**3. Scenarios**

The selection of the scenarios to be modeled plays a crucial role in the choice of the mathematical models and also on the reliability of the modeling results obtained for the relevant physical phenomena.

Herein, four different modeling scenarios are proposed and detailed for driving the proper implementation of the proposed integrated modeling approach (Table 1): climatological, short-term realistic, long-term realistic and operational/forecasting scenarios.

The climatological approach allows reproducing conditions which are not directly related to a specific series of measured data. They are inferred from observations by means of statistical analysis on available data. The aim is to reproduce either frequent (annual or seasonal) or extreme conditions with given average return periods. It should be stressed that statistical analysis seldom allows defining the return levels of relevant driving forces by taking into account also the marginal probability, i.e., that extreme events of different forcing may occur simultaneously. Usually, this approach is used within the frame of the preliminary modeling phase, when most of the information or data are not available yet.

The shortcoming of the climatological approach may be overcome by employing the short-term realistic approach that considers actually observed driving forces for a short duration time window (event scale). Then, it is possible to take into account the interaction of all driving processes typical of real conditions that can significantly affect the actual dynamics. This method is suitable within the frame of both preliminary information phase and preliminary modeling phase, and detailed modeling phase as well. In the preliminary phases, it allows achieving results with low computational costs useful to depict the big picture of the problem at hand when critical conditions are analyzed. In the detailed phases, it can be considered for either validation purposes or to reproduce extreme events observed in the past. It has to be stressed that this approach may be employed only when detailed measurements are available.

The long-term realistic approach has to be used when long-term effects (or project effects) have to be investigated within the framework of the detailed modeling phase. The definition of the long-term scenarios is then based on real conditions which occurred in the past, for a long duration time window selected as representative of the site-specific conditions. Hence, long-term time series (i.e., years) have to be available by means of either monitoring activities or numerical hindcast (e.g., [28]). The long-term realistic approach allows investigating the probability of exceeding thresholds for the variables of interest (e.g., SSC, see Section 8.1) in terms of combined analysis of intensity, duration and frequency.

During the works execution the operational/forecasting approach can be employed to forecast worst scenarios (in term of weather and sea conditions, and sediment dispersion conditions), within the framework of the environmental monitoring plan. This method can be useful for contractors to optimize the work execution. Indeed, safe conditions may be forecast hours or days in advance (e.g., [29]). Moreover, authorities need to make effective the implementation of the environmental monitoring plan and then to limit environmental effects within the framework of the modeling–monitoring feedback system (see Section 8.1).

Table 1 synthesizes the main features of the four approaches for scenarios selection needed to implement the proposed integrated modeling approach.


**Table 1.** Main features of the modeling scenarios proposed within the different accuracy levels of the proposed integrated approach.
