在山西芦芽山地区采取了符合国际树轮库要求的油松样本,通过交叉定年和应用区域生长模型,建立长度为328 a的标准宽度年表。根据RCS序列所揭示的气候低频变化特征,确定1676 AD以来夏季温度可划分为两个时段:1676—1865 AD和1866—2003 AD。在1676—1865 AD时期,夏季温度变化主要表现为“冷强暖弱”,其中1710—1720s为最冷时段。1866—2003 AD时期,夏季温度呈现出“总体持续变暖,冷暖交替频繁”的变化特征。
A tree-ring chronology of 328 years is obtained by using the Regional Curve Standardization method based on conifer samples (Pinus tabulaeformis) collected from Luya Mountain, Shanxi Province. Correlation analysis shows that the tree growth in the study region significantly positively relates with the mean temperature in May and July and with the precipitation in May; and significantly negatively relates with the precipitation in July; and the mean temperature in May is the more powerful limited factor. Based on the low-frequency climate variation from RCS chronology, the summer (May to July) temperature proxy series could be divided into two phases: the first one was 1676—1865 AD and the other one was 1866—2003 AD. During the former phase, the total variation could be described as "more cold summer and less warm", with warm summers in the 1740s and 1830s, cold summers in the 1710—1720s, 1750s, 1780—1790s, 1810s and 1740s—1850s, and a coldest period in the 1710—1720s. And during the latter phase, the total variation could be described as " continuous warming with high interannual variability ", with warm summers in the 1870s, 1890—1900s, 1930s, 1940s, 1960—1970s and 1990s, and cold summers in the 1880s, 1910—1920s, around 1935 AD, early 1960s and late 1970s. The correlation coefficients rang from 0.62 to 0.79 between RCS chronology and six Northern Hemisphere temperature proxy series, and the highest one is coincident with Jones (2004). As compared with temperature proxy series in North China, Tibetan Plateau and Northern Hemisphere, it is found that the RCS chronology of Luya Mountain not only contains the local temperature sign but also well synchronizes with North China and Global temperature variations. All of these indicate that the RCS chronology has a potential to capture regional climate variability on long timescales.