Scenario-Based Diagnosis of Urban Flood-System Responses with FLOOD-ST: A TELEMAC-2D Case Study

Conventional urban flood assessments primarily characterise inundation under predefined events but provide limited information on how controlled changes in hydrometeorological conditions, antecedent catchment state, and infrastructure operation alter the spatial and temporal response of a flood system. This study operationalises Flood-Oriented Operational Stress Testing (FLOOD-ST) in the Roßbach sub-catchment of the Emscher catchment, Germany, using an established TELEMAC-2D rain-on-grid model. A deliberately selected ensemble of 23 scenarios was evaluated through 45 direct matched contrasts in which one investigated condition changes while the remaining scenario settings are held constant. Flood-system response was assessed using inundation extent together with hydraulic and affectedness-based indicators for streets, buildings, and critical infrastructure, including maximum simultaneous and cumulative affectedness, timing, recurrence, duration, and spatial redistribution. Within the investigated scenario space, changes in hydrometeorological event type and antecedent moisture condition produced comparatively pronounced differences in affectedness magnitude and timing; for example, changing from the stationary cloudburst to the moving convective storm increased cumulative affectedness by 14.90% for streets, 15.52% for critical infrastructure, and 5.63% for buildings. Infrastructure-state perturbations generally produced smaller aggregate differences but could substantially redistribute affectedness in space and time. In the BL05–CS11 matched clogging-state contrast, net inundation increased by 34,308 m2 while gross spatial turnover reached 73,104 m2; at building level, a net increase of 25 affected objects coincided with a gross turnover of 61 buildings. The case study demonstrates the diagnostic value of operationalising FLOOD-ST beyond conventional inundation-area comparison by identifying where, when, and for how long affectedness changes across controlled scenarios. These outputs can support the prioritisation of locations and system states for further inspection, operational review, and detailed risk analysis, while the findings remain conditional on the investigated scenario ensemble and model configuration.