watersheds raise questions about the safety of flood control structures. Evaluating the present safety of past designs where storage was involved requires the entire hydrograph or, at least, the flood peak-volume-shape estimates related to the present or future conditions. Therefore, studying the relationships between these quantities is exciting both from the scientific hydrological and engineering points of view. The statistical analysis of flood peaks and volumes in engineering hydrology practice was often dealt with in a multivariate frequency framework. However, examining the interplay of climatic and catchment processes in influencing the probabilities of flood wave peaks, volumes and shapes is a challenging problem for safety evaluation of flood control structures. The proposed framework for Slovakia is based on the inclusion of climatic and hydrologic controls on the flood peak-volume-wave shape relationships and their statistical models. Rather than examining the safety based on a single control flood wave, a scheme was proposed allowing us to arrive at a set of control flood waves with associated probabilistic parameters. These permit to study of several aspects of reservoir safety (hydrologic, hydraulic and geotechnical). The concept is based on flood events assigned to the prevailing flood process types in the region with discernible flood-peak-volume-shape relationships. A set of process representative flood waves is created by base-flow separation. The joint probability distribution of peaks and volumes is copula-based, which are selected by a process-based approach. Besides that, process characteristic wave shapes are scaled to a set of dimensionless shapes. These allow selecting a dimensionless control flood wave shape according to an empirical shape quantile. Finally, a set of control flood waves is constructed by combining volumes conditioned on peaks and associated with shapes of given quantiles. That provides a good variety of control waves for assessing the safety of water structures under extreme situations in a probabilistic and process-based framework in typical risks of failure scenarios.

Acknowledgements: This work was supported by the Slovak Research and Development Agency under Contract No. APVV-20-0374 and the VEGA Grant Agency No. 1/0782/21.