温度的定量化重建是古气候变化研究中的一个关键和难点，也是未来气候预测的前提基础。本项目选择细菌支链GDGTs（bGDGTs）为温度代用指标，建立区域转换函数为量化方法，通过对比分析云南480个湖泊表层沉积物和表层土壤中bGDGTs的分布特征以及与环境因子的关系，查明云南地区控制湖泊沉积物中bGDGTs分布的主要环境因子，确定5-、6-甲基bGDGTs响应温度的差异性和灵敏性；筛选出湖泊沉积物中bGDGTs以自生来源为主的湖泊，结合器测资料，运用统计学方法建立适用于云南湖泊的bGDGTs-温度转换函数。再将该转换函数应用于有较好研究基础的天才湖，高分辨率定量重建不同时间尺度（60年以来、2000年以来、1万年以来）温度序列，与已发表的不同转换函数方法、不同温度代用指标重建的温度记录进行对比验证，评估新建转换函数的可靠性和适用范围。该研究可为理解我国西南季风区气候变化的动力学机制提供科学依据。

A regional network of quantitative reconstructions of past climate variability is required to test climate models. This project aims to develop a temperature calibration model based on the relative abundances of lacustrine sedimentary branched glycerol dialkyl glycerol tetraethers (bGDGTs) and the instrumental temperature. To address this we studied the surface sediments of 480 lakes and soil from Yunnan, Southwestern China, to: 1) select proper lakes for calibration; 2) quantify their bGDGTs composition and investigate the environmental controls on bGDGTs composition; 3) confirm the difference and sensitivity of 5- and 6-methyl isomers in response to temperature; and 4) develop a regional bGDGTs–temperature calibration model for inferring past temperatures from lakes in Yunnan. Furthermore, we test the applicability of the regional temperature calibration model based on a well-dated sediment core on an alpine lakes in Yunnan, Tiancai Lake. This lake was chosen because they have paleotemperature reconstructions derived from proxies independent of branched GDGTs based on a site-specific temporal calibration between down-core analyses of brGDGTs and instrumental data during past 2000 years, and chironomid-inferred temperature changes based on regional temperature calibration model during past 11800 years, allowing us to directly compare our reconstructions to those derived from other proxies. The results not only provides a methodology for quantitative reconstructions of regional temperature, but also acts as proofs for understanding the dynamic mechanism of climate change in southwest monsoon region.