作者: Aleixandre Verger , Iolanda Filella , Frédéric Baret , Josep Peñuelas , None
DOI: 10.1016/J.RSE.2016.02.057
关键词:
摘要: Land surface phenology derived from remotely sensed satellite data can substantially improve our macroecological knowledge and the representation of in earth system models. We characterized baseline vegetation at global scale GEOCLIM climatology leaf area index (LAI) estimated 1-km SPOT-VEGETATION time series for 1999-2010. The phenological metrics were calibrated over an ensemble ground observations timing unfolding autumnal colouring leaves. start end season best identified using respectively 30% 40% threshold LAI amplitude values. accuracy metrics, evaluated available birch forests Europe (and lilac shrubs North America), improved as compared to those MODIS-EVI produced overall root mean square error 7 days (19 days) season, 15 16 length season. spatial patterns agreed well with -NDVI, although start, end, differed by about one month scale, higher uncertainties areas limited seasonality signal systematic biases due differences methodologies datasets. was spatially consistent distributions climatic drivers biome land cover.