Interannual trends in stable oxygen isotope composition in precipitation of China during 1979–2007: Spatial incoherence. (1st October 2017)

• Record Type: Journal Article
• Authors: Yang, Sen; Zhang, Mingjun; Wang, Shengjie; Liu, Yangmin; Qiang, Fang; Qu, Deye
• Is Part Of: Quaternary international. Volume 454(2017)
• Page Start: 25
• Page End: 37
• Publication Date: 2017-10-01
• Subjects: GCMs -- δ18O in precipitation -- China -- Interannual variation; Geology, Stratigraphic -- Quaternary -- Periodicals; Stratigraphie -- Quaternaire -- Périodiques; 551.79
• DOI: 10.1016/j.quaint.2017.07.029
• Languages: English
• ISSNs: 1040-6182

Abstract

Abstract: According to the simulations of isotope-equipped general circulation models (GCMs) compiled by the Stable Water Isotope Intercomparison Group Phase2 (SWING2), the interannual variability and trend of precipitation isotopic composition in China during 1979–2007 are studied. Six simulations of isotope-enabled GCMs are involved, and monthly series of stable oxygen isotopes in precipitation for each grid are applied to calculate the linear trends from 1979 to 2007. Most isotope-equipped GCM simulations show an increasing trend in δ 18 O, while the GISS-E (NCEP) result shows a decreasing trend. In most areas of north China and northwest China, the GCM-simulations positively correlate with surface temperature. The correlation coefficients between oxygen isotope composition and surface air temperature generally increase from coast to inland. The correlation between observed and simulated δ 18 O in precipitation of northwest China simulated by LMDZ (free) is best ( r = 0.73) among different subregions and isotope-equipped GCMs. The comparison of GCMs simulations and BW model all show the low value region of δ 18 O is in the Tibetan Plateau and the high value in south China. China's meteoric water line simulated using isotope-enabled GCMs (slopes range from 7.63 to 8.26, and intercepts range from 9.25 to 11.92) is similar to the global meteoric water line, and all the correlation coefficients are statistically significant at the 0.01 level.