Understanding the uncertainty of the long-term precipitation trend under global warming through the water cycle

It is certain that globally the troposphere has warmed since the industrial revolution because of the greenhouse gas emissions. The water vapour content of air increases as the climate warms. However, there is nonsignificant long-term trend in global-scale precipitation. This study aims to investigate the sources of the uncertainty of the long-term precipitation trend under global warming through the water cycle. The data of the historical simulations (1850-2005) from 30 CMIP5 models are used in the estimate. The p-value is utilized to indicate the significance of the long-term trend of the water-vapour-related variable in the water cycle as well as the air temperature near surface and at the condensation level. For each variable, the p-value is calculated for each of the grid points. The overall spatial distribution patterns of the p-values are compared between two variables that linked directly, and a point-to-point comparison is also performed to the two fields. Results show that over the globe, the near-surface specific humidity increases significantly at most of the grid points while the increase of the specific humidity at the condensation level is significant at some, and, however, the long-term trend of the relative humidity at the condensation level is majorly nonsignificant. This suggests that as assessed from the global grid points, the uncertainty of the long-term trend of the water-vapour-related variable enlarges substantially in the process of condensation, and modestly in the process of the transport of water vapour. Therefore, the uncertainty of the long-term precipitation trend under global warming is mainly attributed to the condensation and slightly due to the transport of water vapour. The results obtained from the 30 models are consistent with each other, which supports the reliability of the conclusion.