Abstract: The maximum daily precipitation samples are constructed based on the daily observations of 9 meteorological stations from 1961 to 2008 and the multi-model ensembles of the IPCC Fourth Assessment Report in the water source area for the Middle Route Project of South-to-North Water Transfer. The generalized extreme value distribution (GEV) and the generalized Pareto distribution (GPD) are employed to fit the samples. The results show that during 1961-2008, the maximum daily precipitation conforms to both the GEV and GPD, but the simulated results of GPD model are superior to those of GEV model. Then, the corresponding values of different return periods are evaluated. Finally, the extreme precipitation of different return periods are estimated under the A2, A1B, and B1 scenarios. Extreme precipitation events will be more frequent and intense under A2 scenario than under A1B and B1 scenarios, indicating that impacts of the high emissions scenario on future extreme precipitation will be more intense than those of the medium and low emissions scenarios.

Key words: extreme precipitation, generalized extreme value distribution, generalized Parato distribution, climate change scenario