Predicting the Venice tide using State Space models and the Kalman filter

The core idea of this research is addressing the need for enhanced tide level forecasting inside the area of the Venice Lagoon. Multiple literature studies employ various approaches to model hydrometeorology. The objective is to review the conventional method and propose alternative approaches to improve the forecasting framework in this field. The main approach involves the decomposition of the observed tide level, into different points inside the Venice Lagoon into its two main components: The non-stochastic part given by the astronomical tide which level depends on semi-fixed value given by the position of the sun and moon based on the point in which the measure is taken and the highly stochastic weather driven component. The idea is to perform this decomposition using a State-Space model, allowing for a robust representation of component dynamics. The model state is continuously updated and optimized using a Kalman Filter This specific application introduces an innovative pathway for tracking and projecting the influence of the tidal component.