树轮宽资料所指示的川西过去气候变化

PAST CLIMATE ON WEST SICHUAN PLATEAU AS RECONSTRUCTED FROM RING-WIDTHS OF DRAGON SPRUCE

本文利用采自我国川西高原4个地点的树轮资料，在建立树轮宽指数序列和分析所建序列与气候要素间关系的基础上，重建了川西地区1650～1994年冬季平均最低气温距平序列，并分析了其高频、低频和周期变化的特征。结果指出，在重建的345年中，冷冬发生过22年，暖冬发生过26年；有过5个变冷、5个变暖和1个稳定阶段；19世纪前半叶为暖冬期，并变幅最小；本世纪的后半叶为冷冬期，但自60年代以来的增温显著；重建序列中存在着60a左右的长周期变化。

Well-replicated tree-ring increment cores from Picea ba@uriana were sampled for four sites in western Sichuan Province, China. Standard ring-width chronologies were developed after cross-dating and measurementS of cores using stalldard techniques. Inspection of year-to-year variation of ring-width indices revealed a high degree of agreement among sites indicating that cross-dating among four sites can be done Perfectly, although the distant between tWo farthest sites was about 200 km. Response function and correlation function were utilized to study the response of tree radial growth to climate elements. The results of response function showed that from 48% to 73% of growth variance could be accounted for by climatic variables. Correlation function analysis indicated that the most statistically significant climatic variables correlated tO ring-idth indices were winter and June temperature and that the strongest and most consistent relationships between ring-width indices and climatic variables tended to be for winter mean minimum temperature. Therefore, transfer functions were developed for the regional departUres of winter mean minimum temperature. The Percentage of variance explained by the selected transfer function was 60.4%, and the reduction of error was 0.520 indicating the predictive capability of the regression model.A regional departUre series of winter mean minimtnn temperatUre on WestSichuan Plateau has been reconstructed for the Period A. D. 1650 to 1994. Onenotable feature of the reconstrUchon was the posihve mean for the wholereconstFllChon Period due tO the cold statUs during the calibrahon period. Taking l.5stalldard deviation of the reconstructed series as a criterion for cold and warm winter,it was found 22 cold winters and 26 warm ones in the 345--yr reconstttichon. While,there were only I cold,but 9 warm winters in the 19th centUry, implying a warmcentUry. With 10 clod and 9 warm winters, the 20' centUry exhibited the largestntunber of extrCme years in centllry scale. The variahons of the ZI--yr moving averageof the reconstmchon represenhng the low frequency change displayed several cold andwarm periods, in which the coldest Period aPPeared in 1954-- 1979 and the warmestPeriod in 1712~ 1734. It was also shown that there were 5 bines of cooling, 5 timesof warming, and I level off stage during the reconstahon Period. The present is ina warm winter Period. However, it seems another cooling has started since 1990' s. Inorder to describe quanhtahvely the warm and the cold winters, statishcs such as meanand variability were calculated for each decade, each half centUry and each century.The resultS showed that 1810's was the warmest decade with the smallest variabilityand 1960's was the coldest decade. 1710's was the decade with the largest variability.For the half centUry scale, the warmest Period appeared in the first 50--yr of 19thcentury and the coldest Period in the farest 45 years. The Period 1700-- 1749exhibited the largest variability. From 1650 tO the end of 18' centUry the means havebeen negahve in centUry scale, and the 19' centUry was the warmest centllry. msaught indicate the Little Ice Age and the ending of iL The current centIJry is the onewith negahve mean and the largest variability.