
PhDTalks | Uncertainty quantification and spatial biases assessment in precipitation forecasts: a methodology for real-time flood forecasting applications
March 18 @ 17:00 - 18:30

The next appointment in the PhDTalks series will take place on Tuesday, March 18, in the Fassò lecture hall (Building 4A), from 5:15 PM to 6:30 PM CET.
PhDTalks is a series of seminars and discussions among PhD students. The events aim to provide a space for networking among PhD students and connecting with the many projects developed within our department.
The speaker, Enrico Gambini, will lead a seminar titled “Uncertainty quantification and spatial biases assessment in precipitation forecasts: a methodology for real-time flood forecasting applications.”
At the end of the event, a small refreshment will be available, funded by the department.
It will also be possible to attend the conference online at the following link.
Abstract
Accurate meteorological forecasts are essential for mitigating flood impacts, but uncertainty in convective precipitation positioning may significantly undermine flow predictions in small watersheds. A methodology is proposed to assess spatial biases in convection‐permitting model rainfall forecasts, the procedure identifies preferential misplacement directions made by the meteorological model and incorporates this uncertainty into hydrological predictions. The approach was conducted on 64 significant convective rainfall events in the “hydraulic node of Milan” (Northern Italy) during 2013–2022. Quantitative precipitation forecasts from the MOLOCH (ISAC-CNR) meteorological model were compared with observed rainfall fields. Displacement errors were quantified using pattern matching, revealing a consistent northeast bias. Kernel Density Estimation was then used to derive a bidimensional displacement probability density function, this enables probabilistic forecast generation to quantify rainfall misplacement uncertainty effects on hydrological forecast.
Speaker’s bio
Enrico holds a M.Sc in Environmental and Land Planning Engineering, obtained at Politecnico di Milano in 2021. He is currently a PhD candidate (37-th cycle) in Environmental and Infrastructure Engineering, in his PhD he is collaborating with the civil protection of Lombardy Region (CFMR) doing research on the enhancement of regional flood early warning system. His research is mainly focused on the use of machine learning algorithms for flood forecasting, nowcasting and flood warning applications, as well as uncertainty quantification of meteorological predictions.
In his spare time, he enjoys playing chess (with poor results), hiking, climbing and skiing.