Comparison of the Variability of Precipitation and Column Water Vapor Between Satellite Data and Model Simulations

Abstract:

Precipitation plays an important role in the hydrological cycle on Earth. Based on the long-term Global Precipitation Climatology Project (GPCP) V2.2 precipitation datasets, our recent study [Li et al., 2011; Trammell et al., 2015] revealed that the precipitation increased over the wet area (i.e., monthly precipitation > 200 mm) and decreased over the dry area (i.e., monthly precipitation < 50 mm) during the past two decades. To better understand the temporal variations of precipitation and water vapor, we studied the recycling rate of atmospheric moisture, a ratio between precipitation and column water vapor, from 13 CMIP5 models and two observational datasets (SSM/I and GPCP). The comparison between observations and model simulations suggests that most CMIP5 models can capture the two main characteristics of recycling rate: (1) long-term decreasing trend of global-average oceanic recycling rate and (2) dominant spatial patterns of the temporal variations of recycling rate (i.e., increasing in Inter-tropical Convergence Zone (ITCZ) and decreasing in sub-tropical region). All models except one successfully simulate not only the long-term trend but also inter-annual variability of the water vapor column. The agreement between simulations of precipitation in different models is relatively poor, which lead to the discrepancy in the recycling rate between the observations and CMIP5 models. Comparison of spatial patterns also suggests that the CMIP5 models simulate column water vapor better than precipitation. Our comparative studies indicate scope for improvements in the simulations of precipitation, especially for the relatively short time-scale variations, to better simulate the recycling rate of atmospheric moisture, an important indicator of climate change.