Uncertainty of Vegetation Green-Up Date Estimated from Vegetation Indices Due to Snowmelt at Northern Middle and High Latitudes

摘要: Vegetation green-up date (GUD), an important phenological characteristic, is usually estimated from time-series of satellite-based normalized difference vegetation index (NDVI) data at regional and global scales. However, GUD estimates in seasonally snow-covered areas suffer the effect spring snowmelt on NDVI signal, hampering our realistic understanding responses to climate change. Recently, two snow-free indices were developed for detection: phenology (NDPI) greenness (NDGI). Both found improve detection presence snowmelt. these tested several field camera sites carbon flux sites, a detailed evaluation their performances large spatial scale still lacking, which limits applications globally. In this study, we employed NDVI, NDPI, NDGI estimate northern middle high latitudes (north 40° N) quantified snowmelt-induced uncertainty estimations three (VIs) by considering changes VI values caused Results showed that compared with both NDPI accuracy estimation smaller below 55° N, but higher (55°N-70° N), all exhibit substantially larger uncertainty. Furthermore, selecting use depends types. All performed much better deciduous forests, especially well (5.1 days uncertainty). arid semi-arid grasslands, are more reliable (i.e., uncertainty) than NDP-based (e.g., vs. 4.3 d 7.2 Mongolia grassland 6.7 9.8 Central Asia grassland), whereas American prairie, performs slightly (GUD 3.8 4.7 d). central western Europe, acquired only those years without snowfall before green-up. This study provides insights into application of, in, scales, particularly seasonal snow cover.