Impact of initial moisture conditions on the generation of flood events

Extreme events, such as floods, are increasing in frequency and intensity, making it essential to understand the factors that influence them. The objective of this thesis is to analyze the impact of initial soil moisture conditions on the generation of flood events and develop a statistical tool to study the impact of climate change on the magnitude of floods. Specifically, the study is conducted by exploiting rainfall and streamflow records observed in five study sites located in the Veneto region. The analysis starts by defining a bivariate joint distribution of peak flow rate and antecedent cumulative soil moisture. It was possible to determine that the antecedent moisture calculated over the past 5 days is more correlated with the discharge as compared to the antecedent moisture calculated over longer time-windows. Three models were compared to study extreme events: the Generalized extreme value theory (GEV), the Simplified Metastatistical Extreme Value Distribution (SMEVD) based on a single class of data and the moisture accounting SMEVD. In particular, the SMEV distribution is derived from the more generalized Metastistical Extreme Value (MEV) approach, which allows the use of a much larger sample of data than GEV, enabling a better description of extreme events, even in the context of climate change. The moisture accounting SMEVD is a model that allows the fit of flow events by dividing them into classes. In this case, three classes were used, created according to moisture level: dry soil, average wet soil and wet soil. Different methodologies such as Goodness of fit and Cross Validation were used to measure the performance of the models. Goodness of fit showed better performance using the GEV, the moisture accounting SMEVD has lower performance only for the extreme event related to storm Vaia. Cross-Validation, on the other hand, showed better performance using the moisture accounting SMEVD model, highlighting the advantages of using this model over GEV. The moisture accounting SMEV model, can also be used to make assumptions about climate change, in fact in the last phase of the research it was used to hypothesize three climate change scenarios. The first scenario was developed, assuming a 50% reduction in the frequency of events in the soil class with the highest moisture content. The second scenario instead assumes a 15% increase in the scale parameter of the Gamma distribution used to fit the events to the model. The third scenario (the most pessimistic) combines the changes assumed in the first two. The first and last climate change scenarios identify a decrease in magnitude for events associated with low return times, while an increase in magnitude is observed for events with high return times. In contrast, the second scenario shows an increase in magnitude for events associated with all return times.