Interannual variations in spring phenology and their response to climate change across the Tibetan Plateau from 1982 to 2013

Lingling Liu, Xiaoyang Zhang, Alison Donnelly, Xinjie Liu

Land surface phenology has been widely used to evaluate the effects of climate change on terrestrial ecosystems in recent decades. Climate warming on the Tibetan Plateau (1960-2010, 0.2 °C/decade) has been found to be greater than the global average (1951-2012, 0.12 °C/decade), which has had a significant impact on the timing of spring greenup. However, the magnitude and direction of change in spring phenology and its response to warming temperature and precipitation are currently under scientific debate. In an attempt to explore this issue further, we detected the onset of greenup based on the time series of daily two-band enhanced vegetation index (EVI2) from the advanced very high resolution radiometer (AVHRR) long-term data record (LTDR; 1982-1999) and Moderate Resolution Imaging Spectroradiometer (MODIS) Climate Modeling Grid (CMG; 2000-2013) using hybrid piecewise logistic models. Further, we examined the temporal trend in greenup onset in both individual pixels and ecoregions across the entire Tibetan Plateau over the following periods: 1982-1999, 2000-2013, and 1982-2013. The interannual variation in greenup onset was linked to the mean temperature and cumulative precipitation in the preceding month, and total precipitation during winter and spring, respectively. Finally, we investigated the relationship between interannual variation in greenup onset dates and temperature and precipitation from 1982 to 2013 at different elevational zones for different ecoregions. The results revealed no significant trend in the onset of greenup from 1982 to 2013 in more than 86 % of the Tibetan Plateau. For each study period, statistically significant earlier greenup trends were observed mainly in the eastern meadow regions while later greenup trends mainly occurred in the southwestern steppe and meadow regions both with areal coverage of less than 8 %. Although spring phenology was negatively correlated with spring temperature and precipitation in the majority of pixels (>60 %), only 15 % and 10 % of these correlations were significant (P < 0.1), respectively. Climate variables had varying effects on the ecoregions with altitude. In the meadow ecoregion, greenup onset was significantly affected by both temperature and precipitation from 3500 to 4000 m altitude and by temperature alone from 4000 to 4500 m. In contrast, greenup onset across all elevational zones, in the steppe ecoregion, was not directly driven by either spring temperature or precipitation, which was likely impacted by soil moisture associated with warming temperature. These findings highlight the complex impacts of climate change on spring phenology in the Tibetan Plateau.