Stable water isotopes of precipitation in China simulated by SWING2 models

Abstract

The stable water isotope ratio in precipitation is a useful tracer of atmospheric circulation. Such observations, however, are very limited in space and time. To solve this problem, many isotope-enabled general circulation models (GCMs) are used to help the interpretation of isotope proxies. In this paper, several isotope-enabled GCMs released by the second Stable Water Isotope Intercomparison Group (SWING2) were selected to assess the spatial pattern of deuterium (δD) and the deuterium excess (d) of precipitation in China. The isotopic data of the Global Network of Isotopes in Precipitation (GNIP) and the Chinese Network of Isotopes in Precipitation (CHNIP) were also applied to verify the simulations. The results indicate that these models accurately simulate the spatial characteristics of δD and d of precipitation in China. The correlation between the observations and simulations for LMDZ is the highest among these models, while the root-mean-square (RMS) and standard deviation are not perfect. In addition, LMDZ is worse than other models in capturing the low signal in certain regions, such as CAM, GISS_E, and MIROC. For the monthly variation, most SWING2 models underestimate δD of the precipitation but overestimate the value of d, except for isoGSM. The simulated monthly variation of the water isotopes from SWING2 models is in general similar to the observations, and the trend corresponds to the monthly variation in the Northern Hemisphere. Moreover, all models are good at illustrating the temperature and precipitation amount effects, while they exhibit varying skills in interpreting the altitude and continental effects.