Regional-scale normalized difference vegetation index (NDVI) derived from satellite remote sensing observations and gridded climate data were used to study the seasonal responses and underlying mechanisms of vegetation growth over Tibetan Plateau to Pacific Decadal Oscillation (PDO) at period of 1982-2015, by performing Spearman correlation analysis and enhanced multivariate regression model: structural equation model (SEM). The results showed that there was significant negative correlation between PDO index and mean growing-season (April-October) NDVI over Tibetan Plateau; however, marked seasonal divergence in the relationship between PDO and vegetation growth existed among different seasons. It characterized with stronger negative correlation between PDO and NDVI in autumn than in summer, and winter PDO had significant effect on consequent summer vegetation growth. Additionally, it showed great divergence in control processes of PDO on vegetation growth among different seasons, with significant control of PDO on both temperature and precipitation in summer, and significant control of PDO on temperature only in autumn.
利用卫星遥感观测的区域尺度归一化植被指数(NDVI)和格点气候数据,借助Spearman相关分析及基于多变量回归分析的结构方程模型,研究了1982—2015年青藏高原植被生长季节变化对太平洋10年际涛动(PDO)的响应格局及机理过程.结果表明: 青藏高原生长季(4—10月)平均NDVI与PDO指数存在显著的负相关关系,但是PDO与不同季节NDVI之间的关系呈现出明显的季节分异,具体表现为PDO与秋季NDVI的负相关关系强于夏季,且冬季PDO显著影响次年青藏高原夏季植被生长.另外,PDO对青藏高原植被生长的调控过程在季节间存在明显分异,夏季表现为PDO对温度和降水的共同调控,而秋季则以对温度调控为主.
Keywords: PDO; Tibet Plateau; climate change; seasonal response; structural equation model (SEM); vegetation growth.