Phenological responses of vegetation to climate, in particular to the ongoing warming trend, have received much attention. However, divergent results from the analyses of remote sensing data have been obtained for the Tibetan Plateau (TP), the world's largest highelevation region. This study provides a perspective on vegetation phenology shifts during 1960-2014, gained using an innovative approach based on a well-validated, process-based, tree-ring growth model that is independent of temporal changes in technical properties and image quality of remote sensing products. Twenty composite site chronologies were analyzed, comprising about 3,000 trees from forested areas across the TP. We found that the start of the growing season (SOS) has advanced, on average, by 0. April-June and August-September minimum temperatures are the main climatic drivers for SOS and EOS, respectively. An increase of 1°C in April-June minimum temperature shifted the dates of xylem phenology by 6 to 7 d, lengthening the period of tree-ring formation. This study extends the chronology of TP phenology farther back in time and reconciles the disparate views on SOS derived from remote sensing data. Scaling up this analysis may improve understanding of climate change effects and related phenological and plant productivity on a global scale.tree rings | cambial activity | plant phenology | climate change | Tibetan Plateau P henology has a profound impact on vegetation growth (1), carbon balances of terrestrial ecosystems (2), and climate change feedback mechanisms (3). The importance of phenology has prompted many studies, mainly using ground-based observations (4-7), which provide useful phenological information at the species level. However, such studies are also quite time-intensive and typically focus on a few individuals in restricted geographic areas, which often limits their applicability to larger spatiotemporal scales. Changes in plant phenology can be detected on larger spatial scales through near-surface remote sensing, using digital repeat photography (8), but this approach remains limited to the stand level. Another commonly used approach is satellite remote sensing, which can cover large areas (9-11); however, this method has yielded inconsistent results on the Tibetan Plateau (TP) (9, 12, 13).The TP, with an average altitude of over 4,000 m above sea level (a.s.l.), covers more than 2 million square kilometers and is strongly affected by ongoing climate change. Due to its vast area, and its position in subtropical latitudes with high incoming solar radiation, changes in vegetation period duration may have major consequences for regional climate and for carbon sequestration in
SignificanceInconsistent results regarding the rate of change in spring phenology and its relation to climatic drivers on the Tibetan Plateau have been obtained in the past. We introduce and describe here an innovative approach based on tree-ring data, which converts daily weather data into indices of the start (and end) of the growing season. This method pro...