Tree-growth responses and four centuries of winter temperature reconstruction in the Gongga Mountains, southeastern Tibetan Plateau

Abstract

Under the intensifying anthropogenic climate change, understanding regional long-term climate change and its mechanism become even more crucial, especially for regions like the Tibetan Plateau (TP) that play an essential role in global atmospheric and hydrological cycle. Climate change studies have been conducted on the TP for decades but were hindered by the scarcity of meteorological data and limited time span. Tree-ring records are valuable in illustrating regional climate change over a long period of time with annual resolution, especially for areas with a lack of long-term climate records. In this thesis work, four tree-ring width chronologies with the longest one spanning 446 years (1569-2014) were developed from three species (i.e., Abies georgei Orr, Sabina tibetica, and Tsuga chinensis) at three upper treeline sites in the Gongga Mountains, southeastern TP. Significant (p < 0.05) positive correlations were detected between tree-rings and minimum temperatures, in particular in prior winter (previous December to current March, pDec-cMar), implying the role of winter freeze stress. However, a loss of sensitivity to winter temperature was observed after 1998, accompanied by a strengthening of positive response to the current growing season temperature and negative response to the previous growing season temperature. The divergence was more evident at high-frequency instead of low-frequency band, probably because the shifting of controlling factor from winter to growing season temperature did not affect the trees’ ability to capture the overall warming trend. Based on the climate-growth relationship, pDec-cMar minimum temperature was reconstructed for the past 367 years (1648-2014), which accounted for 52% of the instrumental variance. The reconstruction exhibits three major cold (1669-1748, 1804-1864, and 1875-1950) and four major warm (1648-1668, 1749-1803, 1865-1874, and 1951-2014) periods in the past four centuries. The extreme cold (1685, 1703, 1741, 1815-1817, 1906, and 1908) and warm (1778, 1796, 1977, 1985-1989, and 1991-2014) winters were also identified. Regional long-term temperature variations in the Gongga Mountains showed similarities but also discrepancies with temperature reconstructions in other parts of the TP. Pronounced warming of regional temperature was identified in the late 20th century with winter temperature rising more rapidly than that in warm seasons and minimum temperature increasing more rapidly than the mean and maximum temperatures. Spatial correlation analysis revealed good spatial representativeness of the reconstruction. In addition, a strong positive correlation was detected between the reconstruction and the Atlantic Multidecadal Oscillation (AMO) indices with the warm/cold periods coincident with the positive/negative phases of the AMO, showing that the AMO may be a factor affecting regional climate in the Gongga Mountains.