树轮记录的阿尼玛卿山区过去 830年夏半年最高温变化*

利用采自黄河源区阿尼玛卿山的祁连圆柏(Juniperus prezewalskii)样芯,采用国际规范的树轮学研究方法对所有树轮样芯进行了精确定年。针对各样芯不同的生长趋势分别采用负指数函数、线性函数和样条函数拟合与树龄有关的树木生长趋势。对个别在样芯中存在缺失段或缺失年轮过多难以准确判定缺失位置的样芯,采用"搭桥"的方法建立完整序列,最后用ARSTAN软件计算合成树轮宽度年表,最长序列为1532a,在满足复本要求的基础上建立了长度为830a的树轮宽度年表,并对年表的基本统计特征进行了分析。初步的分析结果表明该年表的年轮指数与研究区夏季温度成显著负相关,并在此基础上重建了研究区过去830年来夏半年(4~9月)平均最高温的变化。重建序列包含有大量的高频变化信息,又有十分显著的低频变化趋势。研究区在重建时段内经历了3个较为显著的降温期,其中1820s~1900s的降温趋势不同于该时期全球平均温度的升高趋势。另外,分析表明在重建时段内升温期较短,但是升温幅度大于降温幅度,迅速升温可能导致研究区严重干旱事件的发生。

MAXIMUM TEMPERATURE RECONSTRUCTION FOR ANIMAQING MOUNTAINS OVER PAST 830 YEARS BASED ON TREE-RING RECORDS

Due to the weakness of ecological systems caused mainly by enhanced human activities as well as regional climate changes, the ecological environment in the Yellow River source region is seriously deteriorated. To examine the reasons and mechanism of ecological environment changes in the region, its history should be understood first. Dendroclimatology is one of the important methods for examining the past global climate changes and the tree ring data was used widely as an important archive in paleoclimatic research because of its precise dating, annual resolution, and high correlations with instrumental climate data. Based on tree-ring width and/or density chronologies, long-time-scale climate variations have been reconstructed on both regional and hemispheric spatial scales. The study area, the Animaqing Mountains is located within the Yellew River source region. Cypress (Juniperus prezewalskii) tree-ring cores were taken from living trees from the Animaqing Mountains. Samples were cross-dated with three methods: skeleton-plot cross-dating, visual cross-dating of the ring-width measurement series, and testing of cross-dating using the computer program COFECHA. The oldest tree was sampled as 1532 years old. The cross-dated tree-rings were measured on a Velmex tree-ring width measuring system ( 0.001mm). The computer program ARSTAN was used to develop the chronology. Most tree-ring series were detrended with negative exponential curves or straight line that conforms to the expected biological growth trend model. Few of those that did not fit the negative exponential curves or straight line were standardized using a 67 % smoothing spline. A tree-ring chronology of 830 (1172~2001) years was developed. Before 1172, the expressed population signal (EPS) of the chronology is lower than the recommended threshold of 0.85 because the sample depth is insufficient. The analysis demonstrates that the tree-ring index is correlated negatively with the summer half year (April to September)average maximum temperature significantly.The maximum temperature of the research area was reconstructed based on the correlation analysis model. Test results confirmed the validity of our regression model. The reconstruction contains lots of high frequency signals as well as low frequency trends. The study region has been experienced three significant cooling periods in the past 830 years. One of the cooling periods is during 1820s~1900s, while the global warming is significant. However, the analysis showed that the warming periods are much shorter than the cooling period. While warming speed is higher than cooling, it could lead to severe droughts in the research region. Several warm periods reconstructed coincide with severe droughts that recorded by other archives.