Project description
Future predictions of intense rainfall
Changes in climate extremes can be seen and felt all around. An increase in extreme weather events is one example that is also highlighting the importance of future meteorological predictions. While it is crucial for the public to prepare for intense precipitation and related weather extremes, it’s not easy. There is a lack of historical data for statistical models. In this context, the EU-funded PACE project will fill the knowledge gap by developing a physically informed statistical model based on the framework of precipitationshed (the upwind land and ocean area that contributes evaporation to a given location’s precipitation). This will be combined with a prediction of their changes and a new account of dynamics and thermodynamics changing precipitation conditions, probability, and distribution for improved forecasts of rainfall and extremes.
Objective
Characterizing the frequency of future precipitation extremes is a fundamental task in order to understand and mitigate their effects under future climate conditions. However, predicting these changes is a challenging task as intense rainfall is poorly represented in climate models, and extrapolations based on historical data are highly uncertain. The objective of the PACE project is bridging this disconnection by developing an innovative physically-informed statistical model for rainfall extremes. The proposed methodology explicitly accounts for the interannual variability of atmospheric moisture fluxes and their future predicted changes. PACE will achieve this objective by using an innovative precipitationshed perspective. Precipitationsheds - the regions of Earth from which most precipitable water originates before reaching a target site – will be computed using a moisture tracking technique. By including in the statistical representation of rainfall extreme key physical information integrated over a precipitationshed, PACE will provide a novel representation of how dynamic and thermodynamic changing conditions in upwind areas affect the frequency and probability distribution of downwind precipitation. By learning rainfall statistical properties from historical data and relying on climate model predictions only for large scale climate features, the PACE action will improve probabilistic rainfall predictions under future climate scenarios. The Candidate, recently graduated from Duke University (Durham, NC) will join the Host at TU Delft (Delft, The Netherlands) to achieve this research objective, and to continue his postgraduate professional development aimed at creating an independent research program in hydroclimatology.
Fields of science
Programme(s)
Funding Scheme
MSCA-IF - Marie Skłodowska-Curie Individual Fellowships (IF)Coordinator
2628 CN Delft
Netherlands