The significance of recent and short pluviometric time series for the assessment of flood hazard in the context of climate change: examples from some sample basins of the Adriatic Central Italy

Numerical hydrological models are increasingly a fundamental tool for the analysis of floods in a river basin. If used for predictive purposes, the choice of the “design storm” to be applied, once set other variables (as basin geometry, land use, etc.), becomes fundamental. All the statistical methods currently adopted to calculate the design storm, suggest the use of long rainfall series (at least 40–50 years). On the other hand, the increasingly high frequency of intense events (rainfalls and floods) in the last twenty years, also as a result of the ongoing climate change, testify to the need for a critical analysis of the statistical significance of these methods. The present work, by applying the Gumbel distribution (Generalized Extreme Value Type-I distribution) on two rainfall series (1951–2018 and 1998–2018) coming from the same rain gauges and the “Chicago Method” for the calculation of the design storm, highlights how the choice of the series may influence the formation of flood events. More in particular, the comparison of different hydrological models, generated using HEC-HMS software on three sample basins of the Adriatic side of central Italy, shows that the use of shorter and recent rainfall series results in a generally higher runoff, mostly in case of events with a return time equal or higher than 100 years