Facing storm surges in Venice: operational system and uncertainty

Dealing with storm surges always involves working with forecasting systems, as they can provide crucial information on the evolution of sea level phenomena, fundamental for coastal flooding risk prevention.
This is one of the main reasons why ISPRA developed and continuously updates an integrated operational monitoring and forecasting system, specifically focused on the North Adriatic Sea and the Venice Lagoon. The system consists of three integrated components: the monitoring system, which is based on the National Sea Level Measurement Network (36 stations) and the North Adriatic and Venice Lagoon Sea Level Measurement (33 tide gauges); the forecasting component, based on the deterministic hydrodynamic finite element numerical model SHYFEM; the probabilistic module for evaluating uncertainty in sea level predictions. The numerical model provides 8 different forecasts each day (up to 144 hours), depending on the spatial resolution, the input meteorological data (ECMWF or BOLAM) and the assimilation of real time observed data. Furthermore, the statistical bayesian processor (Model Conditional Processor MCP), recently integrated in the operational chain, estimates the forecast uncertainty in terms of the probability of an event exceeding a fixed threshold; in the first version (v1) it has been directly applied to the total tidal height, while in a new recent development (v2) it has been provided with information regarding only the meteorological component that predominantly determines the uncertainty, and thus improving the performance. These two information (i.e. the deterministic value and the probability), when combined, make a significant difference, as they provide decision-makers with a deeper understanding of whether or not to take action during a storm surge. In other words, the decision triggering threshold will not be based only on different sea level thresholds (warning level, alert level, flooding level), but rather on different probabilities of a threshold to be overtopped. A very detailed analysis has been carried out to better understand the performance of different model configurations, particularly during the storm surges that occurred between January 2022 and April 2024 (approximately 50 events), and the preliminary results are presented in this study.
Finally, the ex-post analysis of one of the most impactful and recent events (November 22nd, 2022, with sea levels reaching up to 200 cm in the North Adriatic Sea) is presented here to highlight the need for a widespread distribution of measurement stations and very high-resolution forecasts. These are essential to allow a detailed analysis of the effects at both large and small scales (e.g., the lagoons of Venice, Marano-Grado, and Sacca degli Scardovari – Po river delta), even in closely located areas. Once again, the integration and improvement of in situ observations and the modeling system provide a virtuous example of efficiency and functionality in the prevention and mitigation of the impacts of floods and extreme weather-marine events on the coastal environment.