不同时间尺度下降水中氧稳定同位素的时空变化

Comparisons on seasonal and annual variations of δ18O in precipitation

The spatial and temporal variations of stable oxygen isotope in precipitation on different time scales are analyzed according to the data from the IAEA/WMO stations with long survey series in the Northern Hemisphere. Temperature effect is mainly distributed in mid-high latitudes on seasonal scale except for Bamako and Addisababa stations. The δ18O/temperature slope displays the positive correlation against altitude for most of the statistical stations. Amount effect appears primarily in the region south of 30°N and coastal areas. The δ18O/precipitation slope is indirectly proportional to precipitation amount. For some of the sampling stations at mid-high latitudes where their seasonal distribution of precipitation is contrary to that of temperature, coupled with temperature effect, the amount effect appears synchronistically. Either the temperature effect or the amount effect on seasonal scale, there are positive correlations to a certain extent between the annual weighted mean δ18O and the annual mean temperature for almost all the stations. The correlation between composite δ18O and temperature on spatial scale is much more marked, compared with that of individual station. There is a good agreement between 10-year moving average temperature curves I and Ⅱ, with the values of the former all markedly smaller than corresponding ones of the latter, calculated by the monthly mean series group I and the annual mean series group Ⅱ, respectively. However, two calculated dδ18O/dT curves display the distinct difference: the variation amplitude of slope series Ⅱ is larger than that of slope series I. Both curves had similar ascending trend from the 1960s to the 1970s, and then, their variations display the anti-phase. Moreover, the analyses show that there is negative correlation between slope series Ⅱ and temperature series Ⅱ. However, the status is different for slope series I and temperature series I. Both series have contrary trend from the 1960s to the 1970s, whereas the same trend since the 1980s.