Variation of δ~(18)O in precipitation and its response to upstream atmospheric convection and rainout: A case study of Changsha station, south-central China

摘要

The results of analyses of the stable isotopes of oxygen in precipitation (delta O-18(p)) are presented for every rainfall event from January 2010 to December 2017 in Changsha, south-central China. Our aims were to elucidate the variations of delta O-18(p) on different timescales and to identify the main meteorological drivers of variations in the oxygen isotopic composition of precipitation. Results showed that there were no statistically significant and consistent negative correlations between delta O-18(p) and local precipitation amount on either daily or monthly timescale; however, changes in delta O-18(p) in Changsha responded sensitively to the variation of precipitation in the key upstream area along air mass trajectories. Year-to-year, the strongest negative lagged correlations (r') between delta O-18(p) and the preceding average precipitation amount varied from -0.79 to -0.63 (all significant at the 0.001 level) in the warm half-year (from April to September) of 2010-2017. However, in the cold half-year (from October to the following March), corresponding r' values varied from -0.79 to -0.38 that were all significant at the 0.001 level, except for the year 2012. These findings suggest that the amplitude of the isotopic signal was closely linked with the position and intensity of upstream rainout activity. This was supported by strong relationships between precipitation-weighted mean delta O-18 (delta O-18(w)) and average precipitation amount in the key upstream area at the monthly scale; correlation coefficients were -0.76 and -0.57 between 2010 and 2017 in the warm half-year and cold half-year, respectively (both significant at the 0.001 level). Results advance our understanding of the temporal variation of the stable oxygen isotopic composition of precipitation, and demonstrate that local isotopic proxy records may be influenced by upstream rainout processes. (c) 2019 Elsevier B.V. All rights reserved.