Abstract:

Based on five global climate datasets from IPCC AR4 and the observational precipitation data at 162 meteorological stations in the Northeast China, precipitation simulated by five global climate models and the multi-model ensemble (MME) was evaluated. Precipitation change was predicted and assessed under three emission scenarios, i.e. SRES B1, A1B and A2. The results show that monthly precipitation could be well simulated by the global climate model in the Northeast China, while simulated values are systematically wetter than observed values. The MME could well simulate spatial distributions of precipitation, while simulated center lies to the north of its observed position and simulated precipitation intensity is stronger than observed values. Also, the simulated precipitation using this model is more accurate in summer than in winter in the Northeast China. According to the projection, precipitation could increase in the 21st century under three emission scenarios. Increasing amplitude of precipitation is higher in the later 21st century than in the early and middle 21st century, so is in winter than in other three seasons. For three emission scenarios, increasing amplitude of precipitation under SRES A2 are as much as that under SRES A1B, while both are more than that under SRES B1. Distributions of precipitation are almost similar in all scenarios. Increasing amplitude of precipitation is the largest around the Bohai Sea region, and it will increase by 7% in the early and middle of 21st century and 16% in the later of 21st century under SRES A2.

Key words: Global climate model, Multi-model ensemble mean, Precipitation, Model evaluation, Scenarios projections