| Precipitation plays an important role in the process of water cycle.Stable isotopes in precipitation which are very sensitive to record the evolution in water cycle are natural tracers describing hydrological processes.The Tianshan Mountains located in the northwest of China are the typical wet islands in the arid central Asia,considered to be the dividing line of Nouth and South of Xinjiang.The Karlik Range is located in the east of the Tianshan Mountains.In recent years,there is an integrated observation network covering the entire region,including more than 20 stations from the plains to the mountains.But the observation time is not long,only about one year.So we add six GCMs and the data of ?18O from the Miaoergou ice core to study the interannual variations of ?18O from each GCMs and ice core.With correlation analysis,Morlet complex wavelet method,former least squares regression,the trend of six simulations of isotope-enabled GCMs and the ?18O data from ice core are analyzed.The simulated ?18O by GCMs model interpolated by Cressman method and Lagrange solution,the most suitable simulated model is found at the Karlik Range and surrounding areas.And then to analysis the moisture source from the modle.The main results are as follows:The value of ?18O in summer is higher than that in winter,the seasonal variation presents an obviously changing trend.The simulation result from the GCMs on an average annual time scale has the ‘temperature effect' in the Karlik Range.However,there is no significant change in the value of ?18O on an interannual time scale.The correlation between oxygen isotope composition and surface air temperature on interannual timescale is lower than on seasonal timescale,and the correlation coefficient is between 0.6 and 0.9.The seasonal variation of heavy isotopes in precipitation presents an obviously changing during the whole sampling periods.The precipitation is less from January to March and the temperature is low from December to February determine the poverty degree of ?18O and ?D in precipitation,but the value of d-excess shows increase trend.The precipitation increases from June to July and the temperature is also higher determine the enriched degree of ?18O and ?D in precipitation.The correlation between d-excess and relative humidity is only slight but not significant at the central of the Karlik Range.Compared with global meteoric water line(GMWL)and the local meteoric water line(LMWL),the slope and the intercept of the meteoric water line in eastern Tianshan Mountains are smaller.The value of ?18O in ice core and the temperature in Hami station all show decreasing trend.According to the wavelet method,the influence of ?18O in the ice core is also related to other factors.The correlation and partial correlation coefficient between the ?18O data of six GCMs interpolated by Cressman method with the ?18O data in ice core.The correlation coefficients in iso GSM(NCEP),LMDZ(ECMWF),LMDZ(free)and MIROC(free)model are more larger than 0.8,and are statistically significant at the 0.01 level.The correlation coefficients in LMDZ(ECMWF)and LMDZ(free)models are the highest than the others(r1=0.88).The partial correlation coefficient values in MIROC(free)model are higher than thoes in other models and are statistically significant at the 0.01 level.In conclusion,the MIROC(free)model is the most suitable model to simulate the monthly ?18O data at the Karlik Range and surrounding areas.The trend of MIROC(free)model is similar to the Miaoergou ice core(a=-0.01).After the linear trend and correlation coefficient areanalyzed in this paper,and MIROC(free)model is the most suitable model to simulate the ?18O inthe precipitation in the Karlik Range.From the observation of vapor source trajectory,the increased vapor from the high-latitude area resulted in the significant decreasing trend of ?18O in precipitation,and the increased vapor from the mid-latitude area leads to many growth fluctuations of the ?18O. |
PDF Full Text Request