Abstract
Abstract
On 29–30 October 2018, an extreme sea level occurred in the Adriatic Sea, driven by a combination of long-period (storm surge) and short-period (meteotsunami) oscillations. Storm surges were strongest in the northern Adriatic, while shorter-period oscillations dominated in the middle Adriatic, where a meteotsunami was observed and recorded in Vela Luka Bay (Korčula Island). This type of extreme sea level event, arising from the combination of a storm surge and a meteotsunami, is termed a compound extreme and poses both a major threat and a forecasting challenge. Using this event as a case study, we evaluated two different approaches to forecasting compound sea level extremes. The first approach relies solely on numerical modelling: sea level response was modelled using the Semi-implicit Cross-scale Hydroscience Integrated System Model (SCHISM), forced by outputs from the Weather Research and Forecasting atmospheric numerical model (WRF). This approach reproduced the storm surge but failed to reproduce the shorter-period oscillations, including the observed meteotsunami. The second, hybrid, approach combines the first approach for modelling the storm surge component with the shorter-period sea level oscillations obtained from SCHISM forced by the high-frequency component of measured air pressure time series. As the meteotsunami was strongest in Vela Luka, the second approach was tested only at this location. With the second approach, both the storm surge and the magnitude, time of onset and duration of the meteotsunami were successfully reproduced. The results suggest that the second approach could provide basis for an early, site-specific warning system for compound extremes.