Quantifying snow controls on vegetation greenness

Abstract

Snow is a key driver for biotic processes in Arctic ecosystems. Yet, quantifying relationships between snow metrics and biological components is challenging due to lack of temporally and spatially distributed observations at ecologically relevant scales and resolutions. In this study, we quantified relationships between snow, air temperature, and vegetation greenness (using annual maximum normalized difference vegetation index [MaxNDVI] and its timing [MaxNDVI_DOY]) from ground-based and remote-sensing observations, in combination with physically based models, across a heterogeneous landscape in a high-Arctic, northeast Greenland region. Across the 98-km distance from the Greenland Ice Sheet (GrIS) to the coast, we quantified significant inland–coast gradients of air temperature, winter precipitation (using pre-melt snow-water-equivalent [SWE]), and snowmelt timing (using snow-free day of year [SnowFree_DOY]). Near the coast, the mean annual air temperature was 4.5°C lower, the mean SWE was 0.3 m greater, and the mean SnowFree_DOY was 37 d later, than near the GrIS. The regional continentality gradient was eight times stronger than the south-to-north air–temperature gradient along the Greenland east coast. Across this strong gradient, the mean vegetation greening-up period (SnowFree_DOY-MaxNDVI_DOY) varied spatially by 24–57 d. We quantified significant non-linear relationships between the vegetation characteristics of MaxNDVI and MaxNDVI_DOY, and SWE, SnowFree_DOY, and growing degree-days-sums during greening-up (Greening_GDD) across the 16-yr study period (2000–2015). These demonstrated that the snow metrics, both SWE and SnowFree_DOY, were more important drivers of MaxNDVI and MaxNDVI_DOY than Greening_GDD within this seasonally snow-covered region. The methodologies that provided temporally and spatially distributed snow, air temperature, and vegetation greenness data are applicable to any snow- and vegetation-covered area on Earth.