Abstract: The future precipitation change in the Belt and Road region in 2020—2099 is estimated under different representative concentration pathway (RCP) scenarios. The monthly precipitation under different RCP scenarios of global climate model (HadGEM2-ES) is first of all corrected using four transfer functions of the quantile mapping method for a more reliable estimation result. On such basis, the change trend and regional differences are examined. The root mean square error (RMSE) and precipitation deviation (diff) are analyzed to compare the correction efficacy of quartile mapping method. Results demonstrate that the correction effect of linear function PTF1 reaches the optimum. The uncorrected HadGEM2-ES model overestimates the trend of precipitation increase in Central Asia in the RCP2.6 scenario; the RCP4.5 scenario overestimates the increase in precipitation in Central Asia, East Asia and Russia; the other two scenarios RCP6.0 and RCP8.0 overestimate the increase in precipitation in East Asia and Russia. In the four scenarios after correction, the precipitation increases in East Asia, Southeast Asia and Russia, while mostly decreases in North Africa and West Asia. In future, the average annual precipitation is generally increasing, with the average value climbing in East Asia and Russia, whereas declining in West Asia and North Africa. The abrupt change in precipitation differs with scenarios: RCP2.6 scenario has the least abrupt change, while the other three scenarios perform diversely in different regions.The analysis results provide an effective reference for future policy formulation or project construction.

Key words: precipitation change, RCP scenario, the Belt and Road, HadGEM2-ES model, quantile mapping