用树轮δ^(13)C重建1685年以来的大气CO_2浓度变化趋势

RECONSTRUCTION OF ATMOSPHERIC CO_2 CONCENTRATION CHANGES SINCE 1685 BY TREE RING δ^(13)C ANNUAL SERIES

对采自浙江西天目山地区的3株柳杉树盘,交叉定年后,测定了3株树轮δ13C的年序列,分析了3株树轮δ13C序列中所含的共同环境变化信息。用二项式拟合法去除气候因素引起的3个δ13C序列的高频变化,得到3个低频变化序列。分析了theLowDome冰芯记录的大气CO2浓度与树轮δ13C序列低频变化趋势的关系,建立了相应的转移函数,重建了天目山地区1685年以来大气CO2浓度变化。结果表明:用3株树轮δ13C序列重建的结果有较好的一致性,并与南极冰芯的记录及前人研究结果有很好的吻合。这一结果表明用同一地区不同树木个体的树轮δ13C序列的低频变化序列可以重建出基本一致的大气CO2浓度变化历史。

The annual series of δ^13C in three Cryptorneria fortunei Hooibrenkex Otto et Dietr （CF） tree-ring dishes collected from the western Tianmu Mountain, Zhejiang, Province, based on cross-dating tree ring ages, were determined. In the common period for three tree ring δ^13C annual series, statistical analysis showed that they were all correlated with correlation coefficients r1, 3 = 0.65, r1, 2 =0.47 and r2, 3 = 0.52 （n = 146）, respectively, at significance level of α = 0.001. This indicates that the three tree ring δ^13C annual series had similar a high frequency which reflected the effect of climatic factors, and low frequency which recorded the effect of the increasing atmospheric CO2 on the three tree ring δ^13C annual series. After filtering out the high-frequency variation affected by climatic changes by using binomial fitting methods, the low-frequencies were obtained from the δ^13C time series of the three tree-rings. Then, we analyzed scattering diagram between the estimated values of the δ^13C series from the equations （2）, （3）, and （4） and the atmospheric CO2 concentration in air extracted from the Law Dome and DSS ice cores. Based on the distribution characteristics of the scattering diagram and curve estimation, we demostrated the regression equation for reconstructing atmospheric CO2concentration using the least squares and according the atmospheric CO2 concentration measured from the Law Dome and DSS ice cores between 1840 and 1978 and the δ^13C values modeled by the equations （2）, （3）, and （4）. Using the regression equation for reconstructing atmospheric CO2concentration, we reconstructed the atmospheric CO2 concentration changes since 1685. The results showed that the three reconstructed atmospheric CO2 concentration series exhibit a similar increasing trend and agree well with those determined from the Siple ice core record. The reconstructed atmospheric CO2 concentration using the δ^13C series performed the best in reflecting the long-term variation trend of atmospheric CO2 concentration both before and after industrial revolution. The atmospheric CO2 concentration series reconstructed by the δ^13C time series increased slowly before 1840 and significantly after that time. The evaluated results show that the atmospheric CO2 concentration varied in the range of 277-280mol/L between 1685 and 1840, and then exhibited a more rapid rising trend after 1840 with 312.48mol/L in 1940, 318.05mol/L in 1958, and so forth. The fitting precise of the reconstructing equation and the test for independent samples indicate that it is feasible to reconstruct the earlier series of atmospheric CO2 concentration using this method if there are longer δ^13C time series of tree rings. The reconstructed results also indicate that the basically agreement atmospheric CO2 concentration variation trend may be reconstructed by using the δ^13C time series of tree rings from different individuals in the same region.