研究青藏高原高寒草甸生态系统碳水通量及水分利用效率（WUE）特征对准确评估气候变化背景下高寒草地生态系统的碳收支、水循环及碳水耦合具有重要意义。本文利用 2012-2017年青藏高原东部中国科学院西北生态环境资源研究院若尔盖高原湿地生态系统研究站玛曲观测点高寒草甸下垫面的涡动相关系统的观测资料，分析了生长季碳水通量、WUE的变化特征，并结合多元逐步回归和结构方程模型两种统计方法，得到了生长季碳水通量和WUE变化的主要驱动因子。结果表明：（1）玛曲高寒草甸生态系统 6 年的年平均净生态系统碳交换（NEE）、生态系统呼吸（Re）和总初级生产力（GPP）分别为-109. 7、798. 6和 908. 3 gC·m-2·a-1，整体呈碳汇；蒸散（ET）6年年平均值为 446. 5 kg·m-2·a-1；WUE 6年平均值为2. 0 gC·kg-1。（2）生长季NEE和GPP日变化呈现明显的单峰型，在一日14:00（北京时，下同）左右达到峰值，Re全天变化较为平缓，夜晚略低于白天；ET日变化呈单峰型，月平均值和月累积值在7月达到最大；WUE日变化呈不对称的“U”型曲线，最小值出现在13:00-14:00，WUE日值和日变化幅度在7、8月较大。（3）在生长季，多元逐步回归和结构方程模型均证实了温度在控制碳通量中的主导作用，以及辐射在控制ET中的主导作用，温度和太阳辐射是生长季WUE的主要影响因素。

Research on the characteristics of carbon and water fluxes and water use efficiency（WUE）in the al‐ pine meadow ecosystem of the Qinghai-Xizang Plateau is of crucial significance for accurately assessing the car‐ bon balance，water cycle，and carbon-water coupling of the alpine grassland ecosystem under the background of climate change. In this study，based on the observation data made using eddy correlation in the alpine meadow on the eastern Qinghai-Xizang Plateau from 2012 to 2017，from the Maqu observation point of the Zoige Wet‐ land Ecosystem Research Station of the Northwest Institute of Eco-Environment and Resources，Chinese Academy of Sciences，we analyzed the changes in carbon and water fluxes and WUE during the growing season. By combining multiple stepwise regression and structural equation modeling，we explored the main driving factors of carbon and water fluxes and WUE during the growing season. The results indicate that：（1）The average annual net ecosystem CO2 exchange（NEE），ecosystem respiration（Re），and total gross primary productivity （GPP）of the Maqu alpine meadow ecosystem over 6 years were -109. 7，798. 6，and 908. 3 gC·m-2·a-1，respectively，showing an overall carbon sink；the annual average evapotranspiration（ET）is 446. 5 kg·m-2·a-1 the 6-year average water use efficiency（WUE）was 2. 0 gC·kg-1.（2）The daily variations of NEE and GPP during the growing season showed a distinct unimodal pattern，peaking around 14:00（Beijing Time，the same as after），while Re showed a relatively flat diurnal pattern，lower slightly at night than during the day；the daily variation of ET exhibited a unimodal pattern，with peak monthly and monthly accumulative values in July；WUE dis‐ played an asymmetric "U" curve with the minimum value at 13:00-14:00，showing significant daily and monthly variations in July and August.（3）During the growing season，both multiple stepwise regression and structural equation models confirmed the dominant role of temperature in controlling carbon fluxes and radiation in control‐ ling ET. Temperature and radiation were identified as the main influencing factors for WUE during the growing season.